Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2005-06-09

00:00:10Rorican you manually copy that from daily?
00:00:15Rorior does it not work?
00:00:19ehntooit is there
00:00:25ehntooit's just... not being... detected...
00:00:55BippleSongs are now working again...
00:01:28RoriOgg is playing back on my bleeding edge but no wps
00:01:33Roridid I miss something?
00:01:37preglowthere is no wps
00:01:45Roriehntoo you lie!
00:01:51ehntoocheck the rbx homepage.
00:01:57ehntooand ogg is working?
00:02:07linuxstbI just committed a change to make the WPS appear - but it doesn't actually know anything about OGG/FLAC files, apart from the filename.
00:02:17linuxstbNot even the length
00:02:18Rorioh heh
00:02:20preglowdidn't see that
00:02:36Roriyour too fast ;)
00:02:46ehntooand... ogg isn't working still
00:02:58 Join DMJC-L [0] (
00:02:59RoriI will wait for tomorrows build methinks
00:03:15 Quit Aison ("( :: NoNameScript 3.72 :: )")
00:03:40Rorigapless ogg is sex
00:03:47BippleIs it normal to load a ogg and it plays the song but doesnt go to the next stage where is shows the length etc...
00:03:49preglownot quite
00:04:10preglowi'm starting to think we should add a faq on this...
00:04:23ehntooBipple: yes
00:04:27Rorifaq's take up valuable coding time :)
00:04:29Bipplebloody chipmunks its crashed again
00:04:40ehntoowelcome to the wonderful world of pre-alpha software
00:05:13Rorisomeone has to be at the bleeding edge
00:05:23BippleWelcome to the wonderful world of and back to 1.65EU
00:05:31RoriBipple good idea
00:05:43Bipplei know, im a genius, ill wait till its done lol
00:05:50ehntoo*kicks his iRiver*
00:05:51Bippleor maybe until radio is in
00:05:55Roristay safe be well....come again!
00:06:01ehntoowhy isn't ogg working on my unit?
00:06:16Bippleyou'll be lucky if you can get anything working
00:06:27Roriehntoo what build? the last commit from LinusN?
00:06:36ehntoomost recent
00:06:53ehntoolike, within 5 minutes
00:06:58ehntooit's not giving me an error any more
00:06:59Rorilet me try it
00:07:07ehntooit just starts to play
00:07:11ehntoobut then nothing comes out
00:07:17Rorigimmie a minute
00:07:19ehntooand within half a second, the play symbol changes to stop
00:07:22ehntooRori: k
00:07:28 Join Naked [0] (
00:07:29ehntoomp3 is grand
00:07:36ehntoobut I can't live without voris
00:08:53*Rori sings "Banned on the Run..."
00:09:05RoriBanned of brothers
00:10:48[solid]yet another status update... flac has been playing for 5 hours straight now
00:11:12preglowsnot bad
00:11:17preglowanyone know how long iriver fw can do wav?
00:11:28Roriogg still works here
00:11:33Rorilast commit
00:11:40Rorishows wps too
00:11:46 Quit Naked (Read error: 131 (Connection reset by peer))
00:11:57Roribut no time remain etc as LinusN said
00:12:40Rori@ ehntoo
00:12:46[solid]preglow: ~4hrs
00:12:47ehntoo... weird
00:13:03Quertylinuxstb: Thanks for the generic WPS, that's really cool!
00:13:03Bippleok re flashed to iriver, thank you for your help my good people
00:13:09BippleBye all
00:13:12[solid]the battery indicator is showing 50% now
00:13:19 Quit Bipple ("CGI:IRC")
00:14:03[solid]and to make it even more realistic i've been browsing files twice (to spin up the disk)
00:14:12 Join bipak [0] (
00:14:22ehntoomount /dev/sda1
00:14:24[solid]and i check the runtime using the debug menu (backlight every ~15minutes)
00:14:25ehntooeh... wrong window
00:14:28preglow[solid]: sure about that? if so, this is very cool
00:14:44[solid]preglow: yeah.. it's quite unbelievable if you ask me:D
00:14:52thegeekdoes anyone know where I can get replacements for the 3.5 mm headphone jack use on the h120/h140 ?
00:14:56thegeekmine are loose
00:15:03thegeekso I have to solder in some new one
00:15:21preglow[solid]: we should still be able to make flac run much faster than it does
00:15:29[solid]preglow: yeah i know :DDD
00:15:31linuxstb[solid]: Are you just repeating one track, or playing a long playlist?
00:15:41ehntoobah... not again
00:15:46[solid]linuxstb: i'm repeating a specially selected album
00:15:55preglowas long as it keeps reading the disk, it doesn't much matter
00:16:36linuxstbYep, that's all I was checking.
00:17:03[solid]linuxstb: it has a bitrate of ~800 (ie the whole album), is 55 minutes long and has both tracks that fit in the buffer and tracks that don't
00:17:46linuxstbSounds perfect to me. I'm surprised it's so stable so early.
00:18:14RoriHow long for Ogg and Mp3 at present anyone know?
00:19:15ehntoowhy isn't my iriver showing up as a drive?
00:19:31[solid]Rori: i'll be testing that tommorow (or rather tommorow and the day after tommorow if rockbox continues to perform so well:D )
00:20:15Roriat this rate Rockbox will be complete in a couple of weeks :D
00:21:05RoriHey I like to be optimistic (which is unusal for me) :)
00:21:57ehntooI don't like the way that happened
00:22:59 Quit markun ()
00:23:26[solid]oh the battery indicator indeed isn't that reliable
00:23:42[solid]showing 62 now ;)
00:24:10ehntoognome file-roller wasn't extracting my stuff into the right folder
00:24:20[solid]linuxstb: btw, i used max compression (-8)
00:24:55[solid]heh and back at 50 again
00:25:22ehntoo.. and... after clearing my rockbox install
00:25:26ehntooit's still not playing vorbis
00:25:36[solid]back to tha books... hope to give next status report in an hour ^^
00:27:26*Rori goes for a workout. fight the flab and all ;)
00:29:07ehntooanybody taking suggestions on what they should work on next?
00:29:28 Quit Querty ("Leaving")
00:29:38BagderI don't think that's really needed
00:30:01ehntooas amusing as listening to "This Week in Tech" running at 2x speed is
00:30:03ehntooiot's kinda nnoying
00:30:03Roripeople work on what they want to
00:30:30RoriI am sure it will get sorted whenever anyone feels the need to do it. this is not a business :)
00:30:37ehntooI know
00:30:39ehntooI'd do it myself
00:30:44ehntoobut I don't know any C
00:30:51ehntoonot much more than "Hello World"
00:30:54Bagderperfect time to learn!
00:31:05ehntooaye, I know
00:31:07Roripatience linus and slasheri and preglow etc are on the case no doubt
00:31:21ehntoohey... now that finals are over, now I CAN learn
00:32:12ehntoo*digs out gedit and gcc*
00:33:53 Join muesli- [0] (
00:34:06t0masare you around?
00:34:12t0masor Bagder?
00:34:20Bagderstill here
00:34:25amiconnme 2
00:34:33t0masI found a bug... I think
00:35:00t0mascan you verify with last CVS build, that keyboard input (or the whole ON/PLAY button) doesn't work?
00:35:35muesli-btw is there somewhere a changelog for latest builds?
00:35:45t0masand I guess that's amiconn's / my hold button fix
00:36:02ehntoot0mas: yeah... it is broken
00:36:07ehntoojust noticed that
00:36:15t0masamiconn: take a look at our little fix
00:36:16amiconnt0mas: poweroff is working
00:36:16 Quit jn-comp ("CGI:IRC (EOF)")
00:36:21t0masamiconn: the other one
00:36:32t0masamiconn: where you left out some parenthesis...
00:37:08amiconnHmpf :/
00:37:22t0massee what's wrong there?
00:38:27amiconnYes... you set the parentheses wrong
00:38:51t0masif (!button_hold() && (data & (0x20 == 0)))
00:38:55t0maswhat should it have been?
00:39:13t0masif (!button_hold() && ((data & 0x20) == 0))
00:39:15t0maslike that?
00:39:34t0masdamn... that's a pretty obvious one
00:41:07HClyes, i'd say :P
00:41:19 Quit matsl (Remote closed the connection)
00:41:44t0mas3 commits for something we could've done just right :X
00:42:11t0mas1 my fix, 2 amiconn's parenthesis, 3 my really stupid parenthesis
00:42:34preglowwoot, i just turned 24!
00:42:51Rorihappy bday preglow
00:42:54t0masah, same birthday as my girlfriend
00:42:55Bagdercongrats preglow
00:43:01amiconnt0mas: Pause working again
00:43:04preglowhappy and happy
00:43:09 Join Naked [0] (
00:43:14muesli-lalala preglow :D
00:43:18]RowaN[welcome to club 24 preglow
00:43:26t0masamiconn: fixed 3 bugs at once now ;)
00:43:28ehntoohappy birthday
00:43:30*preglow does the thaking and the bowing
00:43:34Bagderand that's not even 0x24
00:43:34muesli-still 3years to go to join MY club ;)
00:43:35Rorit0mas have you committed it?
00:43:44preglowmuesli-: you'll wait for me, right?
00:43:48t0masRori: amiconn had
00:43:52muesli-promised :D
00:43:54*amiconn enjoys hot-plugging the remote
00:43:55Rorik grabs
00:44:10amiconnNeed to take care not to plug / unplug too often
00:44:15preglowamiconn: commit?
00:44:36amiconnAbout to...
00:45:07t0masok, now I can commit my paused -> next -> not paused fix
00:45:30preglowman, rockbox is shaping up with a quickness
00:45:30amiconnI 'abuse' the scroll thread to do the init as well, cause this can't be done directly within the tick function
00:45:37preglowi hope there's something left for me to do come july
00:45:45 Quit courtc (Read error: 113 (No route to host))
00:45:57t0maspreglow: enough ;)
00:46:20Roriwork on getting wps to work on ogg properly then ;)
00:46:21amiconnpreglow: Rockbox is never finished... we have archos support for years, and still there are things to fix/ improve/ extend
00:46:23t0masand if you're getting bored... I can do some more parenthesis... so you can search them? :P
00:46:32preglowamiconn: yeah, i know, just joking around
00:46:46preglowat least there'll probably be some dsp stuff to be done
00:46:54preglowi still want to fulfil my dream of dsp plugins
00:47:03Rorigood dream
00:47:28Rorinice verb with surround expansion
00:47:32Roriane exciter etc
00:47:37preglowwe'll see
00:48:01preglowi guess the equaliser should be put in the core
00:48:12preglowunless people are content with what the uda has
00:48:21HCli fully agree with dsp plugins
00:48:28HClpreferably with chaining support
00:48:32t0masah, the iriver bass and treble is uda stuff?
00:48:34t0masnot software?
00:48:42Rorishame DPL II headphone support is not open source :P
00:48:44preglowt0mas: yes
00:48:54preglowhardware based
00:49:40amiconnpreglow: Probably the simplest thing (given that the uda can't do that) is a stereo matrix (stereo width setting in rockbox)
00:51:39t0masok, paused -> next -> playing bug is fixed :D
00:52:47t0masdoesn't give warnings here... and pcm_playback.c is iriver only... so it should just work :)
00:52:49preglowamiconn: just some simple mid/side calcualation thing?
00:53:17t0maspreglow: maybe it can so something like it for iriver's srs funcion?
00:53:30preglowi have no idea how srs works
00:53:31amiconnYup... you don't even need to explicitly calculate mid & side signals
00:53:44preglowamiconn: well, no, but you can look at it like a m/s calculation
00:53:45ehntoois iriver's srs in software or hardware?
00:53:52preglowehntoo: software
00:54:38t0masok, good night again :)
00:55:19preglowl = l + (l - (l + r)/2)*width, r = r + (l - (l + r)/2)*width
00:55:19amiconnpreglow: The MAS does this in its dsp core. Just mixing left_in and right_in into left_out and right_out using 4 scale factors of -1...+1
00:55:21preglowsomething like that
00:55:42preglowwhere l and r are sample values for left and right
00:55:44 Join courtc [0] (
00:56:11preglowthat's quite basically a mid/side conversion with side amplification
00:56:18amiconnI derived the formula used to calculate the factors from the desired width by calculating m/s
00:57:08amiconnIt's in sound.c
00:58:17preglowyeah, see it
00:59:40Roriyou upped the latest commit with the pause bug fix?
01:00:21muesli-will go 2 bed..have fun girls...
01:02:23preglowdoes anyone have any idea how the eq should be like?
01:02:24preglowlike specs
01:02:45preglowseems like a five band fixed frequency thing is what's most popular
01:03:14 Quit DMJC-L ("Leaving")
01:03:54amiconnpreglow: I'd prefer simple treble/bass, plus a correctly working loudness
01:04:05Rorigimmie parametric any day of the week ;)
01:04:06preglowamiconn: uda has that already
01:04:17preglowamiconn: that is, no loudness
01:04:36preglowi've never done a loudness filter before
01:04:41preglowjust now roughly what it has to do
01:05:23amiconn(UDA) Yeah, no loudness, and no sensible treble/bass range
01:05:40preglowi guess i could make several eqs
01:05:58preglowi've got a feeling people will request a more elaborate one in addition to a bass/treble thing
01:06:25preglowloudness basically just lifts the bass and treble a bit, yeah?
01:06:47preglowto compensate for 4khz peak in hearing at low volume?
01:06:54[solid]imagine having headphones eq presetes for rockbox... extreme l33tness
01:07:02 Quit DomZ ("CGI:IRC (EOF)")
01:07:50amiconnNot that simple. It does it depending on the volume, and (hopefully) matching the ear's sensitivity curve
01:08:00preglowamiconn: do you have any idea how steep the eq filters in the mas are?
01:09:02amiconnAll mas have volume/balance/treble/bass. These are done in the audio codec part
01:09:56amiconnAll mas also have the stereo matrix, done in the dsp
01:10:01preglowi never use eqs myself, so i really don't know what's preferable
01:10:10amiconnThe newer ones (3587 and 3539) additionally have loudness plus 'superbass'
01:10:26amiconn..also done in the audio codec
01:10:51amiconn'superbass' merely shifts the center frequency of the loudness function from 1 to 2 kHz it does nothing with loudness == 0
01:11:33amiconnFinally, the newer mas' also have 'Micronas dynamic bass', done in the dsp
01:11:58amiconnThis is a virtual bass expansion based on the phenomenon of the 'missing fundamental'
01:12:03preglowahh, yes
01:12:30preglowread about that somewhere
01:13:32[solid]ladies and gentleman... proud to announce flac playback has just passed the 6h mark
01:13:47 Quit muesli- (Read error: 60 (Operation timed out))
01:13:59[solid]it's getting exciting
01:14:10[solid]6h was an optimistic estimate for me
01:14:21[solid]...but it keeps on playing
01:14:26preglownot bad at all
01:14:42linuxstb[solid]: Did you remember to remove the charger? :-)
01:14:46preglowamiconn: but i really don't see the point, the should be no missing fundamental with proper earphones
01:14:54 Join bagawk [0] (~Lee@bagawk.user)
01:15:16amiconnI don't know of any earphone that can really handle 20 Hz for that matter
01:15:24amiconnEven 50 Hz is problematic
01:15:30[solid]linuxstb: yeah.. it's so cool i think i'll actually give it a listen now
01:15:41amiconnProper headphones can do that, but certainly no eaphones
01:15:50preglowof course
01:15:56preglowi assume everyone uses great big headsets, like i do ;)
01:16:10amiconnYou said earphones...
01:16:21preglowearphones = headphones, to me
01:16:36preglowi call the smaller ones with shitty sound earbuds
01:17:22preglowbut yeah, that kind of processing will quickly be expensive for us
01:17:40preglowwe'll probably have to do pitch estimation and selective eqing or something
01:18:54amiconnI don't think we'd need that. I don't use it on the archos, because it adds a quite noticeable amount of noise
01:19:50preglowi'd never use it for sure
01:19:59amiconnIiuc what it does is: Filter out bass frequencies (cut off frequency is adjustable), calculate harmonics for them, then add these to the signal
01:20:32 Join midk [0] (
01:20:50preglowwell, yes, that's actually a quite nice way of doing it
01:21:11***Saving seen data "./dancer.seen"
01:22:03preglowperhaps they just distort the filtered out bass
01:22:11preglowthat should generate plenty of harmonics
01:23:19amiconnIt can't be that costly, because the mas manages to do this in addition to decoding mp3
01:23:40amiconn...with its internal frequency of ~24 / ~22 MHz
01:23:40 Join Tomhab [0] (
01:23:44preglowfiltering out the bass and doing some kind of waveshaper on it should be pretty cheap
01:24:59preglowa one pole filter on a dsp should negligable
01:25:03preglowas should a waveshaper
01:28:49preglowbut i need to sleep
01:29:03 Quit preglow ("s")
01:30:43 Quit bagawk ("Leaving")
01:31:13TomhabIs the gmini series dead in the water at the minute?
01:33:28midkyou might check out gmemu.
01:37:21 Quit Sucka ("a bird in the bush is worth two in your house")
01:41:29 Join DMJC [0] (
01:51:14 Quit linuxstb (
01:51:14 Quit Zoom2 (
01:51:14 Quit thegeek (
01:51:14 Quit einhirn (
01:51:14 Quit Bagder (
01:51:14 Quit pabs (
01:51:14 Quit t0mas (
01:51:14 Quit Rick (
01:51:14 Quit Naked (
01:51:14 Quit spiralout (
01:51:14 Quit Seed (
01:51:14 Quit Tomhab (
01:51:15 Quit Rori (
01:51:15 Quit BTKDaImMaikata (
01:51:15 Quit tedboer (
01:51:15 Quit Zagor (
01:51:15 Quit DMJC (
01:51:15 Quit courtc (
01:51:15 Quit bipak (
01:51:15 Quit ehntoo (
01:51:15 Quit Stryke` (
01:51:15 Quit ripnetuk (
01:51:15 Quit Slasheri (
01:51:15 Quit rob- (
01:51:15 Quit silencer_ (
01:51:15 Quit mbr (
01:51:15 Quit odd (
01:51:15 Quit gromit` (
01:51:15 Quit Strath (
01:51:20 Quit lostlogic (
01:51:20 Quit tvelocity (
01:51:20 Quit cYmen (
01:51:20 Quit QT (
01:51:20 Quit crashd (
01:51:20 Quit bill2or3 (
01:51:20 Quit [zmaj] (
01:51:20 Quit ze (
01:51:20 Quit edx (
01:51:20 Quit DangerousDan (
01:51:20 Quit [solid] (
01:51:20 Quit Lynx_awy (
01:51:20 Quit crash_ (
01:51:20 Quit C-Keen (Success)
01:52:50 Join C-Keen [0] (
02:00:40 Quit midk (Read error: 104 (Connection reset by peer))
02:01:25 Quit dwihno (Nick collision)
02:01:25 Quit amiconn (Nick collision)
02:01:25 Join dwihno [0] (~dw@
02:01:25 Join amiconn [0] (
02:01:25NJoinDMJC [0] (
02:01:25NJoinTomhab [0] (
02:01:25NJoincourtc [0] (
02:01:25NJoinNaked [0] (
02:01:25NJoinbipak [0] (
02:01:25NJoinlinuxstb [0] (
02:01:25NJoinZoom2 [0] (
02:01:25NJointhegeek [0] (
02:01:25NJoinehntoo [0] (
02:01:25NJoinspiralout [0] (
02:01:25NJoinRori [0] (
02:01:25NJoinBTKDaImMaikata [0] (~Bager@
02:01:25NJoinStryke` [0] (
02:01:25NJoineinhirn [0] (
02:01:25NJoinedx [0] (
02:01:25NJointvelocity [0] (
02:01:25NJoincYmen [0] (
02:01:25NJoinBagder [0] (
02:01:25NJoinStrath [0] (
02:01:25NJoinQT [0] (as@area51.users.madwifi)
02:01:25NJoinDangerousDan [0] (
02:01:25NJoinSeed [0] (
02:01:25NJoinlostlogic [0] (
02:01:25NJointedboer [0] (
02:01:25NJoincrashd [0] (
02:01:25NJoinpabs [0] (
02:01:25NJoinZagor [0] (
02:01:25NJoinbill2or3 [0] (
02:01:25NJoinripnetuk [0] (
02:01:25NJoint0mas [0] (
02:01:25NJoin[solid] [0] (~solid@
02:01:25NJoinLynx_awy [0] (Lynx@
02:01:25NJoinRick [0] (rick@Rick.user)
02:01:25NJoinSlasheri [0] (
02:01:25NJoincrash_ [0] (
02:01:25NJoinrob- [0] (
02:01:25NJoinsilencer_ [0] (
02:01:25NJoinmbr [0] (
02:01:25NJoinodd [0] (
02:01:25NJoingromit` [0] (~gromit`
02:01:25NJoin[zmaj] [0] (
02:01:26NJoinze [0] (
02:01:29CtcpVersion from freenode-connect!freenode@connect.utility.freenode
02:01:29***Server message 477: 'logbot #rockbox :[freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify:'
02:03:45 Quit Tomhab (Read error: 60 (Operation timed out))
02:04:54linuxstbI'm still not sure if it's a good idea, but I've implemented a FLAC parser to read the metadata independently of the codec. It currently lives in playback.c and populates the "id3" struct with things like "length", "samplerate", "bitrate" etc.
