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 2005-06-10

00:00:14amiconnIt seems that after a while (perhaps caused by some volume adjusting) the decoder can't follow
00:00:29preglowflac needs a lot of work done to it
00:00:34preglowit doesn't even touch iram as it is
00:00:39amiconnThe red led is blinking irregularly, and playback starts & stops
00:00:46preglowi did the lpc opt, but i don't think that helped very much
00:01:14*amiconn just encoded an album with flac to try out flac
00:01:26amiconnDidn't see a single flac file before...
00:01:34preglowyou don't see them often, no
00:01:40 Quit midk_ ("later...")
00:01:42preglowkeep in mind that higher encoding quality == slower decoding
00:01:45linuxstbamiconn: What options did you use with FLAC? What bitrate is Rockbox displaying?
00:01:47[solid]preglow: unless you're into lossless :P
00:01:53amiconnHmm. Now it happens again, within the same file :(
00:01:55*[solid] has 50 dvd-rs full of flac
00:02:03preglow[solid]: not many are, mostly people recording live stuff
00:02:15amiconnlinuxstb: I used -8. Rockbox displays 959 kbps for that track
00:02:26preglowamiconn: what realtime percentage does flac2wav give that file?
00:02:27linuxstbamiconn: I use the same.
00:02:27[solid]amiconn: maybe it's the same skipping bug i was telling Slasheri about?
00:02:46[solid]amiconn: it happened more often with oog vorbis here tho'
00:03:06amiconnThe progress bar occasionally jumps to zero and back to current position when that happens
00:03:10 Quit DangerousDan (Read error: 110 (Connection timed out))
00:03:38[solid]oh so it's not the same
00:04:04Bagderthe flac site only lists iAudio and Rio Karma as portable players that play flac
00:04:30preglowthose or the only two as far as i know
00:04:34linuxstbBagder: Time to send Josh Coalson an email...
00:04:42preglowwe'll be the first to play musepack and wavpack
00:04:48[solid]preglow: yeah ~~
00:04:57preglowand possibly he-aac
00:05:00preglowif i can make it work
00:05:18amiconnHmm. It seems the files with lower bitrate are consuming more cpu
00:05:25preglowamiconn: like i said
00:05:34preglowamiconn: higher compression may mean more cpu for flac
00:05:39preglowamiconn: since it uses higher lpc orders
00:05:44amiconnThe next track didn't stutter so far, but the red led is blinking since the start of the track
00:05:59amiconn..which is > 2:30 now
00:06:12BagderI noticed that it blinks quite a long time
00:06:15preglowamiconn: you can try optimising the default case in colfire.S for a good asm excercise
00:06:20amiconnSeems it can just even cope to with loading
00:06:38amiconn3 minutes now, it stopped blinking
00:06:46[solid]amiconn: weird... i've been playing a lot of flac with no such problems
00:06:47bill20r3the iAudio X5 supports flac, although I dont know if it's actually shipping yet
00:06:55[solid]amiconn: including 8hrs yesterday ;)
00:06:56preglowit is in korea
00:07:49linuxstbIf I just start FLACs playing and do nothing, then they carry on fine. It's only when I try to do other things that things go wrong (seemingly randomly).
00:07:49Bagderthe iAudios seems _so_ much like iriver clones
00:08:06preglowflac runs very marginally for some files, there's a huuuge difference between the files that run well and those that don't
00:08:15bill20r3did everyone see the $150 jukeboxes ?
00:08:19bill20r3refurbs probablly
00:08:37preglowBagder: let's hope they are so internally as well ;)
00:08:42Bagderthey are
00:08:42crwli think my friend just ordered an iAudio X5
00:08:48amiconnlinuxstb: Trying the problematic track with flac2wav now...
00:08:52crwlmy friend here at finland, ordered from germany
00:08:53BagderI saw some details somewhere
00:08:57preglowBagder: they share more than the coldfire?
00:09:13BagderI don't recall, but the coldfire was what triggered me
00:09:22[solid]preglow: i didn't have problems even with 1100kbps -8 files...
00:09:39preglow[solid]: i can't see a common factor for the slow and fast files here
00:09:54preglow[solid]: i've got two files, both encoded at -8, one gives 180% realtime, the other 110% realtime
00:10:24amiconnpreglow: With .wav writing enabled or disabled?
00:10:36preglowamiconn: for flac i doesn't much matter, afaik
00:10:49preglowmight as well keep it enabled to "simulate" the overhead of buffering
00:11:00amiconn...cause... until now... this track decodes at... ~38% (!!)
00:11:10preglowthat's at 49mhz
00:11:14preglowkick the unit up a notch
00:11:20amiconnHmm, doesn't it boost itself?
00:11:21preglowxxx2wav doesn't do that automatically
00:12:08amiconnOkay, trying again...
00:12:12preglowbut yeah, people who want flac playback should start working on it
00:12:16linuxstbpreglow: Maybe doing "flac -a" on the file will give some clues. It produces a "track.ana" text file containing information about how each frame is encoded.
00:12:19preglowit needs tons of work, and the autor hasn't made it easy
00:12:30preglowwith his malloc hell
00:12:34 Join ashridah [0] (
00:13:12amiconnpreglow: Small wonder it stutters. This track decodes at <100 % even at 120 MHz
00:13:21amiconn~95% currently
00:13:27linuxstbpreglow: If I remember correctly, the mallocs are quite limited - i.e. it should be relatively easy to replace them with static buffers. But it was a few months ago when I looked at it.
00:13:33[solid]amiconn: what music is that?:)
00:14:08[solid]abba being too complicated for realtime playback sounds... sad :)
00:14:09preglowlinuxstb: it even uses realloc
00:14:56linuxstbpreglow: Yes, but I think some of them are only for metadata. Would it help if I tried to replace the alloc/realloc calls with static buffers?
00:15:01 Join Axelerator [0] (
00:15:07preglowof course
00:15:10preglowthen you can see what fits in iram
00:15:12preglowand anyway:
00:15:22preglowgetting rid of malloc and realloc is a major goal for codecs
00:15:23 Quit spiralout (Read error: 131 (Connection reset by peer))
00:15:27 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
00:15:30preglowwe don't want to keep a malloc buffer around
00:15:47linuxstbBut with the downside of forking the code to far from the official codebase.
00:15:59preglowwell, that's up to you ;)
00:16:05 Join Stryke` [0] (
00:16:22preglowi place the elimination of mallocs above that concern
00:16:32Bagder"the iPod cannot play songs without a short gap between songs" ... "The processors found in most portable players, including the iPod, lack the speed to process the headers in lossy (MP3, OGG, etc.) files in the short period of time necessary to obtain gap-free playback"
00:16:34linuxstbNow one of my FLAC files I am listening to is stuttering...
00:16:49Bagderwikipedia's page on ipod
00:16:57preglowwritten by an engineer for sure
00:17:06amiconnlinuxstb: Average percentage 93% on that track
00:17:07 Part Axelerator
00:17:08Rori!anime magical canan
00:17:14preglowamiconn: ouch, ouch, ouch
00:17:17preglowsomeone get to work on flac
00:18:01preglowbut people
00:18:08preglowthe stuttering of flac is no surprise
00:18:12preglowflac has seen almost no optimising
00:18:13preglowso get to work
00:18:39amiconnBagder: ".... in the short period of time", hahahaha
00:18:46 Join Zoom2 [0] (
00:18:57amiconnThere's plenty of time to process the headers
00:18:58preglowBagder: i assume you're updating the page as we speak? :P
00:19:25Bagdernah, too much ipodaganda
00:19:41Bagderwouldn't know where to start
00:19:59[solid]hey, that music quiz game sounds nice:)
00:20:06[solid]never knew about it
00:20:12linuxstbDo the ipod-linux people have gapless playback?
00:20:20preglowthey barely have playback
00:20:32Bagderthey have sometimes-playback
00:20:35preglowi guess their gapless status depends on mpd
00:21:02linuxstbOK :-).
00:21:09Bagderactually, I believe they can play CBR mp3 in realtime
00:21:18preglowthey can do aac in realtime as well
00:21:26preglowin addition to vorbis, sometimes
00:22:14Bagdertheir choice of using linux won't help them get the most of their little cpu
00:22:18preglowindeed no
00:22:26preglowi don't get that part of the deal
00:22:49preglowpreemptive multitasking wont make their task easier either
00:23:23preglowand last, mpd most certainly wont help them conserve cpu
00:24:28preglowflac -8 consistently uses 12 sample long lpc sequences
00:24:57 Join zezayer [0] (~chatzilla@
00:25:26preglowlinuxstb: but yeah, if you wanna help flac speed up a bit, search for critical data that can be put in iram
00:25:47preglowdata that is accessed again and again
00:26:00preglowlike lpc coefs, temporary audio data and codebooks
00:26:24 Join spiralout [0] (
00:26:35linuxstbI've already done that in one place in codecflac.c (can someone have a quick check I did it correctly?). I'll try and look at libFLAC as well.
00:27:11ashridahi'm amazed that they've even got enough memory to do the job
00:27:27preglowlinuxstb: looks good
00:27:59preglowlinuxstb: it's bloody big, though, eats over half the iram in one chunk
00:28:59linuxstbRemind me how big the iram is?
00:29:10preglowthe current codec section is 32kg
00:29:34 Join cheriff_AWAY [0] (davem666@wagner.orchestra.cse.unsw.EDU.AU)
00:30:53bill20r3has anyone ever done hardware mods to increase the ram?
00:31:01linuxstbMaybe a change could be to make FLAC produce the audio data directly in the correct format for Rockbox - i.e. not as one 32-bit array per channel (storing 16-bit values).
00:31:05bill20r3seems like if the players could buffer more it'd help battery life a lot.
00:31:55amiconnbill20r3: The iriver already has a relatively big ram.
00:32:18amiconnBigger ram means more audio data can be buffered, so less disk spinups are necessary
00:32:41*ashridah imagines that finding ram with the same pinout but a larger size would be interesting.
00:32:43amiconnOtoh, more ram is worse if you change your mind and start playing another directory or playlist
00:32:51preglowoh yes
00:32:58preglowif you listen to entire albums, it could be good
00:33:21amiconnThen all this preloaded audio data is discarded
00:33:53amiconnThe gain might be smaller than one might expect
00:34:32amiconnOn archos, increasing the ram to 8 MB (from the default 2 MB) increases runtime by ~25 % (with 192 kbps mp3)
00:34:33preglowi don't think there'd be much gain at all, actually
00:34:52amiconn(On disk based units of course)
00:34:54preglowwell, the archos is a pretty extreme example with its 2mb of ram (default)
00:35:12linuxstbMore RAM would help FLAC though?
00:35:26Bagdernow, if we could get more iram... :-P
00:35:29amiconnYeah, and even there it's only 25% by quadrupling the buffer ram
00:35:54preglowwe probably can use more iram for codecs
00:36:01amiconnThe buffer is actually more than quadrupled, since the firmware takes some ram too
00:36:03preglowbut we need to find out how much everything else needs first
00:36:31preglowthe codecs actually use more than 32kb as it is, the stack needs to be pretty large for some codecs as well
00:36:37preglowlike libmad
00:36:56amiconnHmm, now that you mention it...
00:37:17amiconn...maybe the IllIstr reported earlier is actually caused by a stack overflow?
00:37:30amiconnThe address range was in iram iirc
00:37:48preglowmight very well be, what codec was used?
00:38:17preglowamiconn: indeed, it was iram
00:38:32preglowamiconn: but very early in iram, the stack is placed later than that
00:38:47preglowbut then, it does grow downward...
00:39:04amiconn...and the stack check still isn't enabled on iriver, is it?
00:39:11preglowdon't think so
00:40:25preglowmailing list
00:40:34preglowsomeone had to ask sooner or later
00:40:49 Join XavierGr [0] (
00:41:45amiconndoom 1 & 2 might very well run on iriver, but require some graphics squeezing
00:41:51preglowi think they'd run just fine
00:42:09preglowthey ran well on 33mhz 486, for chrissake
00:42:14amiconndoom even runs on 50 MHz Amiga, even with c2p conversion
00:42:23linuxstbWell, the source code is there (at the bottom of the page mentioned in the email)....
00:42:26amiconnThey even did on 386/40
00:42:26preglowc2p? as in c to pascal?
00:42:46preglowthank god
00:43:25amiconnNative Amiga graphics is bitplane oriented, but usually 3d engines output chunky pixels
00:43:30preglowsomeone insane enough is certain to come by one day
00:43:56Zoom2doom 2 you seirous!?!?
00:44:10Zoom2i rmeember it didnt require much specs
00:44:18Zoom2that would be interesting
00:44:19yngwi_awaydo you know how accurate the battery check is?
00:44:21preglowsmaller screen, tons of ram, decent cpu
00:44:23 Nick yngwi_away is now known as yngwi (
00:45:01amiconnThe H-1X0 screen is not exactly suited for action games though
00:45:06preglowgod, no
00:45:12preglowi don't think the h3x0 screen is either
00:45:27preglowbut still, the novelty value alone is worth the effort of porting it, hehe
00:47:09XavierGrdoes wps supports different fonts for different lines?
00:47:30Bagderwe're waiting for your multi-font patch! ;-)
00:47:52XavierGrI am thinking writting something like that
00:47:56XavierGror at least try
00:48:10preglowhow much memory does a font consume, averageish?
00:48:21*ashridah fondly remembers playing doom on a 386 laptop with a greyscale LCD screen
00:48:25preglownot bad
00:48:29XavierGrhey dont laugh at a starters try
00:48:39ashridahmy god that was sickening. turn and "BLURRRRRRRRRRR" <ashridah> "ARGHL, My eyes!"
00:48:54Bagderbut I'd imaging supporting two or max three fonts
00:49:02preglowit would be kinda cool
00:49:23XavierGryes that is what I mean
00:49:29*amiconn tries enabling stov check...
00:49:35XavierGrto support 3-5 five generic defualt fonts for wps
00:49:45preglowwe'll be cloning the iriver gui in no time
00:49:54preglownot that's it's worth cloning...
00:50:01XavierGrwhat about plugin fonts too
00:50:05amiconnMeh, please don't
00:50:13XavierGrthis should be extra handy for new plugins
00:50:31preglowbut people will be requesting custom widgets before long
00:50:54Bagderbut if we just don't read misticriver, we'll feel better ;-)
00:50:57XavierGris there any manual for all the api functions that Rockbox has?
00:51:14preglowoh, they'll find us, don't worry
00:51:21BagderXavierGr: nopes
00:51:52preglowargh, my beer's gone
00:51:53XavierGrat least a list with all supported functions maybe?
00:51:59BagderXavierGr: nopes
00:52:11Bagderfeel free to start writing one
00:52:18XavierGrthen how do developers know which functions is supported
00:52:23amiconnHuh? With the stkov check enabled, rockbox immediately crashes at boot with IllInstr at 00000000 ??!!
00:52:30BagderXavierGr: reading source code
00:52:33preglowamiconn: i believe that's the ol' problem, yes
00:52:42preglowamiconn: there's a reason it's disabled
00:52:50amiconnWth it has to do at 00000000 ???
00:54:28XavierGrwhat files do you think that I have to mod except wps.c to intergrate multi-font support?
00:55:32amiconnfont.c, lcd-*.c, ....
00:55:59Bagderit isn't a small and easy thing to fix
00:56:08Bagderyou're up for a bite
00:56:30XavierGrAnd I know nothing about Rockbox workings great :(
00:56:44Bagderwell, most people start at that point
00:57:26XavierGrI remember that I encountered an putsxy function with four arguments in the API documentation
00:57:35XavierGrthe 4th argument asked font name
00:57:50XavierGrbut normal putsxy has only 3
00:58:01amiconnThere is no such function
00:58:18amiconnThere is lcd_putsxyofs() which has 4 arguments
00:58:29XavierGrand what does it do?
00:58:47amiconn..the additional one being an offset, to chop off n pixels at the left
00:59:03XavierGrnothing to do with fonts then
00:59:03amiconn...used in the scroll thread(s)
00:59:35Bagderlcd_setfont() sets which font to use
00:59:45 Quit Zoom2 ("CGI:IRC (Ping timeout)")
00:59:45Bagdersince we have two atm
01:00:11 Join bg_ [0] (
01:16:26DEBUGEOF from server (No route to host) (snapshot: netstuff.c line 545)
01:16:26***Saving seen data "./dancer.seen"
01:16:26***Started Dancer V4.16
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:16:26***Unable to connect to on port 6667 (tried 8 times)
01:16:26***No seen item changed, no save performed.
01:20:16***Started Dancer V4.16
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:20:16***Unable to connect to on port 6667 (tried 8 times)
01:20:16***No seen item changed, no save performed.
01:24:23***Started Dancer V4.16
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:24:23***Unable to connect to on port 6667 (tried 8 times)
01:24:23***No seen item changed, no save performed.
01:28:11***Started Dancer V4.16
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:28:11***Unable to connect to on port 6667 (tried 8 times)
01:28:11***No seen item changed, no save performed.
01:32:07***Started Dancer V4.16
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:32:07***Unable to connect to on port 6667 (tried 8 times)
01:32:07***No seen item changed, no save performed.
01:35:54***Started Dancer V4.16
01:35:54DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:35:54DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:35:54DEBUGgethostbyname(2) failed for (Resource temporarily unavailable) (snapshot: netstuff.c line 99)
01:35:54***Connected to on port 6667
01:35:54***Logfile for #rockbox started
01:38:01***Server message 501: 'logbot :Unknown MODE flag'
01:38:01Mode"logbot :+i" by logbot
01:38:01CtcpVersion from freenode-connect!freenode@connect.utility.freenode
01:38:01***Server message 477: 'logbot #rockbox :[freenode-info] why register and identify? your IRC nick is how people know you.'