02:07:53[solid]okay, everybody's sleepin' and my iriver is still playing o_O
02:08:03Stryke`the flac test?
02:08:31[solid]7hrs mark in 3 minutes
02:08:49Stryke`hows the battery doing?
02:09:10[solid]the indicator says it's at 38-41 now
02:09:18Stryke`and is it the same flacs over and over, maybe not reading the HDD?
02:09:47[solid]however the voltage is now dropping more rapidly
02:09:53 Join forester [0] (
02:09:56[solid]3.66V atm
02:10:08ehntooit's still going, isn't it?
02:10:15ehntoo7 hours...
02:10:33[solid]it's the same flacs, but it has 32 megs of ram and not 337megs :D
02:10:37foresterAnyone happent o know which format has a better designed and documented decoding api? (Wavpack or FLAC)
02:10:59ehntooI'm assuming it depends on what you want, exactly
02:11:14[solid]forester: flac will probably be better documented and worse designed
02:11:20[solid]forester: with wavpack the opposite
02:11:28forester[solid]: Probably?
02:11:39[solid]yup, i'm no pro
02:11:42foresterI notice it is better documented, much better.
02:11:55[solid]flac is older and much more popular
02:11:55foresterI am not sure it is worse designed simply because it is older though.
02:11:57ehntoomeh... what in the world is with some of these ogg's...
02:12:05ehntoothey worked with yesterday's bleeding edge I was using
02:12:11Stryke`[solid]: what volume are you playing at?
02:12:18[solid]Stryke`: 50%
02:12:44foresterAnyone compare flac and wavpack seeking?
02:13:01[solid]the battery is now draining faster and faster, wonder if it'll reach anytime near 8hrs
02:13:17ehntoolet's hope it does.
02:13:21[solid]forester: i've heard flac seeking == hell
02:13:39foresterI just read 'extremely fast seeking'. :p
02:14:03[solid]forester: hell in terms of coding, not speed:)
02:14:43foresterAny recommendations on where I could get a quick answer to this dilemma?
02:15:01foresterI want to use whatever format I will be using the decoder of later. :p
02:15:08foresterThe decoder api that is.
02:15:26foresterHoping for something better.
02:15:32[solid]but they don't have any comparisons on the developing side
02:17:26[solid]hm... wonder if such a playback test could be used for improving the battery indicator
02:17:59[solid]saving battery states to some buffer and writing them to a file when the disk is spun for audio buffering
02:18:14linuxstbforester: Why are you asking about decoding APIs?
02:18:31ehntoolooks like I'm going digging into the vorbis code...
02:18:50foresterWriting an on-demand on-the-fly-re-encoding server.
02:19:15[solid]forester: you might take a look at squeezebox
02:19:24[solid]forester: it's gpl afair
02:20:17foresterWhat url? Google results confusing me.
02:21:15foresterMore flac supporting hardware doesn't bode well for wavpack.
02:21:21linuxstbAll that I can say is that I've used libFLAC in various applications, and had no complaints with it.
02:21:36foresterlinuxstb: I am optimistic about it.
02:21:56[solid]forester: sure, but wavpack has a huge chance with rockbox
02:22:06[solid]forester: and the other way round too, imho
02:22:25foresterflac has no chance?
02:22:30[solid]not that
02:22:46[solid]i think it might get more popular when it has decent dap support
02:23:15[solid]rockbox will be the first firmware supporting it i think
02:23:24[solid]and it already is realtime
02:24:08zesupporting what, flac? my karma plays flac
02:24:35zebarely ever heard of it :p
02:25:07[solid]wavpack and musepack might bring rockbox to a new level of popularity i think... two formats with very enthusiatic, devoted and usually also knowing-a-tad-more fans that don't have dap support yet
02:25:21Stryke`musepack :-)
02:25:52foresterI think musepack is about done.
02:26:02[solid]you know, the kind of final screw-the-ipod argument for the totally l33t g33ks
02:26:14Stryke`done, how?
02:26:52foresterDeveloper stopped doing much with it many years ago, before it was renamed... promises never fulfilled...
02:27:03foresterYears of enthusiasm is fading away.
02:27:10foresterWith some late-comer echoes.
02:27:11Stryke`oh, yes, but it still can't be beat for quality vs. size
02:27:36foresterStryke`: Many mpc'ers went lossless.
02:27:38[solid]musepack development is resumed
02:27:42foresterThe early ones anyway.
02:27:47forester[solid]: By who?
02:27:53forester[solid]: Frank?
02:28:19forestermp+/mpc are two different beats.
02:28:25forester[solid]: Yeah, I visited that url years ago.
02:28:37Stryke`if my H120 could play MPC, regardless of the future of the format, i'd certainly encode my new CDs to MPC
02:28:42[solid]note there was an update t wo days ago..
02:28:47 Quit linuxstb ("Leaving")
02:29:03foresterBig deal.
02:29:06foresterIt's not the same thing.
02:29:37[solid]but i'm not sure what are they going to do about sv8
02:29:45foresterThere are also licensing concerns actually.
02:29:52foresterThat's why I think weird things seemed to happen.
02:29:55[solid]i don't think so
02:30:02[solid]the one patent that was the problem expired
02:30:23foresterI haven't heard from Andre for years.
02:30:29foresterThen all these leet sites show up.
02:30:30[solid]about the time people picked up development i think... but i'm not sure
02:30:33foresterFrank starts making releases.
02:30:40foresterPromises made my Andre are not fulfilled.
02:30:42[solid]it's all lgpl now..
02:30:45foresterHe is not working on it.
02:30:48foresterAs far as I know.
02:30:57foresterAnd this is a story spanning many years.,
02:31:08foresterYeah, now it's lgpl.
02:31:08[solid]Musepack was created by Andree Buschmann (mirror) (archival site) and is currently developed by Frank Klemm (mirror) and the Musepack Development Team.
02:31:11foresterBecause of all this drama.
02:31:54foresterI think mpc is past its peak.
02:32:22[solid]yeah, probably because of the stop in development back then
02:32:22foresterIt may get better technically of course.
02:32:25Stryke`the point remains, its a very fast encoding/decoding format with high quality at low(er) bitrates
02:32:28 Quit Rori (Read error: 54 (Connection reset by peer))
02:32:33Stryke`sounds pretty good for a portable
02:32:33foresterStryke`: duh.
02:32:34foresterStryke`: duh.
02:32:34DBUGEnqueued KICK forester
02:32:34foresterStryke`: duh.
02:32:35 Join Rori [0] (
02:32:44foresterStryke`: I have many mpc and used it for years.
02:32:47[solid]Stryke`: i always thought it was good at high(er) bitrates :)
02:32:53foresterStop acting like I said otherwise.
02:32:58Stryke`160 - 180 kbps is not exactly high(er)
02:33:05[solid]well, it is
02:33:13[solid]vorbis does sound very good at 128
02:33:46[solid]and that's still not the ultra-lo category
02:34:04 Quit Rori (Read error: 54 (Connection reset by peer))
02:34:08ehntoovorbis sounds... listenable when you get much lower than 128
02:34:11[solid]nowadays you can encode a file at 128 and not be able to abx...
02:34:15 Join Rori [0] (
02:34:28[solid]below ~100 it's becoming a matter of which codec sounds more listenable
02:34:50[solid]because artifacts become quite obvious
02:35:10[solid](and my h120 is still playing)
02:35:45[solid]scary :)
02:36:39 Quit Rori (Read error: 54 (Connection reset by peer))
02:36:44[solid]but musepack could become tha shiznit again imo
02:36:49 Join Rori [0] (
02:37:16[solid]if only frank would create at least a reference sv8 implementation
02:37:26 Quit Rori (Read error: 54 (Connection reset by peer))
02:37:32foresterI said the same thing for years about sv7.
02:37:37 Join Rori [0] (
02:37:51[solid]ooh. you used musepack before sv7?
02:37:55[solid]that's l33t man
02:37:59forestersv7 is new.
02:38:13foresterI waited so long for sv7, for streaming support.
02:38:29[solid]i've yet to see a non-sv7 .mpc file:)
02:38:36foresterWant one?
02:38:40foresterOr hundreds?
02:38:53[solid]oh, so he promised streaming support for sv7 already? XD
02:38:55foresterI don't know, I seem to recall sv3,4 and 5...
02:39:07forester[solid]: It's not there?
02:39:11forester[solid]: I wouldn't be surprised.
02:39:24[solid]nope... it's supposed to be... in sv8 :D
02:39:28forester[solid]: Years ago sv7 was supposed to ...
02:39:34foresterI would still be here waiting.
02:39:41foresterAnd when I talked to him, it was a week or two away.
02:39:43foresterNot 3 years.
02:39:49foresterOver and over.
02:39:50[solid]that sheds a new light on musepack, thanks ^^
02:39:52foresterAnd the same thing now.
02:40:17ehntoowell, when you do something for free...
02:41:22[solid]ehntoo: the only thing that makes my computer slow is a component that for sure wasn't made for free (ide controller)
02:41:42 Join ashridah [0] (
02:41:54[solid]the stuff that was made for free makes it run miraculously fast comparing to windows :)
02:43:30ehntoogotta love Gentoo.
02:44:10[solid]ehntoo: exactly ^^
02:45:08[solid]hmm... iirc the iriver fw has 8-10hrs uptime with ogg vorbis -q6 files
02:47:25ehntoofrom what I've seen, that's right
02:48:08 Quit ripnetuk (Read error: 104 (Connection reset by peer))
02:48:35[solid]looks like a larger battery replacement won't be needed at all...
02:51:07 Quit ]RowaN[ ()
02:55:55ehntoo*updates gcc to hopefully include m68k support
02:57:02[solid]i wonder how much more battery time can be squeezed with further optimisations of the core, the codec and file buffering
02:57:50[solid]i'm also wondering if i haven't accidentaly used ogg files lol
02:58:23Stryke`lol, that would be funny
02:59:14[solid]oh i can check it with rockbox (rename)
02:59:37[solid].flac alright
03:10:09[solid]good god.
03:10:31[solid]i've come to the point where i'm not sure if i believe my own test...
03:11:13 Part forester
03:21:15***Saving seen data "./dancer.seen"
03:25:55spiralouthow much time is elapsed since staring the test? you´re testing flac? do you know how long the hp is mp3 running? and what time is it ? :-P
03:26:31[solid]oh man
03:26:33[solid]it's dawn here
03:27:25[solid]the time is 03:27 so it's about 8hrs 17minutes now since 19:10 yesterday
03:27:49[solid]and yes it's flac, at max compression rate
03:28:06spiraloutomg nice
03:28:19[solid]ogg vorbis and mp3 tests will follow
03:28:36[solid]yeah, i almost don't believe it..
03:28:47spiraloutdo you hear the music or is it just playing
03:28:51Roriflac over 8 hours already that is amazing
03:28:59[solid]i hear it
03:29:04[solid]it's 50% volume
03:29:13[solid]in my koss ksc75's
03:29:42RoriI'll have to convert my ape to flac now :)
03:29:42[solid]and i must admit i got already bored of this album even tho it's lovely
03:30:37spiraloutwhich album...??:-D
03:30:47[solid]but death is finally coming
03:30:57[solid]i sense it;P
03:31:03 Join amiconn_ [0] (
03:31:11[solid]spiralout: antipop consortium - tragic epilogue
03:31:24[solid]amiconn_: 8hrs and counting
03:31:30[solid]amiconn_: how about that?:D
03:33:20[solid]battery indicator is at 6% and blinking
03:33:26[solid]and the birds started singing
03:34:50spiraloutok time to spleep for me now...hmm i think you have to hear until the batterie is dead but that is far away :-P lol ok not really but i hope so...n8
03:35:14[solid]it's almost dead
03:35:23[solid]the disk can't spin up properly
03:35:29[solid]oooh. silence.
03:35:36spiraloutwhat :-OOO
03:35:43ehntoois it still on?
03:35:56[solid]yes, but it's not playing anymore
03:36:11[solid]too little power for the drive to spin
03:36:13 Join textchimp [0] (~text@
03:36:17 Join cheriff [0] (davem666@wagner.orchestra.cse.unsw.EDU.AU)
03:36:20[solid]salvation, i can go to sleep..
03:36:30Roriyou must now do ogg and mp3 ;)
03:36:39spiraloutyou areno your ears are sleeping but your eyes are awake
03:36:47textchimpcan anyone suggest a program which will allow everyone in (for example) an office to add mp3 files to a single central queue to be played from a single machine?
03:36:54[solid]and how am i supposed to do that with a completely discharged unit?:PPPP
03:37:09Roriah true true
03:37:23[solid]haha lol the buttons are not working
03:37:29ehntootextchimp, you mean like requests on a radio station?
03:37:29[solid]i can't shut it down
03:37:41Roricharge it
03:37:47ehntooanother 4 hours
03:38:12[solid]where's that *(%$#&^# pin
03:38:14 Quit amiconn (Read error: 60 (Operation timed out))
03:38:14 Nick amiconn_ is now known as amiconn (
03:38:22Rorinot advisable to let it die completely anyhow
03:38:25textchimpehntoo: well, maybe....just one queue that everyone adds to or modifies, and gets played through a single machine
03:38:28[solid]yeah i know
03:38:33[solid]not with li-ions
03:38:47Roriit's probably why it just stops
03:38:49ehntootextchimp: wouldn't that be kinda chaotic?
03:39:00Rorirather than shuts down when exhausted
03:39:11 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
03:39:30textchimpehntoo: no....i can't believe this isn't something everyone wants to do
03:41:08ehntootextchimp: I can only think of managed systems...
03:41:26ehntootextchimp: what level of users are we talking?
03:42:09ehntoocommon accounting zombies or talented manic programmers?
03:45:37textchimpehntoo: programmers
03:45:50textchimpehntoo: i actually thought there was a way for itunes to do it
03:45:54[solid]iriver fw seems to keep the platter spinning all time time
03:46:55[solid]oh it stops sometimes
03:47:03[solid]it's just too late :)
03:47:12[solid]or too early rather
03:48:24[solid]i'll be doing a vorbis test later on today... however this was such an impressive and surprising result i'm thinking of doing a flac test again
03:48:34[solid]well, anyhow... good night:)
03:52:25 Join d00bie [0] (
03:53:20d00biewutza rockbox?
04:00:34 Part d00bie
04:04:07ehntootextchimp: I'm trying to think
04:04:27ehntooso the central machine is going to be actually playing the files, or just serving them
04:05:31 Join DJ-suparappa [0] (
04:05:48 Join QT_ [0] (as@area51.users.madwifi)
04:10:33Rorirock my box
04:13:54textchimpehntoo: playing them
04:17:26 Quit QT (Read error: 110 (Connection timed out))
04:38:40 Quit ashridah ("Leaving")
04:46:57ehntootextchimp... you could use ssh and mpd
04:47:01ehntoodunno how well that'd work
04:49:07textchimpehntoo: wwwinamp looks like what i want....though i'd prefer a way to get the song db to refresh when required
04:49:17 Join ripnetuk [0] (
05:16:17 Quit cYmen ("zZz")
05:21:19***Saving seen data "./dancer.seen"
05:21:24 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
05:21:50 Join dapureplaya [0] (
05:22:25dapureplayadoes any1 here have playback problems for the latest build?
05:26:06DJ-suparappawhen u create a wps file should it appear in your player browser?
05:26:16DJ-suparappacuz i put it in any folder and it wont show up
05:27:12dapureplayau putting it in the .rockbox folder?
05:27:37DJ-suparappai tried that but after i put it in some other folders and it wont show up
05:27:45DJ-suparappasame folders w/ my music
05:28:06dapureplayau sure u got it in a .wps file?
05:28:14dapureplayaor is it a ***.wps.txt file?
05:28:24DJ-suparappano i renamed it .wps
05:28:35DJ-suparappathe problem is when i goto play it its not there tho
05:28:37dapureplayawell it works fine fo rme
05:28:50dapureplayai just put it in the .rockbox dir.
05:28:58dapureplayaatm playback is acting up on the latest build
05:29:09DJ-suparappahow do i get it to work from there tho?
05:29:20dapureplayafor some reason when u play a song, stop and playback a different song it just doesn't play
05:29:26DJ-suparappathat foler doesnt show up in the browser
05:29:28dapureplayaall i did was saved it into .rockbox
05:29:37DJ-suparappao and it updated itself?
05:29:42dapureplayathen i went into the options and selected 'browse .wps files'
05:29:47dapureplayaand clicked on it.
05:29:49dapureplayaand it showed up
05:30:44dapureplayai'm changin back to the old build. this build is kinda acting up too much.
05:30:44DJ-suparappawheres the brows .wps files?
05:31:04dapureplayageneral settings >> display >> browse .wps files
05:31:10dapureplayamake sure u put it in the .rockbox folder tho
05:31:14dapureplayait only navigatees around there
05:31:22DJ-suparappaah ok
05:31:35 Quit tvelocity (Read error: 110 (Connection timed out))
05:31:36dapureplayait should work.
05:31:50DJ-suparappaah that worked
05:31:52DJ-suparappathanks man
05:32:12dapureplayadid u just try to use browser to open it?
05:32:18DJ-suparappayeah lol
05:32:24dapureplayayeah that wouldn't work
05:32:27DJ-suparappaim a rockbox noob
05:32:32 Join tvelocity [0] (
05:32:34dapureplayaheh don't worry so am i
05:32:49dapureplayaevery1 is except the devs heh
05:32:59dapureplayai learn new things every time i fiddle with this thing.
05:33:01DJ-suparappai went there the other day and saw "plays mp3 and ogg" and i had to get it
05:33:16DJ-suparappayeah thats what i plan to do
05:33:24DJ-suparappaim lovin this real gapless playback
05:33:33DJ-suparappaunlike the official firmware
05:34:14dapureplayahmmm....this recent build is acting up.
05:34:17dapureplayai'm going to put back the old one
05:34:24DJ-suparappawhats wrong?
05:34:25dapureplayait stops playing songs after a while.
05:34:40dapureplayathey implemented 'now playing' screens for FLAC and OGG now
05:34:44DJ-suparappahmm ill test that after im dont makin my wps
05:34:59dapureplayait plays when u first turn it on
05:35:04dapureplayabut once u stop and change to another song
05:35:13dapureplayau can't hear anything
05:36:13dapureplayawhen u do d/l it make sure u have a copy of ur previous build. Coz playback on it atm is kinda acting up. although it's cool it was wps for FLAC and ogg. also they implemented some buttons on the LCD remote
05:36:35dapureplayaBUT me = not a remote user
05:37:06DJ-suparappaoo they did
05:37:11DJ-suparappacuz i AM a remote user
05:37:24dapureplayai think they've only got stop and play working
05:37:47dapureplayaand the hold button aswell.
05:37:49dapureplayabut that's it
05:38:01 Quit textchimp ()
05:38:01dapureplayathey're still making real rapid progress atm.
05:38:44dapureplayawhen u have the time can u check the latest build and get back to me? i wanna know if i'm the only one having this playback problem
05:39:49 Join Harpy [0] (
05:42:06dapureplayaaaaww man i'm having a serious craving for a gourmet pizza. Screw this i'm ordering myself one for pick-up now. Hehe
05:45:24DJ-suparappa%pf : Player: Full-line progress bar + time display
05:45:31DJ-suparappathat didnt seem to work =\
05:45:47dapureplayahmmm.....well i haven'
05:46:01dapureplayahavne't had a go at making a .wps file yet so.
05:46:23dapureplayacan't help ya there :p. I'm still so fascinated with the FLAC Playback. :p
05:46:38DJ-suparappawhats FLAC?
05:46:47dapureplayalossless audio codec.
05:46:57DJ-suparappanever heard of it
05:47:02dapureplayait's cd quality
05:47:15DJ-suparappaso not compressed at all?
05:47:27dapureplayakeep in mind it's just a wav file which is compressed without losing any audio quality
05:47:35DJ-suparappathats cool
05:47:38dapureplayathink of it as a .zip file
05:47:40dapureplayabut audio
05:47:52DJ-suparappai should check that out later
05:47:58dapureplayayeah it's pretty kewl.
05:48:18dapureplayaonly one bad thing about it is that the amount of space it takes up hehe
05:48:36dapureplayai ripped 3 albums. each are about 1/2 gig
05:48:44dapureplayaso 1.5 gigs gooooooooone alreadyhj
05:49:32dapureplayathere's FLAC playback on the Rio Karma. and i think the iAudio M3/X5. So it's about time iRiver can play em hehe
05:50:23dapureplayasome peopel said that FLAC isn't gapless, but the thing is that it sounds pretty gapless to me.
05:50:30dapureplayabrb goign to order pizza
05:51:03 Nick cheriff is now known as cheriff_AWAY (davem666@wagner.orchestra.cse.unsw.EDU.AU)
05:54:29dapureplayadamn. 20 min wait. usually doesn't take that long lol
05:54:47DJ-suparappa:) my wps is commin along nicely
06:05:56DJ-suparappawell theres a couple problems
06:06:03DJ-suparappacant go back a song
06:06:20DJ-suparappaand the comming up doesnt ever change
06:07:04DJ-suparappahmm the comming up is working now ><
06:08:20DJ-suparappaok i see the problem ur having now
06:08:28dapureplayaok cool so it's just not me
06:08:48DJ-suparappasometimes when the song changes it stops and goes to the browser
06:09:06dapureplayatry turning off and on the player
06:09:10dapureplayathen try playing an ogg file
06:09:20DJ-suparappai dont think i have any ogg
06:09:21dapureplayathen press stop and try another file.
06:09:24DJ-suparappalemme check a few games
06:09:40dapureplayagames should be fine it's the playback i'm having troubles with for the new build. i just put back on my old one
06:09:54DJ-suparappano i mean check some games for ogg
06:10:04 Quit Zoom2 ("CGI:IRC (Ping timeout)")
06:10:21dapureplayai'm goign to get this pizza already hehe. bbl
06:10:24DJ-suparappamaybe halflife 2 uses it
06:10:27 Nick dapureplaya is now known as dapureplaya_AWAY (
06:10:33dapureplaya_AWAYDOOM3 uses it
06:10:43DJ-suparappalol damn i dont have that anymorew
06:11:05DJ-suparappao well ill try restarting and just keep w/ mp3
06:11:06dapureplaya_AWAYwell your going to check out FLAC, u may as well rip something into FLAC and OGG if youw ant.
06:11:07DJ-suparappaand see if it happens
06:11:07dapureplaya_AWAYanyway gtg now
06:12:15 Join midk [0] (
06:14:51 Quit midk (Client Quit)
06:15:07 Join midk [0] (
06:19:36 Join Aramil [0] (~tony@
06:21:30DJ-suparappacan anyone tell me how i update the build?
06:21:47midkyou mean when compiling?
06:22:11DJ-suparappawell dont i have to do something like "update firmware" everytime i change it?
06:22:24midkyou mean when you put a new ajbrec on the unit?
06:22:35midkor archos.mod?
06:22:40DJ-suparappafor iriver
06:22:47midkoh.. unfamiliar with that.
06:23:23DJ-suparappacuz the latest build has playback broblems and im not sure how to install the 1 bfore it =\
06:25:56 Quit tvelocity (Read error: 60 (Operation timed out))
06:35:28 Join StrathAFK [0] (
06:44:03dapureplaya_AWAYu don't need to reflash bootloader every time
06:44:15dapureplaya_AWAYall u do is get the "bleeding edge" build and copy it over to the H140
06:44:29dapureplaya_AWAYanyway i'm off. time to eat! Cya
06:44:31 Part dapureplaya_AWAY
06:46:33 Join tvelocity [0] (
06:51:43 Quit Aramil (Read error: 60 (Operation timed out))
06:52:05midkanyone familiar with the cross-compiler? need some help
06:52:09midkfor linux, that is.
06:53:49 Quit Strath (Read error: 110 (Connection timed out))
07:09:06 Quit pabs (Read error: 60 (Operation timed out))
07:10:57 Join pabs [0] (
07:18:15 Join LinusN [0] (
07:21:22***Saving seen data "./dancer.seen"
07:32:33BTKDaImMaikatamorning, guys
07:32:34midkhey linus :)
07:32:42 Nick BTKDaImMaikata is now known as Bger (~Bager@
07:33:08midki was about to message you half an hour ago about an error building the cross compiler.. when i noticed the workaround for SH1 just under gcc build instructions
07:33:30midkthat really should be moved above the instructions so you actually notice it before you rip your hair out :)
07:35:33LinusNmidk: so you're bald now, and it's my fault...?
07:35:50midkyes, mister can't-design-a-webpage. ;)
07:36:41midkhmm. i thought convbdf was automatically built.
07:38:23midki built them manually and it worked.
07:38:27midkWHO'S GOOD?!
07:39:58LinusNmidk: you da man, mister can't-edit-it-yourself-in-the-wiki :-)
07:40:41midkbah you. i'm busy. ;)
07:41:00midkwith very important business.
07:41:04midk*opens rockbox.c*
07:41:12midker. blox.
07:42:16LinusNmidk: are you working on the iriver rockblox patch?
07:42:16midkagh. not the konqueror-refuses-to-open-after-plugging-usb-device problem again.
07:42:34midkno.. i don't have an iriver :)
07:43:21midki think i've got to get one if i want to continue rockbox work though. seems kind of pointless to work on an older platform. :)
07:45:09bill2or3start on a H340 port. :-)
07:46:18midkhaha. hm. that's a great idea. i'll go buy an h340 tomorrow... and randomly press keys indefinitely until i happen to write all the necessary code. :)
07:46:24midkif monkeys can do it.. i sure as hell can.
07:46:49bill2or3you just need 999 friends
07:47:20midkoh, whoops.
07:48:00midkwell. no... since i have an infinite amount of milleniums to do it correctly :)
07:48:04 Join Zoom2 [0] (
07:48:18midkof course.. i should consider the potential life of the actual project.
07:49:11bill2or3I'mshopping for my first player.. about this >< close to ordering an H340
07:49:44midki guess the distance between you and the buy button then depends on your screen resolution and monitor size.
07:50:14bill2or3but in any case, it's pretty close
07:50:19midklooks like about 1 pixel to me... and thus... you should have clicked it about .1 seconds after you alerted me about your readiness to purchase it.
07:50:50midkthis av320 probably won't last much longer.. and i've got that best buy warranty to exchange it for anything i want.
07:51:17midksince it was around $500.. i wonder if i can get store credit for that, to get two different things..
07:51:30midki want that AV700.
07:54:01midkhmm. it's just too bad they rarely stock much in the way of MP3 players and multimedia devices.
07:54:24bill2or3where do you live?
07:54:26midkthey used to have the iRivers.. i don't think they do anymore. i know they've always got ipods.... probably the creative portable media center thing....
07:54:29midkseattle, wa.
07:54:39bill2or3I'm in utah, never managed to see an H340 in person
07:54:46midkhmm. me either.
07:56:28 Join pabs_ [0] (
07:56:55midkas expected. they've got every single color of the ipod mini, every version of the ipod including photo, every color of the zen micro, the touch, and the portable media center...
07:57:16 Quit pabs (Nick collision from services.)
07:57:21 Nick pabs_ is now known as pabs (
07:57:21midkand then a couple random things like the iriver 6gb player, a rio player, and a couple samsung ones.
07:57:28bill2or3and..... nothing
07:57:46bill2or3I saw an H10, dig the slider
07:57:57midkyeah, those H10s.
07:58:03midknever seen one nor tried it.
07:58:47midki have an idea... i could get the most expensive one there, like the portable media center which is $500... and ebay it brand new, and use the cash to get an iriver ;)
07:59:03midkthough.. the portable media center isn't all that unappealing in the first place.
08:00:18midkmaybe they can back-order something else, though... it's kind of unfair that they can limit me to ipods or zen micros..
08:01:05midkspeaking of which, i wonder when this warranty/service plan/replacement deal lasts.
08:01:26midkit'd be just like them to notice how often i've used it and not notify me that it's about to end and thus get me stuck with a broken player :P
08:15:55HClcats are cute, but its a bit annoying that they wake you up at 7 :/
08:17:01zeHCl: only if you're the one who feeds 'em
08:17:08midkze: q3.
08:17:14zemidk: eh?
08:17:20midkquake 3.
08:17:23zewhat about it?
08:17:28midklet's play.
08:17:34zei don't even have it installed
08:17:41zei don't think i even have the nvidia drivers set up for opengl
08:17:42midki was kidding. :P
08:22:06midkwait.. RLD is gone?
08:22:40dwihnoRLD is just something you made up, right? ;)
08:23:20 Quit Zoom2 ("CGI:IRC (Ping timeout)")
08:24:54HClze: he mostly wanted to cuddle and play
08:25:58midkreboot, brb.
08:26:19 Quit midk (Remote closed the connection)
08:28:17dwihnoI'm rather pleased. Seems like I got the VPN stuff rolling quite OK after all!
08:28:56 Join midk [0] (
08:31:09 Join Bipak2 [0] (
08:31:24 Join dapureplaya [0] (
08:32:44dapureplayaIs anyone here having trouble with the latest build for iRiver?
08:33:13dapureplayait doesn't wanna play audio files very much.
08:34:03dapureplayalike i turn it on. play one song......then i press stop and try to play another song in a different format
08:34:11dapureplayaand no sound comes out
08:34:39dapureplayair. i'll play an MP3 file then an OGG file and then ogg won't play.
08:34:54dwihnoHow well is the playback on iriver nowadays btw?
08:35:19Bagderit plays rather good
08:35:28dapureplayawell the previous build i had played everything fine.
08:35:34t0masmorning :)
08:35:54dapureplayaexcept my prev. build doesn't have 'now playing' screen for flac or ogg.
08:35:57LinusNdapureplaya: it doesn't matter which format you play
08:36:10dwihnoBagder: it must've happened lately then... or am I wrong?
08:36:10LinusNit can't play after stopping
08:36:18Bagderdwihno: the last week
08:36:29t0masdapureplaya, LinusN: it can with the newest build
08:36:41dwihnoBagder: really cool! so it's about time to get a h140 then I guess ;)
08:36:41SlasheriHmm, maybe the new wps update has broken this up
08:36:44LinusNi just installed latest cvs
08:36:51SlasheriI have really interesting problems too
08:36:54t0masit works here
08:36:58SlasheriSomething i have already solved
08:37:07t0masSlasheri: check out the pause fix then
08:37:21t0masI'm not sure if that's the perfect place to do it
08:37:30Slasherit0mas: I have done that and in my code there are other updates also that makes pause to work with next
08:37:39dapureplayawell i tried playing one song just then, pressing the stop button and playing a different song and it won't play anymore
08:37:45SlasheriBut it wont play everytime
08:38:14Slasheridapureplaya: Yep, that was one problem i just fixed (not yet committed)
08:38:38t0masSlasheri: what are your other new problems?
08:38:47Slasherit0mas: just a moment
08:38:50dapureplayaah cool. i was just checking if i was the only one who had the problem. I was getting concerned
08:39:45LinusNSlasheri: there seems to ba a problem with the buffering, sfter a while i can hear garbage (or rather short chunks of previously played songs) being played in the middle of a song
08:39:57BagderI got that too
08:39:59Slasherit0mas: when i press next track, does it pause the current track and the forget to unpause?
08:40:00LinusNmaybe it is a problem with the wrapping
08:40:01dapureplayahere again
08:40:13Bagderand I've experienced the all-of-a-sudden-stop
08:40:18SlasheriLinusN: Hmm
08:40:20t0masSlasheri: hm? it's playing, you press next, and then it pauses??
08:40:33Slasherit0mas: I think so, i will verify that
08:41:01t0masonly change was:
08:41:01t0mas- if (!pcm_is_playing()) {
08:41:01t0mas+ if (!pcm_is_playing() && !pcm_paused) {
08:41:03CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
08:41:03*HCl mumbles stuff and guesses he'll go buy breakfast...
08:41:35dwihnoHas anyone noted the lock release on playlist end (with no icon change)?
08:41:59t0masno, thanks dwihno... we'll take a look
08:42:38t0mas(dwihno: newest version?)
08:42:58dwihnot0mas: No, but not that old.
08:43:45Slasherit0mas: Hmm, that was not the problem :)
08:43:56t0masSlasheri: found it?
08:44:03 Quit Bipak2 ("CGI:IRC")
08:44:06t0masdwihno: what button was working? remote play?
08:44:10SlasheriStill searching.. I will commit soon when i find it
08:44:12t0masor everythinh?
08:44:30HCl"He almost danced to the fridge, found the three least hairy things in it, put them on a plate and watched them intently for two minutes. Since they made no attempt to move within that time he called them breakfast and ate them"
08:44:40dwihnot0mas: I locked the unit, played a playlist and upon the end, the unit became unlocked and the "lock" icon remained
08:46:26t0masHCl: you live that far from mcDonalds? :P
08:47:32HClmcdonalds isn't much better ;p
08:48:37Bagdernow why have the spam rate increased so much lately?
08:48:56t0masposted your email address on some website?
08:48:57BagderI get 4 or 5 times my regular amount the last 4 days
08:49:14Bagdermy emails are everywhere already
08:49:18Bagdersince many years back
08:49:29*t0mas goes to take his non moving friendly looking breakfast ;-)
08:50:38dapureplayadammit my friggin digital-camera hasn't arrived yet. I better get it from those chumps before my trip otherwise there'll be hell 2 pay! >:-(
08:52:56gromit`do you see Monster Minds in your food ?
08:53:11amiconnGood morning
08:53:31 Join webguest42 [0] (
08:53:42*dwihno remembers the late 80's and watching wheeled warriors
08:53:44amiconnLinusN: Did you try ho pluggiing the remote? :)
08:53:52LinusNamiconn: not yet
08:54:09 Quit ehntoo (Ping timeout: 14400 seconds)
08:54:40Slasheriamiconn: It seems to work :)
08:57:12LinusNSlasheri: susprised? :-)
08:57:39SlasheriLinusN: it's nice to see the hotplug working ;)
08:57:45Bagderand wiki spam again
08:58:56t0masamiconn: lcd hot-plugging works here :)
08:59:11t0maserh: lcd -> remote
08:59:49dwihnoLCD hotplugging support would be nice though :)
09:01:23t0maslittle hard to accompilsh... hardware wise :)
09:01:49t0masman... I'm going to get some more sleep... typing really bad :X
09:04:04t0masoh Slasheri: bug report... when it's paused in the middle of a song...
09:04:23t0masand I press prev, it gets to the beginning, but it doesn't update the screen untill it's unpaused
09:04:56Slasherit0mas: ah, that will be solved later (not the easiest things to do)
09:05:47SlasheriAnd i found the problem..
09:06:47t0masah, where was it?
09:07:38Slasherithere is no single line :)
09:07:56Slasherisolving it requires a new global variable
09:11:39 Join ten-ten [0] (
09:11:40Slasheriit works :)
09:11:52Slasherimuch better next track switching now
09:12:18 Part ten-ten
09:13:04Slasheriaargh.. now pause stopped working :D
09:13:48Slasheriah, that was because i commented it out =)
09:16:59Slasherithere is still some pause problems but i will commit now
09:21:25***Saving seen data "./dancer.seen"
09:22:11BagderSlasheri: I think you should add stubs for thos functions to the sim, then you won't have to add so many #ifndef SIMULATOR, which makes nicer code imho
09:22:23 Join Chamois [0] (
09:24:31amiconnOooh... I just did an experiment... burst mode _does_ work with movem
09:24:44amiconn...but only if the start address is _line_ aligned
09:25:02amiconnSpeed does more than double compared to simple logword writes
09:25:14Bagder_line_ as in cache line?
09:25:36amiconnLine as in 4*4 bytes. There is no data cache
09:25:43Bagderah, right
09:25:47*Bagder slaps forehead
09:28:58SlasheriHmm, somebody should add config menu entries for cross-fader :)
09:34:44 Join linuxstb [0] (
09:37:00Bagderlinuxstb: a minor nit: we write the code within 80 columns
09:44:22linuxstbBagder: OK, I'll stick to that. But who has an 80-column text editor in 2005?
09:44:54Bagderand I know Linus has
09:44:57Bagderand Zagor too ;-)
09:45:03linuxstbOK :-)
09:45:13BagderI have like 3 windows next to each other
09:46:01linuxstbI always just have one full-screen window. Guess I haven't adjusted to this windows business...
09:46:46linuxstbDo you think that FLAC metadata parser is a bad idea? I'm not sure about it.
09:47:21Bagderno, we need it
09:47:53Slasherilinuxstb: it works quite well. Some support for elapsed information need to be added also and more yields to codecflac.c
09:48:23Bagderand when we add a vorbiscomment parser, both flag and vorbis can use that
09:48:43linuxstbBagder: My thinking exactly. I don't think there are too many different tag types
09:49:12Bagderright, there seem to be only a few pretty standard ones we would be interested in
09:51:05linuxstbCan someone explain the "while (!*ci->taginfo_ready) rb->yield()" in plugin_start in codecmpa.c ?
09:51:20linuxstbShould codecflac.c (and others) have the same check?
09:51:37Bagderbtw, are the codecs still loaded in the common plugin space?
09:51:54Slasheriyes they should have it _if_ they need to read some metadata information from ci
09:52:14SlasheriBagder: yes they are, LinusN is working on that
09:52:38BagderI find all the bug reports on crashing plugins while playing music rather amusing ;-)
09:53:02linuxstbAm I right in saying that the code in playback.c runs in a different thread to (e.g.) codecflac.c ?