01:38:02 Join logbot [242] (
01:38:02 Join XavierGr [0] (
01:38:02 Join cheriff_AWAY [0] (davem666@wagner.orchestra.cse.unsw.EDU.AU)
01:38:02 Join spiralout [0] (
01:38:02 Join Stryke` [0] (
01:38:02 Join ashridah [0] (
01:38:02 Join _DangerousDan [0] (
01:38:02 Join bill20r3 [0] (
01:38:02 Join n[o]bby [0] (
01:38:02 Join Naked [0] (
01:38:02 Join Patr3ck [0] (
01:38:02 Join bipak_ [0] (
01:38:02 Join yngwi_away [0] (
01:38:02 Join ehntoo [0] (
01:38:02 Join Ismo [0] (
01:38:02 Join thegeek [0] (
01:38:02 Join tvelocity [0] (
01:38:02 Join xen` [0] (
01:38:02 Join Rori [0] (
01:38:02 Join BossG [0] (
01:38:02 Join [solid] [0] (~solid@
01:38:02 Join _aLF [0] (
01:38:02 Join Bger [0] (~Bager@
01:38:02 Join mbr_ [0] (
01:38:02 Join courtc [0] (
01:38:02 Join Rick [0] (rick@Rick.user)
01:38:02 Join MoosCamaro [0] (
01:38:02 Join edx [0] (
01:38:02 Join elinenbe [0] (~elinenbe_@
01:38:02 Join ]RowaN[ [0] (
01:38:02 Join preglow [0] (
01:38:02 Join linuxstb [0] (
01:38:02 Join pabs [0] (
01:38:02 Join StrathAFK [0] (
01:38:02 Join QT [0] (as@area51.users.madwifi)
01:38:02 Join amiconn [0] (
01:38:02 Join HCl [0] (
01:38:02 Join CoCoLUS [0] (
01:38:02 Join crwl [0] (
01:38:02 Join Plugh_ [0] (
01:38:02 Join C-Keen [0] (
01:38:02 Join dwihno [0] (~dw@
01:38:02 Join Bagder [0] (
01:38:02 Join Seed [0] (
01:38:02 Join lostlogic [0] (
01:38:02 Join tedboer [0] (
01:38:02 Join crashd [0] (
01:38:02 Join bill2or3 [0] (
01:38:02 Join Lynx_ [0] (Lynx@
01:38:02 Join Slasheri [0] (
01:38:02 Join crash_ [0] (
01:38:02 Join ze [0] (
01:38:02 Join [zmaj] [0] (
01:38:02 Join gromit` [0] (~gromit`
01:38:02 Join odd [0] (
01:38:02 Join silencer_ [0] (
01:38:02 Join rob- [0] (
01:38:11amiconnpreglow: Haha, gcc's really fun way of multiplying an address register by 3
01:38:21XavierGrso in order to print a character from a font (with putsxy) rockbox makes the font character a binary bitmap and then prints it, tight?
01:38:25amiconntook me a while to understand: lea %a1@(00000000,%a1:l:2),%a0
01:43:29preglowamiconn: ehhh??
01:44:01preglowgcc does stuff like that ;)
01:44:04 Nick yngwi_away is now known as yngwi (
01:44:22yngwiis it harmles to charge with rockbox??
01:44:24 Join Zoom2 [0] (
01:46:57preglowi've even seen gcc replace multiplies with shifts and adds
01:47:42amiconnpreglow: I know why switch_thread crashes with stkov check enabled...
01:48:15 Join sandstorm [0] (~dxz@
01:48:56amiconnIn that case, switch_thread pushes one register onto the stack, however, this will cause havoc as switch_thread is manipulating the context, i.e. the stack pointer...
01:49:08amiconnThe end:
01:49:14amiconn31028c9c:4cd0 fcfc moveml %a0@,%d2-%d7/%a2-%sp
01:49:15amiconn31028ca0:241f movel %sp@+,%d2
01:49:15amiconn31028ca2:4e75 rts
01:49:20amiconn...lovely :/
01:50:17XavierGris this asm?
01:50:22preglowindeed it is
01:50:32XavierGrgreat! :x
01:50:46amiconngcc does nothing wrong here, but I wonder what we can do...
01:51:33amiconnThe curse of too few scratch registers... :(
01:52:24 Quit Zoom2 ("CGI:IRC")
01:52:45XavierGrin which .c file does the default font gets load?
01:53:21 Part MoosCamaro
02:00:12 Join DomZ [0] (
02:00:54XavierGrhmm for wps intergration two major changes must occur
02:01:26XavierGrfirst I need to rewrite font.c to initialize more fonts
02:02:02XavierGrthen I have to write a funstion to lcd.c that has an argument font type
02:02:20XavierGrand then change wps code to use this new function
02:02:56XavierGrall this is very theoritical from a quick view of lcd-h100.c and font.c
02:03:20amiconn..and change the scroll thread to cope with more than one font displayed at once
02:03:35XavierGrone at a time amiconn!
02:03:43 Join webguest22 [0] (
02:03:48XavierGrI dont think that I can make it
02:03:52webguest22hi all
02:03:53amiconnIf you only display one font, why do you need more than one?
02:04:28XavierGrbut in the wps we need to display multiple fonts
02:04:49amiconnYes, so the scroll thread must be able to handle that
02:05:29XavierGrbut fisrt we must make a function like putsxy that can take a font argument
02:06:12amiconnIt doesn't make sense to always pass paramaters which don't change that often
02:06:46amiconnInstead, it is better to have a separate function to set the font, which is then used in subsequent calls until it is changed again
02:06:53XavierGrthen make lcd_font take more that 2 options
02:07:06amiconnYes, and that's the easiest part
02:07:20XavierGrI cant even imagine the hard
02:07:56XavierGrwhat if I write something like this in a plugin:
02:08:16 Part sandstorm
02:09:34amiconnThis will of course work and display your texts with the 2 available fonts
02:09:57amiconn...because lcd_putsxy(), lcd_putsxyofs() and lcd_puts()
02:10:08webguest22amiconn: scuse for this question, but what's about your much ideas about API graphical and grayscale please
02:10:26amiconn...are all 'fire and forget', i.e. they just paint some graphics
02:10:43 Quit _aLF ("Leaving")
02:11:00amiconn...but this won't work for lcd_puts_scroll(), as this one uses the scroll thread, and has to redraw continuously
02:11:11XavierGrso if setfont can take more than 2 options we could make plugins to be able to print different fonts
02:11:22amiconn...and we surely want scrolling lines in the wps, don't we?
02:11:34XavierGryeah I guess sscrolling will be hard!
02:11:41 Quit webguest22 ("CGI:IRC")
02:12:49amiconnXavierGr: In fact, current rockbox uses both fonts at once in some cases
02:12:58XavierGrdo you think that changeing setfont options will be easy?
02:13:27amiconn...e.g. the status line always uses FONT_SYSFIXED (for numeric battery/ volume display, and also for time display on archos recorder)
02:13:52amiconn...while the file browser of wps use FONT_UI
02:14:03 Quit Patr3ck (Read error: 104 (Connection reset by peer))
02:14:32amiconnsetfont() itself is trivial; font_load() requires a bit more work
02:15:03amiconnThen there is the question how to let the user select more than one font without confusing him
02:15:17XavierGrin the wps?
02:15:31amiconnNo, I mean how to select the font
02:15:37ashridahhave a system font, a title font, and some other font?
02:15:51amiconnCurrently we have 'browse fonts' which selects FONT_UI
02:16:38XavierGrbut the user will not be bothered about the other fonts? Only the wps (or the plugins which is the programmer that makes the choice) will be open for users
02:16:42*ashridah nods
02:16:50amiconnMy idea (which might be a little easier) was to try adding font styles first.
02:17:06amiconn(bold, underline, and perhaps italic)
02:17:37amiconnXavierGr: The user will need a way to select the additional fonts
02:17:52 Quit ]RowaN[ ()
02:17:54XavierGrfor what purpose except the wps?
02:18:05amiconnYes, for wps
02:18:11amiconnThey still need to be selected
02:18:23XavierGrin the wps we can make tokens like %f1 %f2
02:18:36XavierGrthat corresponds to a fixed size font
02:18:45amiconnYes. But how do you map f1 to e.g. atadore.fnt?
02:18:48XavierGrand will be loaded per line
02:19:04XavierGrthese values will be fixed
02:19:05amiconnLoading fonts everytime is a big no-no
02:19:16 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
02:19:19XavierGrwe must select 3-5 default fonts for the wps usage
02:19:27amiconnWe'd need to load all used fonts once
02:19:32amiconn5 ??!!??
02:19:39XavierGror less maybe?
02:20:19amiconnIf they're fixed, who should decide about them?
02:20:43amiconnBtw, I like the current concept way more. The user selects which font to use
02:21:00XavierGrbut nothing will change for that
02:21:29amiconnYes it will if you want predefined fonts for the wps
02:22:02XavierGrpredefined fonts will come into place when someone will utilize the %f1 token for that exact line
02:22:30XavierGrthe rest of it should remain as is between the fixed and ui font
02:24:32XavierGrso you can have the tiny fixed font of rockbox (for values that are not so crucial to the eye from a distance) and then you can use an %f token for information you want to fill the screen
02:25:55preglowgotta go sleep
02:26:00 Quit preglow ("leaving")
02:26:19XavierGrcan the user change the default rockbox font?
02:26:52XavierGreven with having to compile a different file
02:30:30 Join telliott [0] (
02:44:09 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
02:48:30 Part telliott
02:50:10XavierGrI have to go too. Good night all!
02:51:22 Quit XavierGr ()
02:52:45 Quit Stryke` (Read error: 145 (Connection timed out))
03:04:03 Join DJ-suparappa [0] (
03:21:39 Quit BossG (Read error: 110 (Connection timed out))
03:29:46Roriare we all in the UK or something? everyone went to sleep?