09:54:22Slasheriyes, there are two threads in playback.c
09:54:43Slasherione of them runs the buffering stuff and other one the codec
09:56:08SlasheriHmm, and buffer filling seems to be broken with flac again.. Maybe some updates to playback.c killed it
09:56:14*HCl stretches
09:56:27Slasherihi HCl :)
09:58:17Slasherioh, i broke it :)
09:58:57linuxstbSlasheri: Yes, I noticed that - I thought it was my metadata parsing, but then I checked earlier builds, and they were broken as well. Have you found the bug?
09:59:08Slasherifixed :)
10:03:45 Nick cheriff_AWAY is now known as cheriff (davem666@wagner.orchestra.cse.unsw.EDU.AU)
10:05:32 Nick Lynx_awy is now known as Lynx_ (Lynx@
10:06:07*HCl is dead tired :/
10:06:12 Part dapureplaya
10:10:23 Quit ripnetuk ("Off To Download Festi :)")
10:17:32[solid]morning... in case you guys didn't read the logs - flac playback went on and on for a freakin' 8 hours and ~30minutes
10:17:57[solid]kinda unbelievable providing iriver fw is doing 8-10 with ogg vorbis q6, eh?
10:18:29LinusN[solid]: that's nice
10:18:56LinusN[solid]: would be interesting to see how we perform with ogg
10:19:26[solid]LinusN: yup, i'll be testing that as soon as i'm back from the uni today
10:19:41 Join Patr3ck [0] (
10:19:46LinusNhow fun for you :-)
10:20:22t0mas[solid]: maybe post your findings to the IriverBattery page?
10:20:34t0maswith the exact bitrate / quality settings of the music?
10:21:07[solid]LinusN: i'd hope to see at elast 6hrs seeing how well flac flac performed and provided the oggs of the same album are 72MB, not 337MB:D
10:21:41[solid]t0mas: i think i will, but i'd like to perform ogg vorbis and mp3 test beforehand for the sake of completeness
10:22:05[solid]t0mas: the files were amx compression (-8), average bitrate of all files 852kbit/s
10:23:44[solid]t0mas: also, remote was plugged in, and headphones (koss ksc75) into line-out, volume set at 50%
10:24:17t0mashm.... the running time left indicator should be written too...
10:24:33t0mashow have you done the benchmarking?
10:24:45t0maswritten to some memory buffer what the battery level was?
10:25:08[solid]oh it wasn't really a benchmark... i just tested for how long it would play
10:25:21[solid]it stopped ad 3% battery
10:25:23linuxstbSlasheri: I'm still getting problems playing back FLAC files - they seem to stop playing after a couple of minutes. Pressing next/prev brings it back to life.
10:25:42[solid]and the end was qquite funny... there was so little battery that the disk couldn't spin ^^
10:25:49t0mas[solid]: can you benchmark it sometime too?
10:26:12t0masso make it write battery levels to a memorybuffer once in 10 minues or something
10:26:34[solid]t0mas: yeah, i was thinking i could do that if someone provides me with a build that does it, so the indicator could be improved
10:26:40t0mas8 hours = 8 * 60 = 480 minutes... hm... maybe just to 480 readings
10:26:53t0mas[solid]: that's what I was thinking of...
10:27:12t0masbut we'll have to wait for Slasheri to allow plugins
10:27:24t0masas I don't know how to do it anywhere else...
10:28:02[solid]great. i'll be so glad to help!
10:29:40Slasherilinuxstb: Hmm, if you can fix it, please do so :) I have to go now
10:30:28t0mas[solid]: maybe I can hackup the power thread to do it...
10:30:34t0masthat would save us waiting for Slasheri
10:32:09Slasherit0mas: I think that LinusN is working on separating codecs from plugins
10:32:41LinusNSlasheri: you want a crossfade setting?
10:33:00linuxstbSlasheri: Sorry, I don't have any time to do anything now - need to start my real working day.
10:34:40SlasheriLinusN: gladly, i don't think that even every people wants to use crossfading :)
10:35:32Slasherilinuxstb: as well i have to go to work also :/ cu later :)
10:35:50 Join Aison [0] (
10:36:24amiconnt0mas: There is already some debug code for logging in the power thread
10:36:40t0masbut I've to run now...
10:36:42t0mastrain to catch :)
10:37:25amiconnSome nice memset hacking due this evening.... speed increase from burst mode is really noticeable :)
10:39:22amiconnWriting 8000000 bytes (cpu @48 MHz): single bytes: 3.11 s, longwords: 0.78 s, line bursts: 0.32 s
10:40:48 Join ehntoo [0] (
10:42:51LinusNamiconn: i thought you didn't see any difference with the burst mode
10:43:04amiconnThat was because I'm silly
10:43:17amiconnReading the datasheet properly does help a lot ;)
10:43:19LinusNah, that
10:43:52amiconnmovem does only burst if the start address is aligned to line size
10:44:32LinusNdidn't know that
10:45:23amiconnDo you think the refresh rate bug of the v2 bootloader should be added to the wiki page?
10:45:46amiconn(cuurently it says no known bugs)
10:47:57 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
10:51:27[solid]oh boy, so nice all supported files now have a working wps... that might finally motivate me to make one
10:52:35linuxstb[solid]: Have you tried the latest bleeding-edge builds? I have problems with FLAC playback stopping after a couple of minutes - I haven't tried other formats.
10:53:15[solid]linuxstb: i just got the bleeding edge, so i'll give it a shot
10:53:54linuxstbThe latest daily build should have the same problem - unless it's just me.
10:55:35 Join preglow [0] (
10:55:58preglowi'm not the only one having trouble connecting to sourceforge, no?
10:59:10linuxstbWhich part of sourceforge? web, cvs, shell ?
10:59:25preglowpatch right now
10:59:42preglowso web
11:01:06Lynx_how do i get rid of the recycle bin and system restore folders on an external drive again?
11:08:36[solid]linuxstb: it's still playing... elapsed time would be lovely for such tests ;)
11:12:20preglowwhat're you testing now?
11:14:09[solid]10:52 < linuxstb> [solid]: Have you tried the latest bleeding-edge builds? I have problems with FLAC playback stopping after a couple of minutes - I haven't tried other formats.
11:15:25preglowwhat format is next up on the battery testing? read about the flac playback time, and it's quite amazing
11:15:45dwihnoThere's already been some testing?
11:15:50dwihnowhoa! in the wiki?
11:16:08 Join bobTHC [0] (
11:16:19[solid]dwihno: you can play 8hrs of flac with a rockboxxxed h120 :)
11:16:28[solid](well, over 8hrs)
11:16:33bobTHCmornin' all !
11:16:37dwihno[solid]: I don't know if that's good or bad :)
11:16:45[solid]dwihno: incredibly good imo
11:18:05 Quit webguest42 ("CGI:IRC (EOF)")
11:18:15amiconn[solid]: It should still get better. There are some things which currently limit runtime
11:19:18[solid]amiconn: yeah yeah i know:) i'm thinking another hour might be added, you reckon that possible? with codec and core uber-optimalizations and a tuned buffering system...
11:19:42[solid](these are what i heard that are going to be done)
11:21:10[solid]and with the remote buttons working i can finally use rockbox when outdoors too... pimpin'
11:21:16linuxstb[solid]: Is FLAC still playing OK for you? For me, it is stopping before it's finished one track.
11:21:23[solid]linuxstb: yup
11:21:28***Saving seen data "./dancer.seen"
11:21:29amiconnThere's currently the problem that the cpu is always boosted, hence it draws more power
11:21:55linuxstb[solid]: thanks. Not sure if that's good or bad news. I'll do some more tests myself.
11:22:18preglowoh, yes
11:22:19[solid]linuxstb: maybe something went wrong with updating? if it's stoppping before a track ends it could be connected to buffer refilling which should be okay in the build i have
11:22:21preglowso that's still there
11:22:22amiconnAnd making the codec thread yield more often will result in less buffering time, hence ´the hd will spin less
11:22:32amiconn...and _that_ will increase runtime for sure
11:23:16Slasheriamiconn: the cpu is not always boosted and i tested yielding codec thread much more often -> no decrease in buffering time. Maybe more tests should be done
11:23:34[solid]still 8hours is something you could literally boast about... ipods can do 6hrs of alac and 2hrs of flac (with ipodlinux) iirc
11:24:31dwihnoHow's the audio quality of the iriver compared to the archos?
11:24:40preglowwavpack is very efficient now
11:24:46preglowby far the most efficient codec in rockbox
11:24:48amiconnSlasheri: The cpu is always boosted for me, with cvs from this morning
11:24:53preglow400% realtime WITH disk writing
11:24:54amiconnCan't test atm
11:25:01[solid]preglow: w00t
11:25:11[solid]preglow: would love to test it also ^^
11:25:26Slasheriamiconn: hmm, when you are playing? it should be boosted always while buffering, after that only when it's necessary
11:26:00Slasheriamiconn: try playing longer and keep eye on the boost indicator in debug menu
11:26:57amiconnIt's boosted right at rckbox start, and stays there
11:27:14amiconn(playing mp3)
11:27:17preglowcommited a few new opts now
11:27:34[solid]okay, time to go pimpin' with lossless + gapless chemical brothers... the gaps on this one were *so* irritating. vorbis test as soon as i come back.
11:28:35amiconnSlasheri: re the buffering time - perhaps the ata transfer is slower than I thought... in that case it should be possible to do something about it
11:28:36Slasheriamiconn: Hmm, yes. That startup boost thing should be fixed. But on my iriver, it's not boosted always when playing mp3s. Just tested
11:28:58amiconnWe are able to read 4 MB/s on archos...
11:29:31[solid](i just tested it also, while playing flac, and it wasn't boosted indeed)
11:29:36Slasheriamiconn: iriver is still able to buffer everything in few seconds without any other load. But running codecs requires quite much cpu power
11:29:57preglowSlasheri: yo, any reason for the AFMT flags being formatted to fit a bitfield?
11:30:08Slasheripreglow: that's not mine idea :)
11:30:16preglowthis limits us to 32 codec
11:30:17Slasherii got the defined from wiki site
11:30:30preglowi can see no reason for this being a bitfield
11:30:52Slasheriok, maybe that can be changed.. but no i have to go again ->
11:31:11amiconnSlasheri: Then I think there's still something wrong with the yields. If libmad is able to run at 300% realtime, and we're just playing, it should draw ~33 % cpu power
11:31:24amiconn...and that should not even double the buffering time
11:31:31linuxstbSlasheri: My FLAC playback problem was a false alarm - I just had some corrupt FLAC files.
11:32:09linuxstbIt may be useful to keep them as bitfields (capabilities of codecs). Why don't we keep them as they are for now, and change them if we ever need 33 codecs.
11:36:41amiconnlong long would allow for 64 codecs ;)
11:39:23linuxstbI noticed that the id3 struct still contains lots of ints - should these be changed to longs?
11:40:00amiconnOnly if the value might not fit in a 16 bit int
11:41:02linuxstbThere is filesize, length, elapsed...
11:41:44amiconnlength and elapsed are in milliseconds, correct?
11:41:55amiconnIf so, they should really be changed to long
11:42:03LinusNand filesize too
11:42:26amiconnof course
11:42:28 Join ]RowaN[ [0] (
11:42:41preglowmy, isn't source forge responsive today
11:42:44 Quit Aison (Read error: 60 (Operation timed out))
11:43:04]RowaN[should this experimental crossfader work on its own or do i need to enable it somewhere?
11:43:39 Join Aison [0] (
11:47:18preglowworks fine here
11:47:23preglowand i haven't enabled anything
11:47:24LinusN]RowaN[: it is enabled by default, but it is very short
11:47:34]RowaN[k, bout to try it
11:47:42LinusNi just committed a setting for it though, which is off by default
11:48:28LinusNGeneral Settings->Playback->Crossfade
11:49:31linuxstbI've just added some code to codecflac.c to keep track of the elapsed time (it's also done in codecmpa.c), but it doesn't seem right to do it in the codecs.
11:50:55preglowmy player hangs at 'starting original firmware'
11:51:02preglowperhaps it's time to recharge it a bit :)
11:51:57preglowwhy is it that it actually refuses to start with too little battery?
11:52:05preglowcan't imagine that's caused by anything other than an explicit check
11:52:23Bagderdisk spinup problem perhaps?
11:52:40preglowmp3 playback has some nasty playback bugs
11:52:46preglowwelcome to anyone who wants to debug my code
11:53:12LinusNpreglow: what bugs?
11:53:21preglowLinusN: it sounds like shit
11:53:45LinusNlike "pfrrrrbbrr"?
11:53:45preglowLinusN: gimme a sec and i'll provide an example
11:53:52preglownasty background noise
11:54:01LinusNyeah, we all know what shit sounds like
11:54:37preglowand it's all over the playback!
11:54:39preglowbut no, it's there
11:54:50preglowit might come from the use of lower precision in synth.c
11:55:22preglowi could try fixing that to see how bad the performance hit is anyway, i guess
11:55:48DMJCany eta on when lower than 41000 bitrates will work?
11:56:17preglowLinusN: and btw, is switching sample rate to 22050 etc. any problem?
11:56:45LinusNno, it's not a problem
11:57:05LinusNhowever, we should probably resample anyway
11:57:30preglowto 44.1khz you mean? if so, why?
11:57:30LinusNotherwise the S/PDIF output won't work
11:57:38DMJCone track I have sounds like chipmunks playing it heh
11:57:39preglowspdif supports other sample rates
11:57:46LinusNpreglow: yes, b ut not all
11:57:58DMJCactually it's hillarious because it's a metal piece
11:58:29LinusNs/pdif supports 48, 44.1 and 32 iirc
11:58:37preglowLinusN: but yeah, i did an experiment with linear interpolation and resampling from 48khz to 44.1khz, sounds completely decent
11:58:57preglowthere's a wee gain change
12:00:13preglowbut linear interpolation prolly won't do for 22050 -> 44100
12:00:19preglowat least, there'll be tons of high frequency noise
12:00:52linuxstbWe couldn't do crossfading if we didn't resample everything to 44.1Khz. So in a way, the lack of 48KHz support simplifies everything for us.
12:01:21linuxstb... until the next hardware Rockbox-with-software-codecs is ported to.
12:01:54preglowcrossfading across tracks with different sample rates does make sense, i guess
12:02:13DMJCwill rockbox goto h300?
12:02:22LinusNDMJC: yes
12:02:29DMJCmate of mine will want that
12:02:37linuxstbDo we know if the h300 does native 48KHz?
12:02:46preglowalmost certainly not
12:03:49LinusNno it doesn't
12:04:13LinusNnot doing 48000 is a flaw in the coldfire cpu
12:04:51preglowany of many :)
12:04:56preglowone of many as well!
12:05:03preglowi suck profoundly
12:05:40LinusNpreglow: yes you do
12:05:48linuxstbpreglow: Happy birthday, by the way.
12:06:35preglowthanks, thanks
12:07:17preglownot very happy, though, the weather is still depressing
12:07:44LinusNpreglow: congratulations
12:15:29preglowplease, god, don't let this be a bug in imdct36
12:21:25LinusNlogbot seen god
12:21:36LinusNdoesn't seem to be around
12:25:07preglowelusive fellow
12:25:21preglowdang, seems like i might actually make this a bit faster as well
12:26:56LinusNcool, how?
12:28:00preglowlike it is now, dct32 does a lot of shifts to be able to use 32x32 -> 32 mul
12:28:24preglowthink i'll be able to eliminate all of those, and a mac.l + movclr.l should take exactly as much time as a muls.l did anyway
12:29:52preglowwe'll see, i have been mistaken about optimisation benefits in the past...
12:34:56linuxstbpreglow: Did you say you had a (non-working) codecwavpack.c ?
12:35:09*[solid] is back
12:35:51[solid]charging for testing vorbis now..
12:36:51preglowlinuxstb: david bryant is on it
12:37:10linuxstbOK, I'll leave it with him then.
12:37:48preglowLinusN: what are your thoughts on giving him cvs access, btw? he seems to know what he's about, seems a bother to have us commit his patches all the time
12:38:08preglowinsert words at your convenience to enhance readability
12:38:18preglowthe wavpack fellow
12:38:41LinusNdavid bryant
12:38:48LinusNyes, that would be ok
12:38:57LinusNsend him my way
12:38:58preglowi emailed him and asked if he was interested
12:39:03preglowi'll send him your way if he is
12:40:26linuxstbDoes anyone fancy writing a simple OGG parser to extract the metadata - similar to my FLAC parser?
12:42:09preglowargh, someone utterly insane needs to rewrite synth_full in pure asm for me
12:44:40Zagorlinuxstb: you mean the tags?
12:45:27linuxstbInitially just things like totalsamples, frequency etc, but yes, eventually the VORBIS_COMMENTS as well.
12:45:55Zagorok. i was going to say bagder did that, but that was in perl... :-)
12:46:02linuxstbSee my recent commit to playback.c that does the same thing for FLAC.
12:46:10Bagderyeah, and I only found an existing package ;-)
12:46:17Bagderdidn't do much myself
12:46:49linuxstbThe thing that worries me about vorbis comments is that they are just TAG=VALUE - the user can use any tags they want. But I'm guessing that applications follow conventions.
12:47:20preglowamiconn: and btw, mp3 playback isn't 300% realtime, as it currently uses a very inefficient dither
12:47:22[solid]linuxstb: aren't apev2 tags doing it the same way?
12:47:34linuxstbI have no idea, but it wouldn't surprise me.
12:48:00Bagderlinuxstb: I think we should just assume the most commonly used comments, and ask users to adapt
12:48:27linuxstbYes, I'm sure that's what every other application does.
12:48:44BagderI guess so too
12:49:02preglowyou've got stuff like TITLE, ARTIST, ALBUM
12:49:12preglowthey're pretty standard
12:49:16linuxstbBut for example, the specification suggests (encourages) that users use tags such as ARTIST multiple times to indicate multiple artists.
12:49:31preglowi've never seen that actually done
12:49:46Bagderyes, that scared me too
12:50:20linuxstbIt would be nice (eventually) to allow the WPS to access any vorbis tag - so if the user wanted to store unusual tags, Rockbox could display them.
12:50:42[solid]TRACKNUMBER,TRACKTOTAL and DISC are also commonly use
12:50:42 Join midk_ [0] (
12:50:45Bagderand potentially quite memory consuming
12:51:29[solid]+ the replaygain ones, but these are of no interest yet i guess
12:51:30preglowit actually ended up a bit faster indeed
12:51:40 Quit midk (Read error: 54 (Connection reset by peer))
12:51:45linuxstbAnyone know what wavpack uses for metadata?
12:52:05Bagderno idea
12:52:17[solid]linuxstb: i think apev2, but i'll look that up
12:52:43dwihnoI saw apev2 on a mp3 once
12:53:14preglowwavpack uses apev2 a lot
12:53:22[solid]yeah you can use apev2 wiht mp3's
12:53:27preglowlinuxstb: david bryant said he'l look at that as well
12:53:50dwihno[solid]: If you ask me, id3v1 and v2 is all that's OK for mp3...
12:54:07linuxstbSo it looks like we have ID3, ID3v2, VORBIS_COMMENTS and APEv2 (with the last two looking very similar)
12:54:17[solid]dwihno: id3v2 is very limited, let alone id3v1
12:54:55dwihno[solid]: It's metadata. It's not supposed to be all shiny and stuff :)
12:56:14[solid]apev2 support for mp3's also would be great, that would mean foobar's tags (with unicode and replaygain data) for mp3's would work...