03:31:05 Join amiconn_ [0] (
03:31:49ashridahquite a few of the rockbox developers are from northern europe
03:35:58***Saving seen data "./dancer.seen"
03:37:09 Join DMJC [0] (
03:38:18 Quit amiconn (Read error: 60 (Operation timed out))
03:38:18 Nick amiconn_ is now known as amiconn (
03:46:08 Quit _DangerousDan (Read error: 54 (Connection reset by peer))
03:49:27 Join DangerousDan [0] (
04:00:31 Nick cheriff_AWAY is now known as cheriff (davem666@wagner.orchestra.cse.unsw.EDU.AU)
04:05:20 Join QT_ [0] (as@area51.users.madwifi)
04:05:44 Quit ehntoo (Read error: 110 (Connection timed out))
04:09:04 Quit QT (Read error: 60 (Operation timed out))
04:14:36 Quit DJ-suparappa (Read error: 54 (Connection reset by peer))
05:09:50 Join DJ-suparappa [0] (
05:36:01***Saving seen data "./dancer.seen"
05:43:31 Quit HCl (
05:43:31 Quit CoCoLUS (
05:43:31 Quit crwl (
05:43:31 Quit Plugh_ (
05:43:31 Quit C-Keen (
05:45:03NJoinC-Keen [0] (
05:47:55NJoinCoCoLUS [0] (
05:48:22NJoinHCl [0] (
05:48:31NJoinPlugh_ [0] (
05:52:21 Join Zoom2 [0] (
05:54:34NJoincrwl [0] (
05:56:37 Join webguest79 [0] (
05:57:08webguest79Hi everyone, just curious... is there any reason why some of the custom WPS's in the gallery show a little symbol that looks like an angstrom symbol?
05:58:36webguest79for example, Flol's wps screen has little A's just before every dot
06:02:30Zoom2yeah I dont see it
06:02:40webguest79seems to appear everywhere
06:03:22Zoom2what browser are you using?
06:03:44webguest79see those dots between all the numbers? I have a little angstrom symbol between everything
06:03:49webguest79sorry, I meant on the iriver itself
06:04:01webguest79the browser shows what I want to see, but the iriver shows a little angstrom between everything
06:04:25Zoom2oh, I am not sure, havent used that WPS
06:04:53webguest79well, that friendly little symbol appears in almost all the wps screens
06:05:01webguest79Tang's screen shows the same symbol just before the first line
06:05:55Zoom2are you using the most current rockbox?
06:06:15webguest79yeah, bleeding edge actually
06:06:20webguest79I wanted to try the crossfader
06:07:11Zoom2I am using one from last nite, I will try updating it later and see if I have the same problem - its not with me right now
06:18:59webguest79ok, so it's definitely related to fonts, because if I change to a different font, the char disappears
06:26:03Rickwhat font?
06:33:36webguest79the rockbox_default font gives me the angstrom, some of the others give me a weird symbol for the dot, others don't have a dot, some display correctly, and others have the angstrom as well
06:33:39webguest79I tried a lot of them
06:40:14webguest79well, if anyone else notices this, or finds a solution, just leave a message. I'll search the logs... gnite everyone!
06:40:16 Quit webguest79 ("CGI:IRC")
06:41:22 Join Strath [0] (
07:04:00ashridahhrm. should have gotten him to take a screenshot using his player
07:04:15ashridahi'm not seeing anything weird with that wps and the latest cvs checkout here
07:04:16 Quit StrathAFK (Read error: 110 (Connection timed out))
07:09:30 Quit tvelocity (Read error: 60 (Operation timed out))
07:29:51 Quit DJ-suparappa ()
07:36:05***Saving seen data "./dancer.seen"
07:55:59 Join LinusN [0] (
07:59:03 Quit bipak_ (Remote closed the connection)
07:59:05 Join bipak [0] (
08:12:33amiconnGood morning
08:19:00 Quit Rick (Read error: 54 (Connection reset by peer))
08:20:32 Join Rick [0] (
08:22:15amiconnLinusN: I know why rockbox on iriver immediately crashes when the stack overflow check in switch_thread() is activated.
08:22:42LinusNyes, i know that too
08:22:49amiconnThe rockbox scheduler has a bug (or rather: design problem) depending on how this function is compiled
08:23:11 Nick QT_ is now known as QT (as@area51.users.madwifi)
08:23:16LinusNgcc save an extra register when adding the stack check
08:23:20amiconnWe already have a possible solution - do it like jyp did on the gmini
08:24:18amiconnDoing most of the context save on the stack is debatable; it has both advantages and disadvantages
08:24:35amiconnI mean the 'started' flag
08:24:59 Quit Seed (Read error: 145 (Connection timed out))
08:25:48LinusNyeah. i'm not particularly fond of his solution, but it works
08:25:56 Join Harpy [0] (
08:27:47amiconnI think the started flag, and doing the first call with a direct jump is in fact a cleaner solution than to 'craft' a return address on the stack and the let the return do the work
08:28:19amiconnAs I said, the second part of this solution (saving most registers on the stack) is debatable
08:28:27 Join webguest98 [0] (
08:28:47amiconnIt saves space in the context save array, so having a higher max_threads doesn't eat that much iram
08:29:10amiconnOtoh a thread's stack might not be in iram, making the context switch a bit slower
08:31:25amiconnFor the extra register problem, there is another solution, but it's still dirty imho. We could put the crafted return address on the stack twice in create_thread()
08:35:40amiconnI'd prefer the 'started' flag though
08:36:40amiconnOur current solution is highly compiler dependent
08:36:46*LinusN has an H100 now
08:36:50LinusNamiconn: yes it is
08:37:30amiconnWhat do you think about that flag? I'd really like to have a working overflow check....
08:37:46LinusNsure, why not?
08:39:14amiconnI *could* also be done for sh, although not necessary due to the many scratch registers, but it would unify the code (?)
08:40:05amiconnI think about coding the check in load_context() in one asm block, i.e. possibly more efficiently than using 2 asm blocks and doing the check in C
08:41:54LinusNamiconn: that's also an alternative
08:41:58LinusNSlasheri: hi
08:42:25amiconnLinusN: ???
08:42:40amiconnDo you mean coding the whole switch_thread() in asm ?
08:43:55LinusNhmmm, what happens if you move the panic call to a void tpanic(void) function?
08:45:05amiconnHmm. That might work, although it's still compiler dependent
08:45:43 Join ghode [0] (
08:48:07 Join Chamois [0] (
08:48:18 Quit ghode (Client Quit)
08:49:15 Quit Zoom2 ("CGI:IRC")
08:58:41 Join Seed [0] (
09:01:36 Join ghode1 [0] (
09:02:34 Quit ghode1 (Client Quit)
09:11:06 Nick Naked is now known as Hadaka (
09:28:02 Join bobTHC [0] (
09:32:10LinusNSlasheri: you there?
09:32:24SlasheriLinusN: yes i am :)
09:32:42LinusNi wondwr about your cpu_boost fix
09:32:55SlasheriHmm, which one?
09:32:56LinusNyou can cpu_boost(false) in the init
09:33:34LinusNin pcm_playback.c
09:33:41Slasheriah, yes. I did that in audio_init but i am not sure why the cpu was originally boosted at startup
09:34:03LinusNi think we should find the real cause of it rather than "painting it over"
09:34:53Slasherithat's true
09:35:37SlasheriHmm, btw i might know what calls it
09:36:09***Saving seen data "./dancer.seen"
09:36:22SlasheriWhen the sample full of zeros is played at startup and buffer goes empty, pcm watermark callback is called and that will boost the cpu..
09:38:27SlasheriHmm, maybe pcm_boost false could be added to dma_stop so that would unboost cpu everytime playing stops?
09:38:33LinusNwhy is that played btw?
09:38:52SlasheriThere is a really bad dc offset unless something is played..
09:38:54LinusNremember that cpu_boost() is stacked
09:39:22Slasheri(that's only on headphone jack, not on line-out)
09:39:23LinusNeach cpu_boost(true) must have a matching cpu_boost(false)
09:39:55Slasheriyep, pcm_boost keeps that in order
09:40:20 Join markun [0] (
09:41:09LinusNmaybe the pcm watermark callback should only call cpu_boost() if it knows that it is really needed
09:41:42markunI sugest not turn on the headphone amplifyer at all until playing starts, if we can turn on the amp quietly.
09:41:43SlasheriHmm, it calls pcm_boost and that will check if it's needed
09:42:13LinusNstill it leaves the cpu boosted
09:42:16Slasherimarkun: I don't know if it's possible at all to turn it on quietly..
09:42:17LinusNit seems
09:42:35LinusNSlasheri: the iriver firmware seems to handle it well
09:43:06markunyes, but I still couldn't find out how to do it. Spend a lot of time reading the uda1380 datasheet yesterday.
09:43:25SlasheriLinusN: I will try to move the pcm_boost to dma_stop, maybe that will fix all the boost issues :)
09:45:20markunMaybe instead of the uda1380_init functions we should have a more fine-grained control of which parts to turn on and of as needed.
09:49:11markunDon't know if it saves a lot of battery. When turned on the dac+headphone draw 9 mA, is that a lot compared to wat the other iriver components draw?
09:49:58 Join Zoom2 [0] (
09:50:24 Quit Rick ("I… don't need to be here.")
09:50:38Zoom2hey is crossfading working?
09:50:39markunLinusN: Will you do some current measurements soon?
09:51:11 Quit DangerousDan (Read error: 131 (Connection reset by peer))
09:53:37 Join Rick [0] (
09:54:01LinusNmarkun: no, not soon
09:55:19SlasheriZoom2: partially if you enable it. But the crossfading period is quite short, something like 2-3 seconds
09:55:32SlasheriIncreasing it would require also increasing audio buffer size
09:56:03Zoom2so if we increase the "skip mode" the crossfade will increase?
09:56:14LinusNi noticed that it could snip off a fraction of the end of the song in non-crossfading mode
09:56:26Slasheri"skip mode" does nothing at the moment
09:56:36Zoom2I thought that controlled the buffer
09:56:46LinusNZoom2: on the archos, yes
09:56:48Slasherimaybe it will in future :)
09:57:01LinusNZoom2: currently it doesn nothing
09:57:04LinusNon the iriver
09:57:14Zoom2well good work on the crossfade regardless, I was so excited to see that as a possibility
09:57:32Zoom2all my music is sounding better on the rockbox firmware
09:57:48Slashericurrently the buffer is 1 MiB but it should be possible to change the size on the fly (that would require at least restarting audio playback)
09:58:50dwihnoMan, I need to get a ihp-140 asap :)
09:59:06Zoom2do we know what causes the crossfade to work/not work?
09:59:18Zoom2all I notice is a slight "skip" in the next song that plays
09:59:29Zoom2although gapless is looking good
10:01:41Slasherithe crossfade might be slightly buggy at the moment
10:02:29Slasheriand you will hear only a skip if there is some gap between the songs
10:02:56Slasherithat because the crossfade buffer is too short..
10:03:53Zoom2ahh, I am guessing that is not a major change?
10:04:48Slasherinope, the buffer size should be soon configurable :)
10:05:23Zoom2another random question, I was searching on the webpage for but couldnt find an answer... the "battery capacity" option does what exactly?
10:05:32Zoom2yeay! the crossfade worked for one of my songs
10:05:35Zoom2that made me happy
10:05:57LinusNZoom2: it tells rockbox the capacity of the battery, so it can calculate the battery percentage correctly
10:06:38Zoom2and on the iriver 120 what should it be?
10:06:47Zoom2I think I had changed it by accident
10:07:33Zoom2ok thanks
10:12:31markunI've put a 2200 mAh battery in my iriver. The voltage right after charging is higher than the voltage for the old battery.
10:13:03Zoom2wow that was odd.
10:13:16Slasherimarkun: Hmm, would you recommend those 2200 mAh ipod batteries or 1900 mAh ionity? I think i am going to buy a new battery soon
10:13:29 Join B4gder [0] (
10:13:37Zoom2I was listening to a mp3, and when it ended it went to a another song 2/3 into it, and display's a competly different song
10:13:44Zoom2is this a known issue?
10:14:05markunI bought the 2200 mAh from Hong Kong for just EUR 13,70 including shipping. It was too cheap not to buy :)
10:14:15Zoom2now it repeated the song before that, and displaying another song
10:14:16SlasheriI think yes, but i haven't been able to reproduce it yet
10:14:26Slasherimarkun: oh :)
10:14:36 Join sox [0] (
10:14:54Zoom2does this only happen with crossfade mode?
10:14:59markunHCl also bought the same. On the original firmware mp3 played for 29 hours.
10:15:13linuxstbZoom2: Are you using the very latest build? Slasheri made lots of playback fixes last night.
10:15:17soxLinusN: where should I add the WPS req list?
10:15:25SlasheriZoom2: I am not sure, i think not
10:15:28Zoom2yes, I am using the latest bleeding edge
10:15:56Zoom2with crossfade enabled
10:16:33Slasheridid you do anything else like file seeking, next/prev etc. while listening?
10:16:37Zoom2what I was doing before was skipping tracks manually to get to a song I wanted in the playlist because it had stopped randomly in the middle of the song
10:16:37linuxstbSlasheri: [solid] and I experienced different behaviour yesterday with crossfade on and off, but I haven't been able to reproduce them since your last set of fixes. But it could still be an issue.
10:17:02SlasheriZoom2: Hmm, interesting
10:17:18Zoom2Same issue happend with you?
10:17:52Slasherilinuxstb: yep, i am still trying to figure that out.. No luck so far reproducing it :/
10:20:20Zoom2The onlything I was doing was skipping numerous tracks to get to the one I wanted at the time, let that play full and then things got weird. I didnt do any other "extra" behavior. no FW or Rewind
10:20:57Zoom2and this was all in a "playlist"
10:21:02SlasheriZoom2: ok, i will try that
10:21:44Zoom2I will brb though
10:24:12linuxstbSlasheri: I started a directory (of FLACs) playing. I skipping forward/back a few times (which worked) and then let it play. It then stopped playing after a couple of minutes and next/prev didn't work any more. I pressed stop, and then play again in the same directory, and after a lot of disk activity, there is still no sound. The elapsed time indicator is staying at zero.
10:24:53linuxstbAh, attempting to seek in my FLAC file brings it back to life.
10:26:11Zoom2was taking some sleeping pills, back now - let me see if I can reproduce it
10:30:35soxHere's a Wiki page where everyone can add their feature requests for the WPS:
10:30:37webguest98markun: where have you bought the 2200mAh battery ?
10:30:57 Join Patr3ck [0] (
10:31:02webguest98thanks I'll try immediately
10:32:10dwihnoI think bold and italics go under the "different fonts" note
10:32:24markunwebguest98: look for "ipod 2200"
10:33:03markunand don't forget to change the polarity before you install the battery..
10:33:13Zoom2Resume: Off −−- Shuffle: On −−- Selected: Playlist −−- Jumped: Multiple tracks −−- Result"
10:34:04Zoom2Crossfade: On too
10:34:37Zoom2I will try it once more. see if it happens again
10:34:57ashridahi've been finding that mp3s are still skipping the last chunk of a track when changing to the next one.
10:36:03linuxstbashridah: Even with the very latest builds? Is that with crossfade on or off?
10:36:26ashridahand yes
10:36:30webguest98markun: thanks
10:36:31linuxstbSlasheri's fixes last night fixed that problem for me (with FLAC files).
10:36:32Zoom2I dont know if this is related, but I think if you "jump tracks" faster than it buffers the first 10 tracks you run into problems
10:36:34ashridahthe two tracks are part of the same cd
10:36:43Zoom2I am using mp3 files
10:37:18SlasheriZoom2: ah, thanks. trying that
10:37:19linuxstbI have a good test CD (a radio recording - DJ talking over the starts and ends of tracks) so it's easy to spot if the gapless playback is not 100% perfect.
10:37:40linuxstbAnd it's currently perfect.
10:38:42Zoom2yeah it just stops "jumping" after so many songs
10:39:14Zoom2only way to get outta the frozen screen is to hit stop
10:39:14ashridahit's definently broken here. the WPS display changes to the next track before it starts playing and sits at 0 for a few, then all of a sudden, it skips the last block (and possibly the first block of the next track) without joining them together.
10:39:46linuxstbashridah: I get the same "WPS changing too early" problem. But the audio is fine (with FLAC).
10:41:43ashridahodd. it seems to actually be misguessing the length of the track
10:42:15ashridahthe track claims to be 9:18 on the iriver, but 10:11 normally
10:44:30ashridahah, no, it's not getting the length wrong
10:44:33ashridahthe other app was :)
10:44:35Zoom2yeah I reproduced it again, this time outside of a playlist - just jumping tracks in a folder
10:44:48Zoom2settings are all the same though
10:44:50B4gderrockbox does the same
10:45:01B4gdera dir or playlist is no difference internally
10:45:08ashridahZoom2: i think there's a couple of things that are broken while it's in the process of buffering
10:45:14Zoom2I had figured, just thought I would try it out
10:45:49Zoom2I presume that a "buffered" track loads instead of the inteded "shuffled" track
10:46:12Zoom2because when the HD stop spinning it doesnt respin for the song after
10:46:22B4gderI doubt that
10:46:45B4gderthe code asks for the next song and that is loaded
10:46:45Zoom2because I played a few seconds of the intro song, and last night I was told it buffers the first 10 tracks or so
10:46:54ashridahthe problem is, the two tracks i'm getting this problem with are largish, i can't put them anywhere that won't attract attention :(
10:47:02Zoom2but I skipped at least 9 or 10 tracks
10:47:15ashridahZoom2: it'll depend on how big the tracks are
10:47:16Zoom2how large?
10:47:53Zoom2ashrdah: do you mean all the tracks that are being buffered from the first song? or the song I am playing
10:48:28ashridahZoom2: well, no, i mean the number of tracks that can be buffered will depend on how big they are
10:48:29Zoom2because the odd part is, after it skipps to the wrong song, it goes back to the song I had "seeked" but updates the next song ment to be played correctly
10:49:57Zoom2For example, I skipped some files to the Kinks "girl you really got me", after that song it went into a middle of a miles davis song, but showed playing "turn turn turn" by the byrds. After this the Kinks song played again, but the track info that was updated was a slipknot song
10:50:35LinusNclear as mud
10:53:29Zoom2Ok, I opened an mp3 folder, skipped a few files to file "a". that played fully and loaded song "b" that started playing midway. song "b" was not supposed to be the next track. When song "b" finished. song "a" replayed but the track info was that of song "c"
10:53:33Zoom2does that make any sense?