12:56:25Zagori'd say id3v2 contains way more tags than 99% of people are using
12:56:51dwihnoid3v1 is all you ever need :)
12:57:08dwihnothe 1.1 extension is cute too
12:57:23[solid]besides id3's suck when you have national characters in your tags
12:57:41[solid]i have, for example, many polish tags and many western-european ones
12:57:44Bagdernot really
12:57:46Zagorworks for me
12:57:47Bagderits unicode
12:57:52dwihnoworks nicely for me too
12:57:57[solid]there's no way with id3 for them to work...
12:58:02Bagderyes there is
12:58:06[solid]you always have to choose one codepage
12:58:22Bagderread up
12:58:30Bagderon the recent specs
12:59:46[solid]so how come it doesn't work with oh-we-are-so-conforming-to-everything foobar2k? ^^
12:59:58[solid]i've tried many times but with no success :(
13:00:27Bagderare you seriously asking me?
13:00:47Bagderread the spec, you can clearly see unicode all over
13:01:11Bagderif players don't support that, if you use an older id3 version or whatever, I can't possibly tell
13:01:17[solid]i think i'll give up, spec's are scary
13:01:28[solid]oh... so it's v2.4, right?
13:01:42BagderI don't recall
13:01:57[solid]that would explain a lot
13:03:50[solid]i'm a ogg zealot anyway ;)
13:04:59[solid](and i agree that id3v2 contains way more tags than 99% of people are using)
13:05:17 Quit DMJC ("Leaving")
13:06:44 Join zezayer [0] (~chatzilla@
13:11:10amiconndwihno: id3v1 doesn't suffice for most of my mp3s
13:11:24amiconnThe limited field length almost always hits me
13:11:31Bagderyeah, that's pretty lame
13:11:33 Join DMJC [0] (
13:11:34preglowyes, indeed
13:11:41dwihnoamiconn: Old standard, small fields.
13:11:49preglowtoo simple standard
13:11:50dwihnoOld standard, non-dynamic fields :)
13:12:09dwihnoOld song, magnetic fields ;)
13:12:14 Quit edx (Read error: 110 (Connection timed out))
13:12:37amiconnHowever, I always add an id3v1 in addition to id3v2 where id3v2 is necessary, and id3v1 only when it is sufficient
13:13:26*preglow giggles at flac playback
13:14:56Zagoramiconn: boo, then you waste battery by forcing a seek ;-)
13:16:20amiconnIsn't the seek aways necessary, to check whether there is an id3v1?
13:16:31 Nick QT_ is now known as Qt (as@area51.users.madwifi)
13:18:22[solid]mmm... elapsed time for flac is lovely ^^
13:18:44Zagoramiconn: not if an id3v2 is found, iirc
13:19:15 Join spiralout [0] (
13:19:53amiconnZagor: Hmm, maybe if tag priority is set to id3v2 first (like I have it set)
13:20:37 Quit [solid] ("leaving")
13:20:52linuxstbCan anyone find a straightforward specification of the .ogg format? I'm going round in circles on xiph's website. I just want to find the samplerate, totalsamples and VORBIS_COMMENTS...
13:21:32***Saving seen data "./dancer.seen"
13:21:32 Quit ehntoo (Remote closed the connection)
13:22:12amiconnZagor: Anyhow, I'm not that concerned about the seek, especially on the Ondio ;-)
13:24:28spiraloutis it a known bug that after playing one song the same song starts again?
13:28:23LinusNspiralout: yes, it has been reported earlier
13:28:31LinusNeven on the archos
13:29:13spiraloutoh ok thanks than is allright you guys will do it
13:30:53ZagorLinusN: have you been able to dig up any more clues about that than are in the bug report?
13:31:14LinusNlogf may save us here
13:32:29 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
13:35:48amiconnLinusN: That's really a known bug? I never observed it...
13:36:01LinusNwe have a bug report
13:36:23LinusNand i've seen it on the iriver
13:37:28preglowusing full accuracy didn't help a bit
13:38:43preglowcould someone please give me a peak realtime percentage reading for iriver.mp3 (using mpa2wav av 120mhz)
13:38:52preglowso i can see if i should commit this nonetheless
13:39:31amiconnLinusN: You said even on archos.... I have 3 archoses, and I really never observed it
13:40:52LinusNamiconn: the bug report says archos, i can't help that
13:42:12 Quit DJ-suparappa ()
13:42:23Zagori don't feel comfortable releasing 2.5 with that unresolved
13:42:50amiconnLinusN: Hmm, different thing: Did you verify the MFDR(2) settings?
13:45:46preglowno takers? :>
13:48:14*HCl went from dead tired to simply pretty tired
13:48:21linuxstbI have crossfade turned off, but I seem to be losing a couple of seconds from the end of my FLAC files.
13:48:48linuxstbDoes anyone else get that?
13:51:00preglowlinuxstb: do you have any idea why mpa2wav realtime percentage increases steadily for some files?
13:51:13preglowlinuxstb: for iriver.mp3, for example, it never stops increasing
13:52:20linuxstbpreglow: No.
13:55:11 Join Musicmad [0] (
13:58:13preglowit went from 116.78% to 119.01% with disk write, so i guess i'll commit it
14:03:34Musicmadjust a thought - would it be possible to make non-lame mp3s gapless by skipping the last frame entirely or something like that?
14:04:28preglowbut then you'd hear a great big click
14:04:35Musicmadok :)
14:05:47linuxstbI don't use MP3s, but there must be a tool out there to add lame headers to non-lame mp3s - i.e. to do the silence detection "off-line".
14:06:48preglowof course
14:06:58preglowbut you'll never be certain it's correct
14:07:06preglowbut that doesn't matter much as long as you can't hear any gap, i guess
14:08:18*t0mas is back
14:09:25linuxstbIMHO, trying to detect and remove silence from poorly encoded MP3s is a waste of time. But everyoine has their own interests.
14:13:54t0maswhat should I include for the logf() function?
14:16:10 Join Shagnar [0] (
14:16:55Zagorlinuxstb: why a waste of time?
14:16:56preglowi solved it!
14:17:03 Join elinenbe [0] (~elinenbe_@
14:17:41preglowmp3 playback is now smooth like a babys bottom
14:19:23linuxstbt0mas: Are you trying to use logf in a plugin, or in rockbox itself?
14:19:28Shagnari know it's no real bug, but today i listened to some files in shuffle mode... sometimes a file is played 2 times...
14:19:45preglowShagnar: i don't think that should happen
14:19:56Shagnarthis morning in the train... it happened
14:20:09amiconnpreglow: What was the problem?
14:20:13elinenbepreglow: everything? ff/rw/pause/crossfading/ff through crossfading/rw through crossfading/pause during crossfading/etc?
14:20:34preglowamiconn: too little precision in libmads default multiplication routines
14:20:43preglowelinenbe: no, i mean mp3 playback, as in the libmad specific part
14:21:04ZagorShagnar: was that on iriver or archos?
14:21:13Shagnaririver ;)
14:21:32Shagnarbleeding edge version updated this morning...
14:21:59linuxstbZagor: Because it seems the least elegant solution. To me, the solution is for the MP3 file to know it's own length (i.e. via the lame header).
14:22:01t0maslinuxstb: rockbox itself
14:22:09preglowamiconn: however, it does seem to cost a tiny bit of performance, i however think that's acceptable
14:23:09Zagorlinuxstb: right, but when the format itself is bad we can only do as best we can. saying "screw all you who don't rip your own albums" isn't really helpful.
14:24:05Zagorunless we can create a fast and accurate offline tool to add this tag to mp3 files
14:24:21preglowthe glitches i could hear with loud bass drums have vanished as well
14:24:58linuxstbt0mas: #include "logf.h" :-)
14:25:21t0mask :)
14:25:26t0masit logs to the remote on iriver right?
14:26:08amiconnYou need to configure a logf-enabled build
14:26:26linuxstbZagor: That's what I was asking - does such a tool exist already? If so, then maybe we don't need to worry about doing silence detection in rockbox so much.
14:26:54 Join edx [0] (
14:26:59t0masamiconn: I already have :)
14:32:49preglowdoes logf work from a plugin now?
14:33:38amiconnlinuxstb: Silence detection won't work for finding how much of the last frame isn't supposed to be there
14:34:22preglowthere is no silence in any frame that wasn't pure digital silence before it was encoded
14:34:30amiconnAn mp3 frame can't be partially silent, due to the way mp3 compression works
14:34:30preglowany non-zero sample will be smeared over the entire frame
14:35:19preglowhowever, the waveform will decay, so some kind of thresholding can be done
14:35:25preglowit won't be accurate, though
14:37:37linuxstbThis (old) thread on HA talks about a fb2k plugin that allows you to modify the lame header. Not sure if it's automatic though.
14:38:51Musicmadis crossfading being worked on? (or do I misunderstand the cvs commits).
14:42:13preglowit's partially working
14:42:24preglowthat is, it does work
14:42:28preglowbut it's not very customisable
14:43:18elinenbeSlasheri said yesterday that he was working on adjustable cross-fade time and curves... although he might have just been working me!
14:45:09preglowso, what's the general consensus on something that really enhances mp3 quality but shaves off a wee bit of performance?
14:45:17preglowshould i put it in?
14:47:42amiconnI'd say just do it
14:48:13linuxstbDefinitely - as long as it is > 100%, I say go for quality.
14:48:37preglowoh, it should be by far
14:48:57preglowi'll compare the runtime to the daily mpa2wav.rock compile
14:49:11linuxstbIf it improves quality that much, then IMO it's a bugfix, not an enhancement.
14:49:22preglowit's _really_ noticable to me
14:49:30preglowespecially during more quiet parts
14:49:36Musicmadwhat is it?
14:49:45amiconnpreglow: Btw, I also had the impression that libmad does something slightly wrong, but I wasn't sure at all
14:50:28amiconnI think libmad will exceed the MAS decoding quality with your latest fix
14:50:53preglowwe'll see
14:51:04preglowlibmad should spit it audio at around 20 bits precision now
14:52:46preglowis CONFIG_CPU defined in a Makefile?
14:52:48preglowi guess not
14:53:02linuxstbShould the dither routine take account of that precision? I'm guessing that the code I plagarised assumes 24-bit precision, but I never really tried to understand it.
14:53:55preglowno, dithering only cares about output device precision
14:54:13preglowand assumes we've got more precision from the codec than it can provide
14:54:22preglowwhich is certainly the case now
14:54:25 Join MoosCamaro [0] (
14:54:38 Quit Rori ()
14:54:50MoosCamaroHi all
14:57:23t0maspreglow? do you know if it's ok to use strtok() in rockbox?
15:00:13linuxstbt0mas: If it's defined in firmware/include/*.h (which it is - string.h), then I would guess it's OK to use.
15:01:10amiconnt0mas: rockbox has strtok_r(), not strtok()
15:04:06 Join ashridah [0] (
15:06:26 Part LinusN
15:09:23preglowcan anyone tell me how i go about adding architecture specific stuff to makefiles?
15:12:29 Quit Musicmad (Read error: 145 (Connection timed out))
15:13:34preglowshould i perhaps add the arch dependent define i want to config-h100.h ?
15:16:36amiconnWhat do you need/ intend to do?
15:17:54preglowadd the define FPM_COLDFIRE_EMAC for mcf5249 build
15:17:57preglowDEFAULT for everything else
15:18:02preglowFPM_DEFAULT, at that
15:18:49amiconnWhy can't you just check for CONFIG_CPU == MCF5249
15:19:03linuxstbIs anyone else experiencing Rockbox cutting off the last two seconds of FLAC files - with crossfading disabled?
15:19:15preglowamiconn: does that work in makefiles?
15:19:33amiconnNo; why do you need this in a Makefile?
15:19:35preglowamiconn: ahh, no, because i need to get rid of the FPM_DEFAULT default define to make it work better
15:20:08preglowamiconn: when libmad sees FPM_DEFAULT defined, it starts doing weird things i don't want to work around with hacks or #if 0 sections
15:20:11amiconnlinuxstb: Sorry, not a single flac file here, and iriver @home...
15:20:45amiconnpreglow: Iirc there is a central .h for libmad
15:20:53amiconnYou could just do
15:21:01 Quit Bger (Read error: 60 (Operation timed out))
15:21:03amiconn#if CONFIG_CPU == MCF5249
15:21:03preglowbut that's not very nice, now is it...
15:21:12amiconn#undef FPM_DEFAULT
15:21:36***Saving seen data "./dancer.seen"
15:21:36amiconnWhere is FPM_DEFAULT defined?
15:21:37 Join Bger [0] (~Bager@
15:21:44amiconnIn the libmad Makefile?
15:21:49preglowit's defined in the makefile now
15:23:01amiconnYou could move it to global.h then
15:24:15preglowi'll just add a note in the Makefile where to look for it
15:24:42preglowsince this really is more of a build system decision
15:26:07 Join [solid] [0] (~solid@
15:26:27[solid]anyone experienced problems with ogg vorbis playback?
15:26:42[solid](bleeding edge)
15:27:05[solid]i'm experiencing some weird skipping
15:28:12[solid]don't know if it is skipping, but it sounds like sometimes it's playing not the samples it should
15:28:41[solid]also i had a problem with flac (same build) - it sometimes doesn't switch to next track, it repeats the same one
15:28:52[solid]if you press next - it skips two tracks
15:29:02[solid](ie it behaves like it thinks it's playing the right track)
15:29:07linuxstb[solid]: I've only been playing FLACs, but I'm getting exactly the same problems as you.
15:29:33linuxstbIt does sound like samples are being skipped occasionally.
15:30:13linuxstbI'm playing an OGG now, and it's skipping heavily.
15:31:20[solid]did anyone have this track repeat problem yet?
15:31:27[solid]that one seems totally random
15:31:51linuxstb[solid]: Do you have crossfade turned on or off?
15:31:57ashridahwhoa, yeah, it's buggy as hell
15:32:18ashridahcrossfade's off here
15:32:31linuxstbI'm going to try OGG with crossfade on.
15:32:44[solid]linuxstb: it's off now
15:33:33ashridahi press stop and it fades out the song, then the song kicks back in at the original volume even tho the menu's showing
15:33:49[solid]linuxstb: but this track repeating happened to me today, before linus' setting commit
15:34:09[solid]linuxstb: so crossfade was on then
15:34:19linuxstbI think track repeating and the skipping are different problems - I'm more worried about the skipping.
15:34:39[solid]i also think they are different, and also think skipping is more 3vil
15:34:44ashridahand it's also decided to start rejecting some mp3's i've got again
15:35:31ashridahhrm, actually, i saw something flash on about it being unable to load a codec
15:35:45[solid]i'm playing some mp3's to see if the skipping is there also
15:36:02linuxstbI'm not sure, but I think enabling crossfading eliminates skipping in the middle of OGGs.
15:36:03ashridahworked after i restarted it
15:37:06[solid]linuxstb: i'll test it first then
15:38:13preglowthe old problem i had returned when i moved the FPM def to global.h
15:42:02t0masdoes anybody know how to save am image in black/white in the gimp?
15:42:16t0mas(so 1 bit per pixel)
15:42:32[solid]linuxstb: indeed, with crossfade on the skipping seems to be gone
15:43:50linuxstb[solid]: Yes, OGG hasn't skipped for me since I turned crossfade on.
15:43:50linuxstbI'll try turning it off again...
15:44:03preglowbloody libmad
15:48:24 Join webguest92 [0] (
15:49:50 Quit webguest92 (Client Quit)
15:52:39[solid]linuxstb: it's crossfades fault indeed
15:52:56[solid]linuxstb: i've turned it on and it started skipping immediatly after switching to next song
15:53:04linuxstb[solid]: I can't make it skip any more - with crossfade on or off!
15:53:21[solid]i've turned it off*
15:53:56[solid]now, the same track with crossfade on..
15:54:05[solid]works great.
15:54:09DMJChow optimized can this firmware get?
15:54:17DMJCI've been noticing lag when navigating menus
15:54:23[solid]but i prefer gapless to crossfade i must say;)
15:55:45[solid]turned gapless off... and it's skipping
15:56:24linuxstbYou mean "turned crossfade off"?
15:56:28[solid]oh yes
15:57:28[solid]hmm... i'll try now if that other bug, with track repeating is also dependant on crossfade on/off
16:00:19[solid](from what i recall it never happened with crossfade off)
16:04:40[solid]it happened again
16:05:59[solid]so there's one bug when crossfade is on (track repeating) and one when it's off (skipping)
16:05:59[solid]tough choice :)
16:07:46 Quit Rick (Read error: 54 (Connection reset by peer))
16:09:04 Quit Chamois (" HydraIRC -> <- IRC has never been so good")
16:09:19 Join Rick [0] (
16:09:46preglowsilky smooth sound
16:10:15[solid]i was wrong
16:10:28[solid]it repeated with crossfade off now, too
16:11:18[solid]but i haven't experienced it with yesterdays build... (it didn't repeat any track for... 8hours:P)
16:11:59linuxstbThe other changes which could be the cause were mine - the code which enables the WPS.
16:15:24linuxstbpreglow: Are you happy with your libmad changes now?
16:17:43preglowyeah, mp3s sound smooth as cream now
16:18:19amiconnpreglow: commit!
16:18:51preglowamiconn: did it five minutes ago
16:19:08 Join postglow [0] (
16:19:24*preglow greets his evil twin
16:20:10 Quit postglow (Client Quit)
16:20:50preglowsomeone with decent ears should check out my improvements
16:21:38preglowthe difference was very noticable for some tracks
16:28:15 Join asdsd____ [0] (
16:28:51 Part asdsd____
16:37:58HClwhat was it?
16:38:30ashridahhmm. does seem to be a bit fuller here, but that could be the earphones i have on
16:38:32*HCl prods preglow
16:38:41preglowread cvs description
16:39:49t0masis %xnxxxyyy as a syntax for placing images on the WPS to difficult?
16:39:58t0masn = image number
16:40:02t0masxxx = x coordinate
16:40:10t0masyyy = y coordinate
16:40:29t0masso %x0010010 would place image 0 at 10, 10
16:40:44t0masand %x1000000 would place image 1 at 0, 0
16:41:30preglowashridah: i didn't notice it usually, bet the background noise sounds completely different now
16:41:45t0maspreglow? what do you think? :)
16:42:10HClt0mas: can't you use some html style stuff?
16:42:25t0maseverything is in the %XX syntax
16:42:39HCli'd more vote for
16:42:43t0masbut the xxx and yyy can ofcourse be done better
16:42:51HCl :/
16:42:56HClits not very readable either way :/
16:43:03HClat all :/
16:43:10t0mas%x|n|x|y would be nice
16:43:11preglowit's just a tiny wps file
16:43:16preglowreadability doesn't matter very much
16:43:32t0maswell... I tried doing that last one... but it took a lot more resources to do...
16:44:00t0masas reading 3 bytes, atoi() em and add 3 to the pointer is much easier than strtok_r() and splitting to some buffer...
16:44:45HClokay, well.
16:44:51HClas long as the wps format is very well documented
16:44:54HClit should be okay
16:47:08linuxstbHow do images interact with the rest of the WPS? Will text be displayed on top of the image? What about scrolling lines?
16:47:28t0maslinuxstb: scrolling lines is something Linus said he would fix
16:47:32t0masto make them use a margin
16:47:52t0masand the images will be painted over the text... but without clearing...
16:48:13t0mas(or should I make that a setting?)
16:48:23t0masso you can clear out text with images...
16:48:57 Nick Qt is now known as QT (as@area51.users.madwifi)
16:49:50linuxstbt0mas: I don't know.
16:50:25t0mashm... if loads of people want to overwrite text with images (seems weird to me...) then I'll make a setting somewhere
16:52:28 Quit ashridah ("Leaving")
17:00:09 Join webguest21 [0] (
17:03:45t0masthere it is :)
17:03:47preglowmy, mp2 playback works like a charm
17:03:47t0masfirst test
17:04:28thegeekpreglow : I agree
17:04:34thegeekyour patch really does improve sound quality
17:04:38thegeeka lot more details now
17:04:51*t0mas updates :)
17:04:55thegeekwas as if it was blurred before
17:04:57preglowshould be, the old multiplicator had mad roundoff errors
17:05:12thegeekthough I can't hear much difference using cheap headphones and no amp
17:05:17preglowthere is still a bit more precision to be gained, but that would make libmad a lot slower
17:05:36thegeekwith a cmoy and senn 595 it's quite noticeable
17:05:51preglowmuch slower
17:06:02preglowso don't expect that to happen
17:06:07preglowi think this is more than precise enough anyway
17:07:24preglowi do so want to make libmusepack realtime
17:07:31preglowbut i have to do other stuff now
17:10:45t0masthegeek: do you have compiler/tools installed?
17:12:57linuxstbI'm not sure what the correct fix is, but I don't think the call to pcm_crossfade_start() in playback.c should happen if crossfade is disabled. This seems to be the cause of my FLAC tracks missing the last few seconds.
17:14:51 Join Sucka [0] (
17:18:13 Quit bobTHC ("Smoke Weed Every Day !")
17:19:36[solid]linuxstb: sounds more than obviously logical :)
17:21:40***Saving seen data "./dancer.seen"
17:22:40linuxstbYep - the audio is going into the crossfade buffer, but never coming out...
17:22:53linuxstbBut I'll leave it to the people that wrote that to fix it.
17:22:55thegeekt0mas: yes
17:23:03thegeekI got it setup on my gentoo serverbox;)
17:23:48thegeekwant me to test the patch?
17:25:08preglowi occasionally get load errors for codecmpa.rock
17:25:14thegeekapplies just fine
17:26:46 Quit zezayer (Read error: 60 (Operation timed out))
17:27:40preglowcan't hear a single flaw in the sound now
17:28:06thegeekdoes optical out work atm?
17:28:12thegeekor is that way off
17:28:14preglowdon't believe so
17:28:38thegeekI think you should perhaps sometime make it possible to run that "superhigh" quality stuff
17:28:48preglowthere is no difference for rockbox
17:28:54preglowit only supports 16 bit quality
17:29:00preglowand it'll be a Major Bother to code it
17:29:07thegeekwould be kinda useful in a studio sitution
17:29:10thegeekwhen using optical out
17:29:18thegeekbut again
17:29:22preglowyou need to fetch extra precision bits out from the accumulator extensions for every accumulator fetch you do
17:29:32thegeekI have no idea what that means
17:29:33preglowand then recombine the result and bla bla bla and then you slit your throat
17:29:43thegeekbut I'll take your word for it when you say it would not be worth it;)
17:30:07preglowthere's no way in hell i'm doing it anyway, heh
17:30:34preglowit would require me being bored five miles out of my mind
17:30:38preglowand i'm never bored
17:30:40thegeekI like the bootloader usb mode
17:30:47preglowyeah, it's great
17:31:21preglowi might have sped up mp2 playback somewhat without knowing it
17:31:36preglowoh well
17:31:40preglowi'll vanish for a bit
17:31:48linuxstbCan you speed up my 48KHz mp2s?
17:31:54 Join webguest82 [0] (
17:32:41 Quit webguest82 (Client Quit)
17:33:07preglowlinuxstb: yeah, but not know
17:33:21preglowi didn't mean that kind of speeding up, btw ;)
17:33:26preglownot NOW
17:34:07preglowbut this rocks
17:34:16preglowmpa, ogg, flac working, and they all sound just great
17:34:41linuxstbYep. It was no effort at all :-)
17:37:09thegeekthank god Slasheri came along;)
17:37:30thegeekhe really is just what rockbox needed
17:38:27 Nick cheriff is now known as cheriff_AWAY (davem666@wagner.orchestra.cse.unsw.EDU.AU)
17:42:58thegeekt0mas: I'll see if I can test the patch a little
17:43:00 Quit webguest21 ("CGI:IRC (EOF)")
17:43:08thegeekI'll try to replace the "Track" part of my wps
17:43:11thegeekwith an image
17:45:07thegeekthere are sometimes high clicks/glitches when starting a new track
17:45:18 Join ghode [0] (
17:45:22thegeekusing usb while playing/stopped/paused is buggy
17:46:00ghodehey preglow , thanks for thaqt patch you submitted, lotr ost sounds even better now.
17:46:24thegeekand btw t0mas: is there no way to use grayscale in the wps?
17:46:31thegeekit would be nifty;)
17:46:34preglowghode: good to hear
17:47:36 Join ehntoo [0] (
17:48:27ghodeonly listening on some sony ex71s, but the difference is notiscable, now i just need my shure's ;/
17:51:25 Quit ghode ("CGI:IRC (EOF)")
17:52:04 Join DMJC-L [0] (
17:52:26 Quit DMJC-L (Client Quit)
17:52:39thegeekseems to work great t0mas:)
17:54:24 Quit DMJC ("Leaving")
17:54:31 Quit courtc (Read error: 110 (Connection timed out))
17:55:01 Join courtc [0] (
17:55:34t0masthegeek: no...
17:55:49t0masthegeek: for grayscale we need to upgrade the grayscale stuff... and the bmp loading code
17:56:34 Quit Patr3ck ()
17:58:06t0masBagder? or Zagor? does the mailinglist have some kind of spamfiltering?
17:58:22t0masor how does it react on two emails from 1 person... shortly after eachother?
17:59:47preglowt0mas: that rocks, man
18:00:11t0maspreglow: have you read the code?
18:00:34t0masit's a bit dirty...
18:03:30 Join Iriverowner [0] (
18:04:15 Quit Iriverowner (Client Quit)
18:04:41thegeekbtw, when loading a new wps
18:04:51t0masit shows the old one...
18:04:54t0massubline by subline
18:04:56t0masI noticed too
18:05:05thegeeknot too big a deal
18:09:47*t0mas goes to read a french book :X
18:10:41t0masstupid teacher...
18:10:50t0mashe's really thinking I can read it :P
18:11:21t0masbut while enjoying some music... it's less irritating to lookup 33% of the words :P
18:11:30MoosCamaroHi t0mas: pourquoi c'est quoi comme livre? :)
18:12:37MoosCamarowhat french book it is ?
18:14:04 Quit [solid] ("leaving")
18:15:18 Join dapureplaya [0] (
18:18:06 Join webguest24 [0] (
18:18:31webguest24Where is the Crossfade setting for iriver box?
18:20:31Zagorrockbox has no crossfade
18:20:55dapureplayageneral settings >> playback >> Crossfade
18:21:04dapureplayais that what ur asking for?
18:21:13dapureplayahi every1 BTW =P
18:21:25thegeekI can't get it to load more than one image?
18:21:26thegeekif I do
18:21:31thegeekthey just end up in the same place
18:21:43thegeekand I do use different coords
18:21:54thegeekis what I use
18:22:06thegeekthey all end up in the same place
18:22:49ehntoowhat in the world is wrong with my queen ogg's...
18:22:51Zagorman, is my head in the sand... :-p
18:22:55 Part dapureplaya
18:23:36 Quit webguest24 ("CGI:IRC (EOF)")
18:24:11 Join dapureplaya [0] (
18:27:23thegeekI could always do as you do in your example wps
18:27:32thegeekbut It would be useful with more than one image;)
18:32:08 Join tucoz [0] (
18:32:54tucozIs slasheri here? Just wanted to report a behaviour (not bug) with rockbox
18:33:17preglowSlasheri: woop woop
18:33:25preglowwhat is it, btw?
18:33:48tucozwhen buffering a somewhat large song, and I want to shut down, it doesn't shut down until the buffering is finished
18:33:54preglowi know
18:33:58preglowannoys the hell out of me
18:33:59tucozok, hehe, great
18:34:42tucozGreat work on libmad by the way. So smooth now.
18:34:48preglowi agree
18:34:54preglowit certainly does sound nicer
18:35:57tucozA lot more detail on metal-tracks
18:36:12tucozi.e. newer death metal, with good production
18:36:20preglowperhaps, i noticed it on a semi ambient track, actually
18:36:24preglowas some kind of jittery background noise
18:37:28tucozIt might be placebo, but I usually listen to that with some bass-boost and trebel-boost in iriver fw, but I think it sounds as it should now in rockbox
18:38:18 Join mbr_ [0] (
18:38:19 Quit mbr (Read error: 104 (Connection reset by peer))
18:39:11dapureplayawe dun have to reflash the bootloader for this improved sound do we?
18:39:27tucozJust download the latest build
18:39:30dapureplayak. well i'm d/ling the new build now then
18:39:38dapureplayadoes this apply for all codecs?
18:39:56tucozNo, it's only for mp3's
18:39:58Plugh_I heard about someone testing audio equipment on audiophiles
18:40:25dapureplayaah ok
18:40:28preglowdapureplaya: only for mp3, mp2, mp1
18:40:33Plugh_they turned up the volume a couple dB and the audiophiles thought it was equipment costing thousands more
18:40:54 Join merbanan [0] (
18:41:08ZagorPlugh_: yeah, that's a well-known fact. wma uses that effect in low bitrates.
18:41:11tucozLot's of psychology involed there. L
18:41:56Plugh_is that what they mean by psychoacoustic models? ;)
18:42:56 Join [solid] [0] (~solid@
18:44:37tucozbtw, I am in no way an audiophile. But I do notice the difference between flat and wide sound.
18:46:46tucozand I'm quite pleased with how the iriver sounds. Has always been. Hmm, there are some glitches on track changes now
18:46:56preglowthey'll be fixed
18:47:01preglowpretty much know why
18:47:04preglowi believe markun is working on it
18:47:23tucozLike small parts repeating?
18:47:54dapureplaya^^ yeah i'm getting that atm aswell.
18:48:11tucozAhh, not plops, 1/10'th of a second of sound repeating just before a track change
18:48:22tucozor something like that
18:48:30 Quit Bger (Read error: 104 (Connection reset by peer))
18:48:44 Join zezayer [0] (~chatzilla@
18:49:02 Join Bger [0] (~Bager@
18:51:08 Quit [solid] ("Lost terminal")
18:51:19tucozThis is not the uda, this has to do with the buffered pcm data playing in some uncontrolled way. Like short bursts of bufferd data beeing played
18:52:26tucozright before a track change. Well, I guess this will be taken care of.
18:52:39tucozgtg, bye
18:52:43 Part tucoz
18:53:41 Join [solid] [0] (~solid@
18:59:39 Quit thegeek (Read error: 54 (Connection reset by peer))
18:59:45 Join thegeek [0] (
19:00:01MoosCamarowhat' happens with wiki page from austriancoder
19:00:13MoosCamarohe put GUI proposal
19:00:48MoosCamarowe can't view topic
19:01:15Zagorwhich topic?
19:01:41MoosCamarohi Zagor: in the wiki page
19:01:48MoosCamaroGUI proposal
19:03:26MoosCamaro there'is not url link for GUI proposal of austriancoder
19:03:53 Join _aLF [0] (
19:04:29Zagor...and what is the url to that page?
19:05:25MoosCamarogreat question
19:05:42linuxstbDo you mean the GUIProposal topic?
19:05:52amiconnMoosCamaro: There is the wiki index page, and you can search for topics
19:06:22MoosCamarothanks guys
19:08:20MoosCamaroI hope for him he solve his player problem
19:12:14 Quit zezayer ("Chatzilla [SUSE 1.0.4-1.1/20050511]")
19:13:27preglowperhaps 'blend' vil du for norwegian as well
19:14:19preglowwhat the hell
19:14:38preglowi'm seriously starting to question the inner workings of my brain
19:15:23amiconnWhy does rockbox on iriver spin up the disk when I pause a track??
19:16:46linuxstbRockbox on iriver is doing lots of strange things at the moment...
19:17:36linuxstbNext/Prev navigation isn't working. The WPS screen is displaying information for completely different tracks.
19:18:05 Nick Sucka is now known as Sucka`away (
19:18:08amiconnCPU is still boosted after boot
19:18:27linuxstbI think both the crossfade changes, plus probably my recent metadata-reading changes are causing problems. And I think there were already bugs before those changes.
19:19:00Slasheriyep, it's very buggy at the moment :/ i am trying to fix those problems but it takes some time
19:19:28[solid]glad to know you'reworking on it already Slasheri :D
19:19:51Slasheri:) i made those bugs and trying to fix them now
19:20:14t0masthegeek: sounds like a bug
19:20:14t0maslet me check
19:20:20linuxstbSlasheri: Am I right to say that you are mainly concentrating on MP3 playback? Most of the code in playback.c seems to be mp3 related.
19:20:41linuxstbOr is it general code, but with "mp3" in the function/variable names?
19:20:49preglowrockbox is very mp3-centric in general
19:20:52preglowfor some reason :)
19:21:03t0masthegeek: they all endup at 0,0 or at the first place chosen?
19:21:04amiconnSmall wonder ;)
19:21:13linuxstbPoor little MP2 - always ignored...
19:21:17Slasherilinuxstb: That's true. When mp3 is fully working, then it's time to support other formats better. But the playback.c should be quite universal regardless of the file type
19:21:18thegeekt0mas : at the first place
19:21:30amiconnlinuxstb: mp2 is handled in rockbox for long
19:21:41amiconnIt's just not mentioned explicitly
19:21:43***Saving seen data "./dancer.seen"
19:21:46linuxstbamiconn: I know - but all the variables etc have mp3 in the names.
19:22:03t0masthegeek: weird :|
19:22:20Slasherilinuxstb: if there is mp3 in function name, then it's meant to support mp3 only (and that function should be moved to another file in future) music just randomly stopped for no reason. Is this stil a known bug or has it supposably been fixed already?
19:22:50t0masthegeek: I'll logf() test it...
19:22:51Shagnardapureplaya same thing happend to me this morning
19:23:16dapureplayawhat were u playing? MP3s?
19:23:24Slasheridapureplaya: currently the playback is full of bugs.. Maybe you would like to use an earlier build
19:23:26Shagnarme: yes
19:23:29linuxstbSlasheri: I think it would be nice to start that seperation asap - I'm very keen on improving the FLAC support, so am happy to duplicate your work, but for FLAC files.
19:24:11dapureplayaslasheri: aaaawww but the 'now playing' and crossfade is so kewl hehe.
19:24:14dapureplayai think i can live with it
19:24:32Slasherilinuxstb: Basically only the metadata and file seeking are codec specific
19:25:04amiconnThe gui lags severely while buffering :/
19:25:15Slasheriamiconn: with mp3s also?
19:25:18dapureplayataht's kinda normal
19:25:26Slasherimp3s should now yield really often
19:26:01amiconnLike, when I push up volume for some seconds, and then release the button, it keeps increasing the volume a while
19:26:05linuxstbSlasheri: How critical are the "elapsed" and "length" variables for your playback code? Do they need to be set accurately for FLAC and OGG?
19:26:07[solid]Slasheri: are you going to possibly implement replaygain globally too? the per-codec approach (as seen in winamp, xmms, etc etc) sucks pretty bad (especially in comparison to foobar which supports rg in the core and the codecs just give the values)
19:26:21Slasheriamiconn: you could try with lower file chunk size. But that may increase the buffering time
19:26:39Slasherilinuxstb: That's not very critical..
19:27:13 Join spiralout [0] (
19:27:21Slasheri[solid]: Hmm, maybe. Everything that is possible to implement can be implemented
19:28:04linuxstbThere is one easy-to-fix (I hope) bug - playback.c is calling the pcm_crossfade_start() function, even if cross-fading is disabled. I think this is causing Rockbox to lose the end of tracks.
19:28:23[solid]Slasheri: i'd be glad to help on that particular thing since it's very important to me (almost as important as gapless and ogg/flac support) but by finals are beginning next week and i would probably suck at coding that
19:28:24 Quit spiralout (Read error: 131 (Connection reset by peer))
19:28:40linuxstbI don't know if the best solution is not to call that function, or make that function return without doing anything (if crossfade is off).
19:28:44[solid]Slasheri: but, when the finals are over... :)
19:28:57Slasherilinuxstb: Ah, yep. I have just changed that if crossfading is not enabled, crosfade_start does nothing
19:29:39[solid]@#&%$ stupid computer... just made dvd's burn and not finish as toasts and the mouse irq got tragically lost in an oops dammit
19:29:41amiconnpreglow: I compared archos with your fixed libmad, using 256 kbps mp3 and real headphones, not my usual earphones
19:30:03linuxstb... and?
19:30:08amiconnI'd say libmad is a tiny bit more precise, but not really much
19:30:31[solid]anyone done comparisons to the iriver fw?
19:30:42amiconnHowever, another interesting observation is that archos can play louder, at least with these high impedance headphones
19:31:25amiconnI have iriver at 100% now, and to get a similar volume on Archos, I need 90%
19:32:02amiconn(600 Ohm headphones)
19:32:41 Quit [solid] ("leaving")
19:32:56dapureplayageez 600 Ohms....what ya got?
19:33:00thegeek600 ohm headphones should have an amp ;)
19:33:09amiconnAKG K141 Monitor
19:33:19dapureplayaAKG. That expllains it
19:33:32dapureplayai'm way out of that league.
19:33:49amiconnI should have gotten the Studio model...
19:33:51dapureplayaMy Senn PX100s are great. i'll stick with those :)
19:34:09dapureplayayeah sure they're not AKGs but they do the job hehe.
19:34:15amiconnMy usual earphones are Sennheiser MX400
19:34:32amiconnThere are better ones, I know
19:34:33dapureplayathey buds right? how do they go?
19:35:18 Quit tvelocity ("Leaving")
19:36:24preglowamiconn: well, i wouldn't expect libmad to be noticably better, really
19:37:13preglowbut i surely expect the mas did better than libmad before today
19:37:28amiconnDidn't test that ;)
19:38:32amiconnSlasheri: Both skipping and searching backwards don't work
19:38:35 Join Stryke` [0] (
19:40:37 Join [solid] [0] (~solid@
19:42:09dapureplayai think it's bedtime for me. cya
19:42:15 Part dapureplaya
19:42:31 Join spiralout [0] (
19:46:14 Join belgarath_the_so [0] (
19:46:38linuxstbSlasheri: Do you think we should keep track of elapsed time within playback.c, instead of the codecs? I think it would also be nice to switch to measuring time in samples, instead of ms.
19:47:57amiconnpreglow: Rockbox clearly sounds better than iriver fw :-) congrats!
19:48:07amiconnThe iriver firmware sounds 'rough'
19:48:24belgarath_the_sodoes anybody know how to switch from time remaining to time elapsed
19:48:33t0masSlasheri: bug report... when it's buffering... and I shut it down... it keeps waiting at the "shutting down..." prompt for the buffering to finish
19:48:44Slasherilinuxstb: It depends on wps. Currently wps wants elapsed time in ms
19:48:55Slasherit0mas: ok :)
19:49:14amiconnlinuxstb: time in samples would imho complicate things when the sample rate changes
19:49:47Slasherit0mas: fixed
19:49:54linuxstbamiconn: The samplerate doesn't change within the same track.
19:50:34amiconnNo it doesn't... but we want to display a time value.
19:52:53linuxstbIf we want gapless MP3 playback for files with lame headers, we need to know the total number of samples - and also how many samples we have played so far - so we can stop at the right place.
19:53:33t0masthegeek: how have you tested that both images get at the same position?