10:54:03B4gderit loads many songs at once
10:54:18B4gderwhen you start song a, it loads song b c and d too
10:55:34Zoom2so what must of happend was when I opened the folder I played the first track, and it must o f loaded some other songs, which I had "skipped"
10:56:13Zoom2hence it played one of the songs correctly, but then the shuffle or fade messed things up due to a buffer error?
10:56:33B4gderI don't think shuffle has anything to do with it
10:57:11Zoom2but what about the loaded songs that I had skipped?
10:57:23Zoom2I have a feeling it had to do with some of those
10:57:42B4gderyes, there are some buffer-related playback problems still
10:58:27SlasheriHmm, i think that problem is almost impossible to reproduce.. The bug has to be figured out by looking carefully the source code
10:58:29crwli've also encountered some quite strange song skipping issues where WPS also showed wrong tracks, and without shuffle
10:59:37Zoom2Slasheri: I have reproduced that 2-3 times so far, buy just loading a file, playing 1-2 seconds of it and skipping 8-9 tracks. The track plays perfectly, and then from there things get messed up
11:00:18Slasheriok, i am still trying
11:01:01Zoom2Open a folder/playlist. Hit play. Skip a good number of tracks. let the song play fully, and then the problem should happen
11:01:22SlasheriNow i started one song, played it 2 seconds and skipped 8 songs
11:01:31Slasheriwaiting this one to fully play..
11:01:46Zoom2do you want to know the settings I was using again?
11:01:56Slasherisure :)
11:02:10Slasherii have now shuffle on and crossfade
11:02:28Zoom2Shuffle, Crossfade, repeat
11:02:54Slasheriok, i think i have the same
11:03:26Zoom2most of my songs are 3-5 minutes long, not sure if that makes any difference
11:08:54 Join narowmind [0] (
11:13:43narowmindanyone know anything abou the Av700 ? there are some indirect reports that it runs linux from the factory
11:14:32narowmindanyone awake ?
11:15:02n[o]bbyi am
11:15:09 Nick n[o]bby is now known as nobbeh (
11:15:26B4gderrockbox is not for any archos multimedia or AV models
11:15:55nobbehand rockbox isnt linux based
11:18:27SlasheriZoom2: btw, are you sure you are running the latest bleeding edge build?
11:18:34Zoom2back. Yes
11:19:09Slasherii was just wondering because the "wrong song display" was corrected. Now it should not display next song at all if it's unknown
11:19:19Zoom2you were not able to reproduce it?
11:19:50Zoom2see under next it displayed a song, that would play AFTER the partial song played
11:19:56Zoom2but all of the tag info would be skewed
11:19:57Slasherihowever, i have had other crashing problems but not that one you describe (i had it before)
11:20:21Zoom2I will try to reproduce it again
11:20:26Zoom2are you using regular mp3 files?
11:20:49Slasheribitrate between 128-240 something
11:20:56Zoom2I have around 150 songs in a folder/playlist
11:21:04Zoom2the bitrate is simmilar
11:21:05Slasheriah, i hadn't that much :)
11:21:18Zoom2I doubt that would make a difference
11:21:20Zoom2but mabye
11:21:23Slasherii will add a few hundreds of songs to playlist
11:23:01Slasherinow there are 592 songs in playlist
11:23:08Zoom2hmmm this time it worked fine. let me try it once more
11:25:15Zoom2gah I skipped 15 tracks and it froze up
11:25:40Slasheriok, i have had that frozing too. Maybe it's related to the problem
11:25:44SlasheriI will try to fix that
11:26:23Zoom2again I have a feeling its buffer related, this time I will skip 14 tracks in the same folder
11:26:23 Quit DomZ ("CGI:IRC (EOF)")
11:26:27Zoom2and see if i can reproduce it
11:27:39linuxstbHas Archos's Linux-based SDK for the PMA 400 been discussed? I know it's of no use to Rockbox, but it's an interesting development.
11:27:47Zoom214 worked fine, let me try playing it on the "15th" track
11:28:17B4gderlinuxstb: you can't even download it without registering your unit
11:28:24B4gderand I have none
11:28:33linuxstbB4gder: Yes, the license is very restrictive.
11:30:20linuxstbI especially like item 17 in their terms and conditions - the right to come to your premises and audit your computer!
11:30:42B4gderthey sure are evil
11:31:15linuxstbI didn't know you had to register a PMA 400 first though - I didn't agree to their terms and conditions, so didn't continue.
11:31:20LinusNcd ..
11:31:34Zoom2Slasheri: it seems I cannot reproduce this problem the last 2-3 times I tried, so I am not sure what is up
11:31:49Zoom2I will try to test it more tomorrow, but the sleeping pills are kicking in so I need to get going
11:31:55SlasheriZoom2: yep, this seems to be more difficult issue..
11:32:09Slasheriok, cu later :)
11:32:18 Quit webguest98 ("CGI:IRC (Ping timeout)")
11:32:23Zoom2good luck bro, I wish I could help you test it some more
11:32:27narowmindLinusN: have you looked at the av700 it looks alot like the pma400 but with a bigger drive and screen.
11:32:30gromit`11:12 < B4gder> -rockbox is not for any archos multimedia or AV models
11:32:35Zoom2I will keep you updated if I notice any changes tomorrow
11:32:38Zoom2goodnite everyone
11:32:45gromit`you told me such things when i asked for the gmini
11:32:45 Quit Zoom2 ("CGI:IRC")
11:32:49gromit`and now it supports it
11:32:52B4gderof course
11:33:02B4gderI talk about present time
11:33:08linuxstbNo means No until it means Yes...
11:33:11B4gdereveryone is free to start a new port
11:33:33B4gderI doubt most people want to know if they can start a new port
11:33:38gromit`maybe it isn't clear that no means "no until yes"
11:33:54B4gder95% of our nos are no until yes
11:33:56linuxstbgromit`: Don't quote me on that.
11:34:10B4gderand btw
11:34:13 Quit Patr3ck (Read error: 54 (Connection reset by peer))
11:34:22B4gderrockbox is not for gmini now either :-/
11:34:37linuxstbOr the current developers all say no, but then a new developer comes along and says yes.
11:35:06B4gderpeople just have to realize that we are an open source project with all what that means
11:35:42gromit`there are several common points between some devices
11:36:10***Saving seen data "./dancer.seen"
11:37:33gromit`i don't blame anything
11:45:43 Quit sox (
11:45:43 Quit QT (
11:45:43 Quit tedboer (
11:45:43 Quit crwl (
11:45:56NJoincrwl [0] (
11:46:31NJoinsox [0] (
11:46:31NJoinQT [0] (as@area51.users.madwifi)
11:46:31NJointedboer [0] (
11:48:21 Part cheriff
11:56:25 Join MoosCamaro [0] (
11:56:31MoosCamaroMorning all
12:02:56 Quit bobTHC (Read error: 104 (Connection reset by peer))
12:03:16ashridahhm. looks like i have more webspace than i thought
12:05:01ashridahSlasheri: you around?
12:05:38*HCl thinks its time for him to actually try the mp3 playback
12:06:04ashridahHCl: feel like poking at a bug?
12:06:33 Quit Seed (Read error: 131 (Connection reset by peer))
12:08:50 Join Enc0der [0] (
12:09:19 Join Lost-ash [0] (
12:09:47Lost-ashHCl: you say anything after i asked you about a bug?
12:09:47Slasheriashridah: yes
12:09:49 Quit ashridah (Nick collision from services.)
12:09:54Lost-ashSlasheri: even better
12:10:00 Nick Lost-ash is now known as ashridah ('s%20Secret%20Service.mp3">'s%20Secret%20Service.mp3 and
12:10:32ashridahthe second one isn't complete, but that shouldn't matter
12:10:43crwlpropellerheads <3
12:10:45ashridahskipping from one to the next should be seamless, but it skips a chunk of the first one
12:11:26Slasheriah, thanks. i will investigate the problem, however, mp3 gapless is not very easy thing to do..
12:11:44ashridahjust figured i'd give you something to poke at that *should* reproduce it
12:12:44 Join Seed [0] (
12:14:16 Quit Enc0der ("Leaving")
12:15:24ashridahi don't particularly want to leave those in my webspace at my isp, so let me know once you've got them, will you?
12:16:15*ashridah notes it's seamless in other mp3s
12:17:05 Join yngwi [0] (
12:18:26 Join Nuxator [0] (
12:18:28Slasheriashridah: ok, i will
12:19:09Nuxatordoes somoene use the remote with rockbox?
12:19:23yngwigood mornig everyone. does any of you get a "nasty" scratch at the beginning of files?
12:19:35Nuxatoryes i do
12:20:05Nuxatorbut not for all tracks
12:20:31yngwiyeah, it does'nt appear with every track
12:21:06Nuxatorig got it last night
12:21:13Nuxatorbut i can't reproduce it
12:21:51Nuxatori jsut get a little pop now
12:22:01Nuxatordo you try the remote with rockbox?