19:53:35t0masmine don't :X
19:53:48amiconnIirc we only need to know how many samples to cut from the last frame we only need the sample count within that frame
19:55:08preglowamiconn: can't say i've noticed that myself
19:55:57amiconnI just tried... listening to the very same track with all of (archos player) (archos recorder) (iriver, original fw) (iriver, rockbox) the same volume, flat eq
19:56:19amiconn..with the AKG
19:57:00amiconnLargish headphone, AKG K141 Monitor
19:57:10linuxstbIt just seems more natural to use samples - one sample is the basic unit of time when dealing with an audio file. But I can't think of any stronger arguments than that at the moment.
19:57:48amiconnWEll, the user doesn't want to know how many samples a track has, but how many seconds it will play
19:58:16 Join BossG [0] (
19:58:23linuxstbYou can convert samples to time, but not the other way around (depending on accuracy of the time).
19:58:46amiconnPlus, imho we should try to unify playback code as much as possible for both hw and sw codec boxes once the sw codec playback stabilises
19:59:02BossGhopefully somebody's awake to help me with a problem
19:59:05 Join Rori [0] (
19:59:26amiconn...and for our hardware codecs, it is impossible to get a sample count
19:59:26linuxstbThat's not an issue - we know how many samples there are in one MPEG audio frame.
19:59:34linuxstbIs it?
20:00:26amiconnWe can count how many frames we played (or rather, how many frames we transferred to the mas)
20:00:37amiconn...but not how many samples were played
20:01:27amiconnWe could just calculate an estimation from the number of samples in a frame,
20:01:32amiconnbut that number is different between mpeg1 and mpeg2, and also between layer2 and layer3
20:01:41preglowamiconn: well, goodie, can't say i hear any immediate difference between iriver and rockbox, but at least i'm satisfied with the sound libmad produces now
20:01:59Roriwhat? what? crossfade? did I read the cvs right?
20:02:04preglowRori: yes
20:02:25amiconnlinuxstb: In addition, there is usually no reason to analyse the bitsream with the cpu, so we don't do this
20:02:29BossGI have the "No FAT32 Partition!" on the Jukebox recorder 20, then shows the partitions that are on there, P:0 has 19000MB worth of data. Only problem is that holding F1 doesn't load up the original archos firmware that might fix it.
20:03:25amiconnBossG: You don't need the archos fw to fix that, it even can't
20:03:51 Join xen` [0] (
20:03:54amiconnYou need to hook it up to your pc and check the formatting / partitioning
20:04:11amiconnWhat does it say after T: for your P:0 ?
20:04:11BossGok, I was hoping I didn't need to do that...I have no choice now.
20:05:08BossGit says P:0 S:100003F T:C 19061 MB
20:05:11linuxstbamiconn: So how do you keep track of elapsed time on the MAS?
20:05:36amiconnBossG: Is it really S:100003F ? Not S:3F?
20:05:51amiconnlinuxstb: We simply count the tick from start of track
20:06:58BossGI'm using rockbox v 2.3
20:06:58 Join XavierGr [0] (
20:07:20amiconnBossG: If there are no other partitions, it seems your partition table is screwed, and you need to repartition and reformat
20:07:55amiconnTry connecting the box to the pc and reading the drive. Does that work?
20:08:00linuxstbamiconn: OK then. It's just that every software codec is counting samples, and then dividing the sample count by frequency before calling the set_elapsed_time function.
20:08:20 Join ac [0] (
20:08:26BossGI haven't tried that just yet...I will today. I'd hate to lose some of that music.
20:08:51accan somebody of the siteadmin fix my error in the wiki " GUIProposal"?
20:10:09XavierGrhi all
20:10:18CoCoLUSsaw the latest cvs commit, about the german translation
20:10:46linuxstbac: Why can't you fix it yourself? It's a Wiki :-)
20:10:56CoCoLUSbest to leave crossfading as "crossfading", "überblenden" is somwhat... not fitting
20:11:05RoriHmm FF/RW on Ogg broke on latest commit
20:11:19Roriprobably to do with the wps
20:11:42aclinuxstb: sure it is a wiki.. but it seems that it is unfixable
20:11:46[solid]did it ever work?
20:11:54[solid](ff/rw on vorbis)
20:12:24Roriwell it was working yesterday
20:12:30aclinuxstb: got it ;)
20:12:57RoriI think when linus added a rudimentary wps to ogg it broke ff/rw (guessing)
20:13:26Rorigoing back to last nights commit to check it folks
20:15:33preglowCoCoLUS: 'crossfading' isn't very german, now is it?
20:15:44 Join Cassandra [0] (
20:16:11Rorimaybe I goofed. Hmm. Try an even older build before wps
20:16:54CoCoLUSnot it isn't, but neither is firmware, and it's common
20:17:08CassandraBleeding Nora - this place got busy.
20:19:10Rorimy mistake
20:19:23RoriI must have imagined ff/rw in ogg ;)
20:19:26 Join niobos [0] (
20:19:42Roriso how does this crossfade work?
20:20:25preglowRori: just try it out and you'll hear it
20:20:38preglowCassandra: yes, yes it did
20:21:03preglowsomehow it coincided with rockbox for iriver getting playback support
20:21:09linuxstbRori: Don't blame Linus - it was me.
20:21:56RoriI enabled it in playback settings. now what? just play s few tracks?
20:22:38preglowjust play a track and wait for it to switch to another track
20:22:38 Quit belgarath_the_so ("CGI:IRC (EOF)")
20:23:44preglowRori: what did you expect it'd do? ;)
20:23:46acwhy is my topic in the wiki not shown as link?
20:24:05XavierGrack!! bug in the snake2 score writing. :x
20:24:07linuxstbac: Which page contains the link?
20:24:20Roriseems to work :)
20:24:26 Quit Ismo (Read error: 110 (Connection timed out))
20:24:29aclinuxstb: i have created a new site:
20:24:48acbut in the main site of the wiki my topic is no link :(
20:25:18linuxstbac: I think you need to manually add the link. Just typing the word GUIProposal should make it a link.
20:25:26XavierGrI fixed it. So much trouble for a bad character!
20:25:39XavierGrcan someone commit it to get along?
20:25:50XavierGrI can send him the .c file for his next commit
20:27:08linuxstbac: Your page appears in the Index -
20:27:24amiconnhi Cassandra
20:27:47aclinuxstb: yeah.. but not as link.. or i am wrong?
20:28:37linuxstbac: It appears as a link for me.
20:28:47linuxstbAre we looking at the same page?
20:29:26linuxstbNo, I was looking at WebIndex
20:30:09linuxstbAre you talking about the "10 most recent changes" table?
20:30:11Slasherit0mas: Hmm, i think i found a problem. When i select a new song while previous one is playing, i got these events: audio_play, audio_current_track, audio_current_track, audio_pause, audio_stop
20:30:14preglowac: oy, are you going to commit your uda changes soon or what?
20:30:19SlasheriWhy on earts do i get audio_stop?
20:30:37CassandraHi ami
20:31:00acpreglow: today or tomorrow
20:31:04Cassandrapreglow: What an odd coincidence. :)
20:32:09preglowac: goodie
20:32:22actime to gooo... see you later
20:32:25 Quit ac ("CGI:IRC 0.5.4 (2004/01/29)")
20:32:46XavierGrdoes anybody want to sent him the snake.c file to include in his next commitment. I found a bug in the hi-score saving routine, and just correct it?
20:32:53preglowThe rude design is done and i am writing code now. <- typo of the year
20:32:54CassandraI need to say that people need to make sure I have all the changes for 2.5 listed on the documentation page.
20:33:07CassandraOtherwise they might not make it into the manual.
20:33:13CassandraWhich would be bad.
20:35:28CassandraAlso, I am unlikely to put Ondio keybindings in the next manual release I'm afraid.
20:35:30t0masSlasheri: who's doing the audio_stop ?
20:35:31t0masand from where?
20:35:51t0masI guess it's your audio_next() code?
20:35:53Slasherit0mas: I am not sure where it comes from, probably from wps
20:36:03SlasheriNope, my code never calls it
20:36:05t0mashow do you select the next song?
20:36:21CassandraIt's looking like I'm going to need to batch convert from OpenOffice to LaTeX to do multiplatform properly, and I'd like to do the iRiver and Ondio platform specific stuff at the same time.
20:36:28Slasherinow i left the current song playing, then go to the file browser and selected new one
20:36:35t0masoh found it
20:36:43t0masif you exit the wps, it sends audio_stop()
20:36:51t0masbut that was there before I did something to it...
20:36:57t0masso I guess it's needed on archos
20:37:32SlasheriSo that's wps's behaviour? :/
20:37:43SlasheriBut who do it send audio_stop _after_ audio_play? :/
20:37:44preglowCassandra: good, please consider not using the default latex font
20:37:53preglowCassandra: on the grounds of it being very ugly
20:37:53t0masLinus did it in version 1.234
20:38:09CassandraI hate the default LaTeX font too. Don't worry.
20:38:17t0masoh, that's just a rename
20:38:35SlasheriHmm, wait a second.. Maybe there is a problem in playback.c also :)
20:38:44 Join tvelocity [0] (
20:39:14amiconnSlasheri: If wps would be acting wrong here, it wouldn't work correctly on archos as well, but it does.
20:39:22t0maszagor in 1.178: mpeg_stop();
20:39:23 Join jipi [0] (
20:39:24XavierGrhi tvelocity
20:39:33tvelocityhey mate
20:40:35t0masSlasheri: it has always been there...
20:40:39t0masthe stop call when you exit wps
20:41:22Slasherit0mas: Hmm, if i select a new song will i then "exit" the wps?
20:41:37t0masdon't know
20:42:14t0masbut that's the only audio_stop() call...
20:43:46CassandraWhere does put the songdb?
20:43:58preglowin the you're standing, afaik
20:44:06preglowin the DIR you're standing
20:44:23CassandraAnd it's supposed to be in .rockbox, right?
20:44:48CassandraMy last run didn't seem to actually write a db file. Weird.
20:45:00preglowworks dandy here
20:45:04 Quit tvelocity (Read error: 54 (Connection reset by peer))
20:45:25CassandraOn iriver?
20:46:55 Join tvelocity [0] (
20:47:19CassandraWell, it's rerunning now. We'll see what happens.
20:47:42CassandraTakes a while, doesn't it?
20:52:12t0masdepends on the pc and platform
20:52:19t0mason cygwin/windows it takes a long while...
20:52:25t0mason linux and a fast pc it doesn't...
20:52:36t0masand ofcourse it depends on how much songs/dirs you have
20:52:41preglowruns fairly fast
20:52:46CassandraI'm running it on Linux. The machine may have USB 1.1 only, I don't remember.
20:52:50preglowcouple of minutes for nearly full h120 with mostly oggs
20:53:33amiconnSlasheri: My iriver didn't shut down at timeout after playing some mp3s
20:53:59CassandraI'm getting a "Not a valid ID3 database" error.
20:54:44CassandraThat's with today's daily build and from CVS.
20:55:46t0masSlasheri: do you yield more in buffer filling code on your local copy?
20:56:33linuxstbThis is probably a stupid question. I don't have any tags in any of my files (I just use the "Album/Artist/NN-Track.ext" directory/filename conventions) - would the ID3 database be of any use to me?
20:56:54linuxstb(I meant to say Artist/Album/Track)
20:56:57CassandraAh, crap. I think I forgot to unmount the disk.
20:57:04CassandraIt's now gone to size 0
20:57:41Slasherit0mas: Not yet but maybe i will lower the file chunk size slightly
20:57:49amiconnDoesn't linux allow to set parameters for portable disk devices?
20:57:50CassandraDamn you Linux and your lack of plug and pray USB.
20:58:19CassandraPossibly. I'm mounting it direct from a root prompt because I can't be bothered to arse about.
20:58:20amiconnWindows does. You can set it to 'optimize for quick removal'; then it won't use any write caching
20:58:40preglowamiconn: windows doesn't by default
20:59:00preglowwith linux it probably depends on the distros hotplug support
20:59:11amiconnI always simply pull the usb plug, and didn't get any problems if this parameter was set correctly
20:59:13CassandraI'm using Ubuntu 4.10
20:59:31preglowi always simply pull the usb plug and have no troubles without having set anything
20:59:48CassandraWeirdly it attaches my USB disk as /dev/sda1
20:59:56preglowCassandra: i'm using hoary, and just pulling the cable here is not very wise
21:00:25Cassandrapreglow: Yeah. I can see it's not wise here either.
21:00:46CassandraGoing to upgrade to hoary at some point, but I'm waiting for my CDs to arrive.
21:01:06 Quit XavierGr ()
21:01:49 Quit thegeek (Read error: 60 (Operation timed out))
21:04:37*t0mas has a script for that...
21:04:43t0masmount /dev/sda1 /mnt/iriver
21:05:04CassandraEerily identical to what I typed.
21:05:05t0masunzip -p /home/tomas/dev/rockbox-devel/build/ -d /mnt/iriver/
21:05:10t0masumount /mnt/iriver
21:05:14t0masecho "Install done."
21:05:28t0maslike that :)
21:05:43CassandraThat wouldn't build your id3db though.
21:06:03t0masno, but I'm not intrested in that
21:06:13t0masI'm building rockbox about 10 times/hour
21:06:17t0masfor bugfixing and testing
21:06:33CassandraWoot. It's working.
21:06:48CassandraRight. I'm off to do some shopping.
21:07:31 Quit Cassandra ("If I were actually witty, this quitline would be funny.")
21:08:49amiconnBah. m68k is sometimes soo limited compared to sh1, despite having more instructions.
21:09:01preglowwhat you trying to do?
21:09:12amiconnAt least lea and add/sub to address registers don't affect the flags
21:09:26preglownot in 68k, no
21:09:42 Join thegeek [0] (
21:09:45*preglow wants arm asm
21:10:00amiconnI'm trying to save registers :/ I want to avoid using the stack as much as possible
21:10:16amiconn4 scratch registers is tight...
21:11:17preglowi also don't get why parameters aren't passed in register on 68k
21:11:21amiconnThe sh convention has 8 scratch registers, plus the first 4 parameters are passed in registers
21:11:27preglowthey could at least have used the scratch regs for parameters
21:12:00amiconnr0..r7 are scratch, with r4..r7 being used for parameters
21:13:49 Join izzy [0] (
21:15:19 Nick izzy is now known as Ismo (
21:16:02preglowi don't think m68k is half-bad
21:16:08preglowor at least the coldfire variant, haven't tried full m68k
21:16:32preglowat least not compared to x86 asm, which is a flaming nightmare
21:16:51preglowapart from that i've only tried 6502 and avr asm
21:17:04amiconnfull m68k has bit rotation, which coldfire lacks
21:17:17Slasherit0mas: I think i have fixed the bugs :)
21:17:19preglowyeah, 'cause it's never used in c
21:17:20amiconnI once tried to code a bit in x86, but failed
21:17:23 Join DomZ [0] (
21:17:23preglowthey optimised the instruction set for c
21:17:34preglowthat's why exg is missing as well
21:17:50amiconnEven earlier I coded a bit in Z80 asm
21:18:28preglowSlasheri: which bugs? commit coming up? :P
21:18:46Slasheripreglow: many bugs, i don't even know how many :)
21:18:56preglowplayback _is_ pretty buggy now
21:19:03Slasheriindeed :D
21:19:39[solid]Slasheri: did you by any chance fix the weird skipping during playback?
21:19:47Slasheriof course this wont fix everything but the most annoying i hope
21:20:05[solid](which happened only with crossfade off)
21:20:08t0masSlasheri: we're becomming the new bug create/fix team here ;)
21:20:10Slasheri[solid]: Hmm, what was that exactly?
21:20:14linuxstbSlasheri: I don't want to delay you committing anything, but if I play one directory of files, and then stop playback, and change to play another directory of files, then I am sometimes getting the WPS for a different track. I mostly have FLAC files, so it could be my bug...
21:20:18Slasherit0mas :D
21:20:47[solid]Slasheri: playback was just... well... uhm skipping samples randomly all over the place
21:20:52Slasherilinuxstb: Hmm.. You should check if the bug is still there after i commit
21:20:53*t0mas goes to smash some of his bugs :)
21:21:02linuxstbSlasheri: OK.
21:21:08Slasheri[solid]: interesting, i am not sure about that
21:21:17[solid]Slasheri: i'll wait for the build and test :)
21:21:25Slasherigreat :)
21:21:28Stryke`no way to center text in a wps?
21:21:47***Saving seen data "./dancer.seen"
21:22:21t0masfuck... I feel like a Microsoft programmer...
21:22:27t0masfixing something a fix broke...
21:22:43t0masStryke`: not yet
21:22:45t0masit's on my wish list
21:22:51t0masand my vacation starts in 1 week
21:22:57t0masso it's comming ;)
21:22:58*amiconn goofed severely
21:23:03Stryke`yeah, i'm designing my own wps and it would look nice, thanks
21:23:07thegeekdoes images work with ?'s in the wps?
21:23:21preglowamiconn: asm related? ;)
21:23:29t0masyou can't add a scrolling or conditional image
21:23:35t0masbecause an image is not a tag
21:23:37amiconnI wondered why my experimental memset crashed
21:23:39 Quit jipi ("Leaving")
21:23:50thegeekthat is desired.
21:23:52amiconnTurned out I reserved 12 bytes on the stack for 6 registers (!)
21:24:05preglowi've managed that as well
21:24:09t0mashm... should've been 24?
21:24:14t0masor am I misunderstanding?
21:24:17amiconnt0mas: yup
21:24:19preglowi've also managed not updating the stack pointer after stowing away some registers
21:24:27 Quit Shagnar (Remote closed the connection)
21:24:45amiconnmovem is nice, isn't it? ;)
21:25:02preglowi wonder why the hell they remover increment and decrement addressing for mvoem
21:25:16amiconnThat would be useful indeed
21:25:20preglow68k has
21:25:24preglowthey just removed it for coldfire
21:25:39amiconnHmm. My test shows 2 things
21:26:02amiconn(1) Doing double-line movems doesn't increase speed when aligned properly
21:26:21amiconn(2) movem uses burst mode as soon as a whole line is covered
21:26:38preglowi thought movem always used bursts
21:26:44amiconnNo it doesn't
21:26:57preglowso i need to transfer four registers before it does a burst?
21:27:22amiconnNot only 4 registers, but the address must be 'line' aligned as well
21:27:33amiconni.e. addr % 16 == 0
21:27:50amiconnI'll utilize that in memset()
21:28:23amiconnIt will be much more difficult to utilize it for both read and write in memcpy()
21:28:29preglowamiconn: well, what performance differences are we talking about? lets say i do a four register movem
21:28:57amiconnIf unaligned, it's just as fast as doing 4 consecutive move.l
21:29:02preglowamiconn: well, for memcpy you'll probably have to make do with just line aligning one of the arrays
21:29:15preglowamiconn: well, a bit faster, since no instructions need to be fetched
21:29:38amiconnIt's exactly the same speed, at least when working in sdram, and the code is cached
21:29:48amiconnI tested it
21:30:18preglowbut ok
21:30:20amiconnMy memset() test uses a number of cases which all add up to moving a total of 8000000 bytes
21:30:30amiconn(except the null test of course)
21:30:32preglowtransfering more than four registers will guarantee a burst
21:30:41amiconnThe largest blocks is 2 x 4000000 bytes
21:31:01amiconnTransferring more than 4 registers still won't guarantee a burst
21:31:16t0mascan I write logf output to a file?
21:31:23t0masor browse up the log some other way?
21:31:45amiconnA burst is guaranteed only if you move at least 7 registers, and the address is long aligned
21:32:05amiconnWith 7 registers, 4 are moved in a burst
21:32:44amiconnA burst is more than twice as fast as non-burst accesses
21:32:45preglowbah, of course seven
21:34:34t0masamiconn: then when you're moving 3 or 4 registers, can't you dummy move 3 to 4 others?