12:22:17yngwiat least it seems to appear every time when the first song starts playing after startup
12:22:23yngwino, but i try
12:22:32Slasheriashridah: got them :)
12:22:38Nuxatori've got problem with fast forward
12:22:51Nuxatorand fast reverse
12:23:04Nuxatori've seen a fix in cvs but it doesn't work for me
12:23:13ashridahSlasheri: thanks
12:23:25Nuxatorif i try ff or fr it goes to next previous track
12:23:46yngwii think ff and fr doesn't work at present, but im sure it soon will
12:25:06Nuxatorit works but not with the remote
12:25:23yngwiyes, thats what I meant
12:25:41yngwiI know it works on the main unit :-)
12:25:55Nuxatori think it's just the same problem for volume
12:26:04Nuxatorthey forgot to put repeat on the buttons
12:26:17 Join preglow [0] (
12:26:24yngwicould be
12:28:02Nuxatorso i think the fix is really easy
12:28:11preglowLinusN: i'll just have david bryant mail you with what username and password he wants for cvs?
12:28:16yngwiNuxator: I also get a small knack at "nearly" every song start
12:28:44Nuxatoryngwi: knack=pop i gess
12:28:54yngwi:-) yes
12:29:09yngwii think so
12:30:21yngwiI sometimes also get a "Io4:IllInstr" error, after which I have to reset, any ideas what this means?
12:30:31Nuxatorillegal intruction
12:30:46Nuxatorso it's a bsod
12:31:04ashridahyngwi: yes. it means rockbox audio playback on iriver is nowhere near complete and stable :)
12:31:20yngwiok, i wasn't worryd anyways, just curious...
12:31:26LinusNpreglow: yes
12:31:33 Join Nibbler [0] (
12:31:37 Join NibbIer [0] (
12:31:49Nuxatori just got it if i tried to view current playlist with more thant 4 or 5 tracks
12:32:28yngwihehe, but honestly, I think this already IS a great piece of software...
12:32:28 Quit Nibbler (Client Quit)
12:33:31yngwii was getting very annyed with the abscence of real shuffle on normal FW after only 9 days of owning a H140
12:33:46Nuxatoron the fly playlist
12:34:11Nuxatorbut it's still half broken in rockbox
12:34:23HClhow so?
12:34:40Nuxatori can't view current playlist and remove tracks in it
12:34:50yngwiis the indicator for the remaining battery time accurate?? if it is, then I may have a Battery Problem
12:36:03Nuxatori don't think it's accurate
12:37:20ashridahyngwi: it hasn't been calibrated yet
12:37:36yngwiok, thanks..
12:37:42ashridahso it probably drops off in a non-linear fashion
12:37:52preglowLinusN:, yes?
12:38:16yngwibecause it says ~8 hours remaining on a fully charged player
12:38:34Nuxatorbut ut doeas played 8 hours flac
12:38:41Nuxatorit does
12:38:57Nuxatorit did play 8 hours flac
12:39:32yngwiyeah, but the 8 hours on mine are before any playback... (and mp3) i also think on normal firmware the battery lives not very long...
12:39:54yngwibtw. sorry for my bad english, I'm out of practice
12:39:57crwlwhere could one order that 2200 mAh battery again? :)
12:40:38Nuxatoranyone for a quick fix for the remote?
12:40:53Nuxatorjust hace to add repeat to some buttons
12:41:28crwlheh, that page crashes my firefox
12:41:34crwlhaven't seen that happening for weeks
12:42:18yngwicrwl: sorry 'bout that, its the thread on the 2100 mAh Li-Polymer Battery
12:43:28yngwithe threadstarter purchased his on:
12:43:55yngwibut it's an ipod battery, if you wanna use it, you have to switch polarity
12:44:47crwli have my warranty left and the original battery in good condition, so i won't be doing anything just yet, but it's interesting anyway :)
12:45:25yngwithere are threads on misticriver on other types of replacement batteries as well...
12:47:53yngwidamn you rockbox, instead of learning latin as I should, i play around with the custom WPS screen :-)
12:49:13 Quit narowmind ("My damn controlling terminal disappeared!")
12:49:21 Quit sox ("Time to move on...")
12:52:26 Join webguest19 [0] (
12:59:47 Join pillo [0] (
13:00:09pillohi all
13:03:13pillojust wanted to let you know that there's a feature request on the tracker that has been completed, it's #989075: bookmarking for text viewer. Direct link is
13:03:38pillostate should be changed to "closed"
13:04:16pillothanks and by the way congrats to everybody working on the sound system on irivers!
13:05:00B4gderthanks, will fix
13:05:01preglowait, i'll close it
13:05:05*preglow awaits
13:09:08 Quit DMJC (Connection timed out)
13:10:15 Join DMJC [0] (
13:12:27 Quit edx (Read error: 110 (Connection timed out))
13:17:44 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
13:20:30amiconnLinusN: I think I'll try to fix the scheduler to work with stack check enabled. Maybe that helps to track down those IllInstr crashes; I suspect a stack overflow somewhere
13:20:55amiconnI'd like to know whether I should go for the cleaner or the faster solution...
13:21:35 Join yngwi [0] (
13:27:43LinusNamiconn: do what you feel is best
13:36:11***Saving seen data "./dancer.seen"
13:37:34B4gdercrosstool is nice"
13:37:48B4gderbuilds a whole crosscompile set in one go
13:45:42 Quit pillo ("CGI:IRC")
13:47:15amiconnLinusN: Then I'll go for the clean solution. I do have 2 nice ideas, I think
13:47:52amiconn(1) It is possible to save the extra flag; the start address field can double as the flag
13:48:13amiconnIf it is non-null, the thread was not yet started, jump there and set it to null
13:49:07amiconn(2) The conditional can be implemented rather efficiently in m68k asm, using the 'instruction shadowing' provided by the tpf instruction
13:50:26 Join tucoz [0] (
13:50:41tucozSlasheri, are you here?
13:51:20Slasheritucoz: yes :)
13:51:44tucozSlasheri: You might know this, but when playing the last song in a directory, it stops playing about 15-20 secs from the end
13:52:01Slasheritucoz: have you tried the latest bleeding edge build?
13:52:42Slasherido you have playlist repeat on?
13:52:47tucozIt's like when there is no more tracks to buffer, it stops.
13:52:53tucozlet me check
13:53:23tucozrepeat - off
13:53:32Slasherimaybe you should try turning it on :)
13:53:37Slasheribut that's still bug indeed
13:54:08tucozTurn it on to One or all?
13:54:26Slasheriafter that the playing should never stop
13:54:52tucozOh, ok. Then this will never occur because rockbox will start to buffer the first song after the last one right?
13:55:32Slasherior maybe you should create a playlist with more songs on it :)
13:56:04tucozNot a problem for me, but still a unhandled event though
13:56:46tucozThanks for the tips anyway
13:56:55linuxstbSlasheri: I've got a new problem. Almost every time I try to play a FLAC file, Rockbox doesn't go into the WPS screen, and the disk activity light is on. Even when the disk stops, nothing happens. This is with latest CVS.
13:57:15Slasherilinuxstb: ah, i will check that
13:57:45linuxstbBut it doesn't always fail - very occasionally, it works.
14:01:58tucozSlasheri: you've been busy today. Downloaded the bleeding edge an hour ago. Didn't notice your latest changes.
14:02:16tucozI'll try again with a newer build. :)
14:04:00nobbehslash > *
14:06:01 Join zezayer [0] (~chatzilla@
14:10:07tucozSlasheri: Still the same. About 20 secs from finishing the last song it stops playing and returns to the file browser.
14:11:10Slasheriok, that will be fixed
14:12:28tucozsounds great
14:15:21*tucoz is away
14:19:09SlasheriHmm, I00: at 40102710
14:23:01B4gderI got the same, at 40102704
14:23:05B4gderjust now
14:23:16B4gdercvs from yday
14:23:59B4gderand no reset "pin" in sight...
14:26:44SlasheriHmm, do you know what the I00 means?
14:27:05Slasheriok :)
14:27:39tucozBtw, do you know if rockbox will be able to detect trackchanges when recording from optical in?
14:28:24tucozFor all I know, irviver's firmware doesn't do this.
14:28:40tucozOr at least, didn't a year ago ;)
14:29:17preglowdepends if we can find the specs
14:30:32tucozpreglow: ok, I remember that my old minidisc did that. So I guess my cd-player sends some kind of information on track changes
14:31:19amiconnSlasheri: How are the playlist viewer crashes related to the plugin_loaded status? (Trying to understand your fix...)
14:32:58tucozhow do I use google to search old irc-logs?
14:33:28Slasheriamiconn: playlist viewer uses the plugin buffer and if plugin_loaded status is not set to true, playlist viewer tries to use the whole plugin buffer
14:33:55B4gderI don't think you should so such fixes
14:34:11B4gderthe codecs are to be moved soon anyway
14:34:33SlasheriB4gder: yes, but that fix was small and it cannot harm anything else
14:34:53B4gderit is unnecessary when the codecs are fixed
14:35:01B4gderand I don't like unnecessary
14:35:06amiconnIt can be taken back when codecs are moved
14:35:30amiconnA comment saying so could be added
14:35:31B4gderI'm just saying that let's not do fixes this way
14:37:16Slasheriwhen codecs are moved, the fix will vanish with them
14:37:32SlasheriIt is defined inside #ifdef IRIVER_H100 and codec_load functions
14:37:39[solid]tucoz: the latest iriver fw can do that
14:38:08tucoz[solid]: ok, sounds great.