21:34:51t0masas if it's more than 2x faster... that would be faster than just moving 4
21:35:01thegeekbtw t0mas
21:35:10thegeekwas that multiple image stuff a bug?
21:35:18t0masI'm fixing it
21:35:56 Quit ehntoo (Read error: 110 (Connection timed out))
21:37:16thegeekbug found
21:37:28thegeeklet me try to write a good description
21:37:44thegeekplay a file
21:37:50thegeekpress the joystick
21:37:52thegeekselect a new file
21:37:53thegeekplay it
21:38:03thegeekthe play/pause icon remains paused
21:38:33thegeekwhile music plays
21:38:40thegeekif you press play the icon changes
21:38:47thegeekwith no other ill effects
21:39:18Stryke`another nice thing for wps would be something that just returned file's extension
21:39:30Slasherithegeek: thanks :)
21:40:01thegeekindeed Stryke`
21:40:13linuxstbStryke`: Do you want the extension or the actual type of the file?
21:40:31Stryke`actual type, but extension is usually just as well
21:40:55thegeekt0mas: we neeed image conditionals;)
21:41:08thegeekso we can have changing bitmaps for different type of files
21:41:46Stryke`i'd love a system like foobar's TAGZ
21:41:55linuxstbRockbox knows the type of the file (but at the moment it is guessed from the file extension) - so it makes sense to make the actual filetype available to the WPS.
21:43:28Stryke`im sure the wps will expand now that we have such a nice lcd available
21:43:33thegeekperhaps just implement it so that it uses whatever rockbox determines the file to be
21:43:56thegeekthen if actual filetype detection is implemented later it won't be a problem
21:47:20t0masthegeek: ghehe... would be a nice idea
21:47:29t0masI'll take a look at it... after everything is working
21:53:09t0maswoohoo :)
21:53:13t0masgta san andreas is ready :D
21:53:21t0masto bad it's on a remote server...
21:53:41*t0mas takes his iriver for a walk... to get his gta images :D
21:54:00Slasheriuuh, there was a huge spider in my room :/
21:54:07t0maskilled it?
21:54:30Slasherinope, i drop it to a box and throw outside :)
21:55:00preglowi frigging hate spiders
21:55:01 Join ehntoo [0] (
21:55:02Slasheriit was quite difficult to get it to the box :D
21:55:03[solid]Slasheri: i think the skipping problem is solved with your latest commit
21:55:07Slasherime too :>
21:55:19Slasheri[solid]: ah, that's great :)
21:56:10 Quit Naked (Read error: 145 (Connection timed out))
21:57:10 Quit Zagor (Remote closed the connection)
21:59:28amiconnSlasheri: Searching & skipping backwards still doesn't work correctly :(
22:00:54 Join ep0ch [0] (
22:01:24Slasherithegeek: i fixed the pause issue :)
22:01:40Slasheriamiconn: Hmm, backward seeking is still buggy indeed
22:01:46t0mashigh loads
22:01:47t0masroot@athlon rockbox-devel # uptime
22:01:47t0mas 22:01:55 up 3 days, 12:36, 5 users, load average: 8.57, 8.47, 8.30
22:01:56t0mason my workstation
22:02:11Slasherit0mas: pretty high loads
22:02:21 Join yngwi [0] (
22:02:34 Nick Sucka`away is now known as Sucka (
22:02:50Slasheribtw, now volume does not lag with vorbis at all :)
22:03:04preglowit's much stabler now
22:03:07amiconnHmm. Now backwards does work. Strange.
22:03:07preglowby far
22:03:23[solid]Slasheri: but the wps is a little off
22:03:43t0masSlasheri: yeah... 2731 processes according to top :X
22:03:43amiconnVolume still lagging like mad with mp3
22:03:53[solid]Slasheri: it displays the next track's data before the previous one ends
22:03:57t0masah found it
22:04:03t0masfucking hotplag crashing again
22:04:05Slasheri[solid]: Ah, that will be fixed later :)
22:04:21t0masit doesn't like plugging iriver in and out every few minutes
22:04:32[solid]Slasheri: still working flawless now :)
22:05:09 Join Tangleding [0] (
22:05:38Slasherihey :)
22:05:40TangledingIv'e updated the fornoobs section but only in french for now
22:05:49TangledingHi slasheri
22:06:12Tangledingthnaks again fr your very good work on audio stuf
22:06:22Slasherinp :D
22:06:56[solid]and the ui is *much* more responsive
22:07:19elinenbegood idea from the misticriver forum: Has there been any consideration about letting the .wps files display the file type (Ogg vs. MP3)?
22:07:56t0masI think I can add that
22:08:01t0masif it's not already there
22:08:26TangledingI've to go
22:08:31TangledingBobTHC not here?
22:08:37Tangledingseems not
22:08:54Tangledinghis email seems to be dead
22:09:00Tangleding(delivery failure)
22:09:09 Join bipak_ [0] (
22:09:17t0maselinenbe: extension is no problem
22:09:36[solid]elinenbe: ogg vs mp3 vs FLAC btw:P
22:11:27Stryke`a string compare would be nice, too, i like to show Self-Titled if album == artist
22:11:51t0maselinenbe: I have %fe here
22:11:53t0masfor extension
22:12:07t0masbut I can't commit it.... as the files has many changes
22:12:15amiconnSlasheri:I did watch the cpu boost while playing music now
22:12:30amiconnIt does indeed boost and unboost periodically.
22:12:30Slasheriamiconn: did it work?
22:12:49amiconnI get 16 seconds boosted vs. 11 seconds unboosted with 256 kbps mp3
22:13:03amiconnHowever, boost is still active after boot
22:13:14Slasheriyep. That depends completely how fast the codec is
22:13:45Slasherivorbis should give higher not boosted times
22:14:22Slasheriah that problem.. I will fix it now
22:14:50elinenbealso, another bug frmo the fourm
22:14:54elinenbeIt looks like Rockbox has a problem identifying .OGG files with an upper case extension, as .ogg files work fine.
22:15:48linuxstbelinenbe: Yes, that would happen.
22:16:13amiconnlinuxstb: No strcasecmp used?
22:16:18linuxstbelinenbe: Do you know what extension FLAC files inside OGG containers are normally given? Is there a convention? What about Speex?
22:16:32t0maselinenbe: do you have build tools?
22:16:49Slasherielinenbe: Oh, i have broke that.. fixing too :)
22:16:51amiconnpreglow: I'll try mp2 and mp1 now
22:16:52linuxstbamiconn: It doesn't appear to be - in playback.c
22:17:19t0maselinenbe: then checkout the mailinglist... and test my patch idea... for extension..
22:17:24t0masscream if it works or not
22:18:31[solid]linuxstb: the last bleeding edge seems not to skip no more
22:18:41yngwiHi guys, i dont want to interrupt, but hav a question concerning shuffle..
22:18:53*t0mas is going to have a fight with his pc about hotplug... and now 3173 running processes
22:19:01 Join Querty [0] (
22:19:10linuxstb[solid]: Yep, the latest changes have been been a huge improvement - thanks Slasheri.
22:19:12t0mas22:19:20 up 3 days, 12:53, 3 users, load average: 9.59, 8.95, 8.23
22:19:16amiconnHmm, the browser doesn't show .mp1 as music files
22:19:24t0mastime to reboot linux :P
22:19:47yngwiis it supposed to work only with playlists?
22:20:36Plugh_reboot linux?
22:20:46Plugh_ 1:20pm up 63 days 0:58, 4 users, load average: 1.00, 1.05, 1.10
22:20:52linuxstbamiconn: mp1 will need adding to tree.c I think playback.c will also need changing to use libmad for that filetype
22:21:04Plugh_just kill the runaway processes
22:21:33Plugh_use top to find the offenders
22:21:56amiconnSlasheri: 384 kbps mp2: ~20 s boosted, ~10 s unboosted
22:22:07elinenbeyeah, i have the development chain installed...
22:22:33elinenbei am just a simple programmer though... most of rockbox is above my head −− well more than i want to do.
22:22:46RoriSlasheri have you figured out -nogap in Lame MP3 yet? :)
22:22:50Slasherifixed cpu boost and case sensitivity
22:23:06SlasheriRori: I don't have to to look at it yet
22:23:13Slasheri+had time
22:23:20RoriOK no prob
22:23:28Roriwhenever you get round to it :)
22:23:30preglowlinuxstb: flac and speex in ogg containers are .ogg
22:23:36preglowlinuxstb: like it should be
22:23:44preglowlinuxstb: relying on file extensions is evil
22:24:26preglowamiconn: mp2 is really expensive yet
22:24:38preglowmy libmad fix might have made it worse
22:24:48amiconnAbout the same with 448 kbps mp1
22:25:26amiconnlinuxstb: With layer 1, the bitrate display is wrong
22:25:39 Quit t0mas ("brb, rebooting without hotplug shit")
22:25:46preglowwith mp2 as well
22:25:54preglowno, forget me
22:25:54amiconnIt displays 176 kbps for a real value of 448 kbps
22:25:55preglowit's not
22:26:05amiconnlayer 2 is working correctly
22:26:12 Quit Tangleding ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
22:26:43Roricurrently saving commits as rockbox1 rockobox2 etc etc
22:26:54Roriso I can always go back
22:26:54 Quit bipak (Success)
22:26:56preglow192 kbps performs at the level of 320kbps mp3
22:26:58preglowthat's really, really bae
22:27:19preglowi can't bloody imagine why it performs so badly
22:27:56 Join t0mas [0] (
22:29:14t0masback :)
22:31:17 Nick yngwi is now known as yngwi_away (
22:32:15*amiconn still has no idea how to solve the coldfire scratch register dilemma
22:33:00 Join tucoz [0] (
22:33:43tucozhi, is there function to get the height of a font in the plugin api? haven't found one.
22:36:08tucoznevermind, will try something else.
22:36:25*Rori laughs at the chipmunks reading an audio book
22:36:49RoriThere are chipmunks in my iRiver dammit! :)
22:37:04amiconntucoz: lcd_getstringsize() (main lcd)
22:39:22tucozamiconn: thanks
22:39:49 Quit Stryke` (Read error: 60 (Operation timed out))
22:40:23[solid]Slasheri: it seems i didn't do enuff testing... it does this skipping thingie sometimes still
22:41:52Slasheri[solid]: ok, i will try to find that problem too :)
22:41:57tucozamiconn: doesn't that just return the width of a string?
22:42:18[solid]however it did it only on two tracks and i've been playing with it for a long time now already
22:42:21amiconnAs return value, yes. But there are 2 int pointers, w and h
22:42:44tucozamiconn: yes, I suck. Have to read up the pointers
22:42:52[solid]w00t no shit, shuffle just played the same song twice:)
22:42:55tucoz*on the
22:43:37amiconntucoz: If you need the height only, using rb->lcd_getstringsize("a", NULL, &h); is sufficient
22:43:50tucozoh, cool
22:43:57 Quit ep0ch (" HydraIRC -> <- 100,000+ downloads can't be wrong")
22:44:17amiconnI.e. pointers that are NULL aren't touched
22:44:39tucozok, will keep that in mind. Thanks
22:44:40 Quit cheriff_AWAY (Read error: 60 (Operation timed out))
22:44:49amiconnSlasheri: Hmm. My iriver just played the last track of the playlist twice :(
22:45:23Slasheriamiconn: I know that problem too but i have no idea what causes that :/
22:45:51Roriwhat causes low bitrate encodes to play too fast?
22:46:13Slasheri[solid]: if you can, please disable crossfade and test again
22:46:40[solid]Slasheri: skipping happens exclusively when crossfade is off..
22:46:53Slasheriah, ok :/
22:47:04[solid]Slasheri: and it's much much much more rare since the latest commit
22:47:05Slasherithen crossfade can't cause it
22:47:41Slasherii will test it tomorrow with small buffer sizes
22:47:58Slasherimaybe i can find the actual bug
22:48:38[solid]linuxstb experienced the same thing, so he might also of some feedback help ^^
22:48:58Slasherigreat :)
22:50:02 Quit Sucka ("a bird in the bush is worth two in your house")
22:52:06t0masyou did something with a tft with dead pixels?
22:52:12preglowthey're still here
22:52:15t0masah ok
22:52:19preglowi'm just ignoring them, heh
22:52:26t0maswhat were you talking about then?
22:52:52preglownot much, i was wondering how i could get rid of them
22:52:58preglowthen i found out and stopped thinking about it
22:53:05 Join Naked [0] (
22:53:07t0maswhat did you find out?
22:53:16preglownothing safe
22:53:24preglowyou can just massage them and sometimes have them vanish
22:53:28[solid]at 10000B78
22:53:30preglowbut you can also risk destroying more pixels
22:53:44[solid]happened when... i pressed stop.
22:57:08 Quit niobos ("off to bed")
22:57:29Slasheri| .icode 0x10000bd8 0x82 /export/home/miipekk/cvs/rockbox/iriver/librockbox.a(pcm_playback.o)
22:57:32Slasheri| 0x10000bd8 DMA0
22:57:36SlasheriHmm, i think it happened there
22:57:53Slasherisomething is wrong.. i look more tomorrow
22:59:16preglowdo that
22:59:22preglowgreat work, slasheri, it's tons more stable now
22:59:48 Quit Harpy (Read error: 60 (Operation timed out))
23:01:04Slasheri[solid]: btw, what file type you was playing? ogg/mp3/flac?
23:01:14crwl/home/crawlie/rockbox/rockbox/build/apps/build.lang:2770:missing quotes for LANG_MDB_CENTER
23:01:17crwlmake[1]: *** [/home/crawlie/rockbox/rockbox/build/apps/lang.o] Error 1
23:01:21[solid]Slasheri: ogg
23:01:23crwlhum, cvs version doesn't seem to compile now
23:01:33linuxstb amiconn: I think the MPEG parsing code is just taken from "old Rockbox" - and I don't think the MAS supports MP1?
23:02:01linuxstbSo bugs with MP1 are not surprising.
23:02:30 Join Stryke` [0] (
23:02:31Bagdercrwl: what language have you selected?
23:03:33linuxstbSlasheri: I can't really help much with the OGG skipping problem - it appeared really badly in a couple of tracks, then next time I tried to play them it was fine. I haven't been able to replicate it since.
23:04:13SlasheriHmm, so it's random.. Maybe it has something to do with buffer refilling?
23:04:27 Quit merbanan ("Gone fisting...")
23:04:28crwlBagder, english
23:04:39QuertyIt's almost like random bits of buffer are played
23:04:40BagderI don't get that problem
23:04:46crwlmaybe i haven't
23:05:00QuertyBut I hardly have it anymore
23:05:07crwlyes, english's selecter
23:05:38tucozI have i minor fix for the menu in solitaire, which allows for user selected fonts. Which didn't work for any platform
23:05:39crwlah, the fix was easy anyway
23:06:06 Quit Naked ("leaving")
23:06:07tucozit's just two lines, so I do not think a patch is necissary
23:06:15Slasheri[solid]: How long the ogg songs are?
23:06:34linuxstbMy problem wasn't random sound playing - it was definitely skipping. There was no silence - just missing audio.
23:07:01linuxstbBut as I say, I can't replicate it.
23:07:04tucozSomeone with cvs-access interested in this?
23:07:54*Bagder ehums
23:08:07Bagdertucoz: tell us
23:08:34[solid]Slasheri: mostly between 2 and 7 minutes
23:08:42 Quit tucoz ("CGI:IRC")
23:08:47 Join tucoz [0] (
23:08:50SlasheriHmm, ok..
23:08:56tucozhehe, wrong paste
23:09:03QuertyAlso weird, I've never seen the CPU boost counter go down to 0 for .ogg
23:09:21SlasheriWith very long vorbis songs there can be memory overflow to the file buffer
23:09:21Quertyit always stays at 1
23:09:30[solid]oh my, as much i dislike the chicago font it really makes a nice wps
23:09:54SlasheriQuerty: Hmm, weird..
23:10:07Quertydo you see this too?
23:10:17crwlis this known that FLAC and Ogg files don't seem to initialize sound
23:10:24crwlone first needs to play an mp3 file
23:10:31[solid]crwl: untrue
23:10:44crwl[solid], i disagree :)
23:10:49tucozBagder, I just getstringsize("A", NULL, &fh), and changed the 9's in rb->lcd_putsxy and rb->lcd_invertrect to be fh
23:11:03tucozin solitaire_menu
23:11:10tucozon line 489
23:11:26crwl[solid], odd. after restart, everything worked
23:11:36[solid]crwl: :)
23:11:46crwlmy first start was from bootloader usb mode though (i upgraded rockbox there), if it matters
23:11:51[solid]crwl: i have only a few mp3's and tons of ogg/flac ;)
23:12:03*Bagder never played Rockbox solitaire
23:12:10crwl[solid], yep, my ogg/mp3 ratio is also quite high :)
23:12:38tucozBagder: me neither ;)
23:13:22tucozBut still, it's a part of rockbox
23:15:04tucozBagder: you should also remove the 1 from 17-1 in rb->lcd_invertrect
23:17:26BagderI made it use the fh +1 instead of 9
23:17:46Bagdermatches the old behaviour better
23:18:09Bagdertell me your name and I'll credit you for this
23:18:38tucozI did something earlier on, so no need for that :D
23:19:00Bagderfor the commit message then
23:19:05tucozok, Martin Arver
23:19:55 Join Patr3ck [0] (
23:21:02 Part tucoz
23:21:50***Saving seen data "./dancer.seen"
23:23:08crwlshouldn't rockbox cache the current directory contents (like the iriver fw supposedly does) to speed up the time to enter the filebrowser?
23:23:26thegeekI thought about that earlier today too
23:23:28thegeekforgot it
23:24:04Bagderthe current code is optimized for song buffer memory
23:24:36t0mashuh? doesn't it do that?
23:24:47thegeekI often have to wait for the hd
23:24:49thegeekwhen browsing
23:25:09t0masbut the first dir you open is quick...
23:25:14t0masso I guessed that was cached
23:25:25thegeekI just noticed it was way slower than iriver fw
23:26:53 Join Naked [0] (
23:27:22amiconnRockbox caches the current dir only.
23:27:42amiconnFor quick browsing, don't set the hd timeout too low
23:28:07[solid]Slasheri: another IllInstr, same adress, this time while playing a mp3 file and pressing rew twice
23:29:32t0masgood night :)
23:29:36t0mastime to get to bed :)
23:29:40 Quit t0mas ("c u tomorrow")
23:29:45Slasheri[solid]: thx, fixing tomorrow :)
23:29:48Slasherinight )
23:31:43Quertynight Slasheri, and thanks for the hard work!
23:33:07 Quit Querty ("Leaving")
23:35:31 Join n[o]bby [0] (
23:35:38 Nick n[o]bby is now known as nobby (
23:39:35 Join Zoom2 [0] (
23:40:44 Join bill20r3 [0] (
23:45:00spiraloutis it possible to add the centerscrolling to the settingsmenue as an option ?
23:46:03Bagdernot without additional coding
23:47:01 Join _DangerousDan [0] (
23:48:34 Quit Zoom2 ("CGI:IRC")
23:49:53*preglow has a birthday ale
23:50:11Bagderthe best kind of ale!
23:50:37preglowoh, yes, i can clearly feel the birthday flavourings perking it up
23:50:42bill20r3I'm getting ready to order an H340, anyone know the cheapest place?
23:51:46 Quit Aison ("( :: NoNameScript 3.72 :: )")
23:55:36[solid]preglow: happy birthday xD
23:55:52preglowanother five minutes and you would have been too late!
23:55:57preglowthanks though :P
23:58:18*preglow Shpongle - Periscopes of Consciousness

Previous day | Next day