14:41:36preglowtucoz: it does, my minidisc player too did that
14:42:48 Nick NibbIer is now known as Nibbler (
14:51:20tucozpreglow: Found some specs for the spdif:
14:52:38 Join edx [0] (
14:54:44tucozhmm, I wonder if my cd-player sends that copy-permit bit on copy protected cd's, and if the iriver fw cares about it.
14:54:59tucozHave to try when I get home.
14:55:18preglowsincerely doubt it cares
14:55:58tucozme too, if it does, I'm sure rockbox will...well..neglect it's precence
14:58:56amiconnThe archos recorder firmware usually respects the copy protect bit, but there's a special method to disable this
14:58:56 Quit webguest19 ("CGI:IRC (EOF)")
14:59:06amiconnRockbox on archos doesn't care
14:59:51tucozamiconn: hehe, why not? ;)
15:00:43linuxstbSlasheri: Just to let you know that your latest commit seems to have fixed the "not playing" problem. Thanks.
15:03:45tucozbye bye
15:03:48 Part tucoz
15:08:35preglowvbr mp3 files only have 100 fast seek positions?
15:08:55LinusNpreglow: yes, with the Xing header
15:09:02LinusNthe VBRI header has more
15:09:03preglowthat explains why this big mp3 is so bloody hard to seek
15:09:25preglowit keeps snapping to the same place when i try to seek to a specific place
15:09:28dwihnoRe-synching to a stream can't be that difficult... Or can it?
15:09:33preglowthen again, xmms is crap
15:09:41preglowresyncing to an mp3 stream is very easy
15:09:44preglowjust look for the sync marker
15:09:54dwihnoSeek to an offset, find next valid frame, decode
15:10:10dwihnoso why are there markers then?
15:10:17preglowfor fast seeking
15:10:22preglowwalking through the bitstream isn't very fast
15:10:29LinusNdwihno: because the frames aren't the same size
15:10:42LinusNso you don't know wrhere to seek
15:11:03LinusNCBR streams are easy
15:11:18dwihnoLinusN: as I said, just find an offset and hook on to the start of the next frame
15:11:28dwihnoWhy isn't my way good? :)
15:11:31LinusNstill, rockbox is supposed to interpolate between the markers
15:11:52LinusNdwihno: it's just that tiny part of "finding the offset" :-)
15:12:05preglowdwihno: what if the file has very low bitrates frames at the start and high bitrate frames at the end?
15:12:14preglowdwihno: you can't just seek to the middle and think you're in the middle of the song
15:12:19dwihnoLinusN: you have the filesize. If you seek ~50% in the file it can't be that bad
15:12:23dwihnopreglow: of course you can :)
15:12:30preglowdwihno: and like i said, walking through the bitstream to find the middle is very slow
15:12:34dwihnopreglow: approximation to the people :)
15:12:41preglowfoobar does this
15:12:46preglowand it's very slow even on my computer
15:17:42 Join asdsd____ [0] (
15:17:45linuxstbBut isn't this an unsolvable problem? The only reliable solutions are to either walk through the file from the beginning, or to put the MP3 file inside a container.
15:17:55preglowmp3 has seek tables
15:17:55dwihnoThen VBR is the tool of the devil! Period! :)
15:17:58preglowthat is, extensions
15:18:17preglowmost vbr files i've seen have xing headers which contain 100 seek points
15:18:25preglowdon't know how common the vbri header is
15:18:30linuxstbdwihno: storing raw MPEG audio frames without a container format is the work of the devil.
15:18:52preglowwhy? i think storing raw mpeg audio frames without a container makes less sense
15:19:27linuxstbBecause if the frames are variable size, or if you can have non-audio data (such as ID3v2 tags) in the middle of the file, you can't seek in them.
15:19:47linuxstbMPEG audio is designed to at least be inside "PES" packets - which contain a timestamp.
15:19:54preglowid3v2 doesn't belong in a container
15:20:03preglowthe container is supposed to handle commenting needs
15:20:37linuxstbpreglow: That's my point - there have been so many bad hacks to MP3 files, when they should have been iside a container such as Ogg.
15:20:40preglowif mp3 had been wrapped in a container, we wouldn't have had a gapless issue, or need for hacks like extra stuff like xing and lame headers
15:27:29 Join Musicmad [0] (
15:29:36 Join tvelocity [0] (
15:30:17 Quit elinenbe (Read error: 131 (Connection reset by peer))
15:34:24 Join dapureplaya [0] (
15:36:13***Saving seen data "./dancer.seen"
15:37:22Musicmadhow do you revert to default wps?
15:39:31LinusNrename the custom wps file
15:39:39LinusNor reset the settings
15:41:53LinusNpreglow and linuxstb: remember that mp3 is a streaming format
15:41:56Musicmadweird. Still got no wps other than the battery, volume and play icon.
15:42:10LinusNMusicmad: hmmm, maybe a bug
15:42:58MusicmadLinusN: yeah - tried various mp3's though. Maybe I don't upgrade correctly. I just unpack the zip on my player and run the .iriver file.
15:43:11linuxstbLinusN: I know - that's why we have all these problems.
15:43:16preglowLinusN: sure, you don't use containers when streaming
15:43:22LinusNMusicmad: that should work
15:43:39linuxstbpreglow: You do, but a different type of container - for a different purpose (error detection/correction).
15:43:42preglowbut it was designed for ordinary storage use as well
15:43:50linuxstb"designed" ?
15:43:57MusicmadLinusN: any ideas then :)
15:44:08LinusNMusicmad: reset the settings
15:45:36MusicmadLinusN: thanks that worked.
15:48:19preglowif anyone still can hear artifacts in mp3 playback, let me know. i think i can hear some sometimes, but i don't really trust my hearing
15:48:20LinusNmisa like the idea
15:48:58preglowLinusN: i agree
15:52:00yngwiPreglow: I hear a loud scratch everytime the first song after startup is played (and sometimes on the beginning of a file) and a small knack or pop noise on the beginning of every file. But I guess you don't mean this kind of "artifact"
15:52:00 Join Tomhab [0] (~tom@
15:52:36preglowhmm, that's not exclusively an mp3 problem, is it?
15:53:17yngwiI don't know, I only have MP3.. Sorry
15:54:51preglowit does seem to be an mp3 problem
15:55:46 Quit Nuxator ("CGI:IRC")
15:56:17linuxstbSlasheri: I see you've committed a patch that delays buffering after 2MB. I haven't tested it yet, but how does that work with FLAC files, where 2MB isn't very long at all (maybe 15-20 seconds)
15:57:08linuxstb2MB will obviously be even less if you're playing WAV files.
15:57:20amiconnLinusN: Re the patch you mentioned - it seems like a nice idea. This functionality should probably be added to the PGUP/PGDN handling as well
15:58:15preglowehh, i paused the player, went away for a couple of minutes, and it locked up when i pressed play
15:58:26B4gderI think the 2mb fix was a bit quick to do
15:59:59B4gderit needs to adjust to codec/compression
16:00:05B4gderto be really good
16:00:08Slasherilinuxstb: Hmm, it doesn't work with flac files quite well
16:00:18SlasheriBut with ogg/mp3 it works
16:01:58Slasherilinuxstb: Hmm, maybe with flac files the delayed buffering should not be activated at all?
16:02:09preglowdid mp3 files use to have a click in the start?
16:02:22LinusNpreglow: with archos?
16:02:28preglowno, iriver
16:02:32preglowbefore my sound quality fix yesterday
16:02:46LinusNi have always heard a faint click
16:02:49Slasheripreglow: I don't yet know that causes that garbage when song starts..
16:03:01preglowit's only for mp3s
16:03:04preglowit seems
16:03:05SlasheriAh, hmm
16:03:34B4gderSlasheri: doing such things for specific codecs signals something is wrong
16:03:50preglowSlasheri: no, i was wrong, it happens for oggs as well
16:04:17preglowi'll just ignore it, then
16:04:18SlasheriB4gder: Hmm.. But flac files require so much buffer space that it's not very useful to do that
16:04:40preglowthen the codec should somehow tell you it doesn't want that behaviour
16:04:52B4gderthe codec should rather set the limit
16:04:54Slasheripreglow: yes, that's correct :)
16:05:03B4gdersince it knows the compression etc
16:05:07Slasherii will do that
16:05:22preglowdidn't someone start working on a separate plugin api for codecs?
16:05:27B4gderthen flac can set it to 16mb or something
16:05:31SlasheriLinusN is working on that
16:06:58B4gderno one works on getting wavs to play?
16:07:12preglowno, afaik
16:07:20B4gderkind of funny :-)
16:07:24Slasherisomeone has to write codec for wavs
16:07:28preglowshouldn't be prohibitively hard, hehe
16:07:28Musicmadget a crash when viewing current playlist.
16:07:31Slasherithat's not very difficult
16:07:48preglowjust load what you need to the buffer, with an endian swap
16:07:56SlasheriMusicmad: with latest bleeding edge build?
16:08:10Musicmadlatest daily.
16:08:21SlasheriMusicmad: ok, that's not yet fixed
16:08:25Slasheritry the bleeding edge one
16:08:28Musicmadi will
16:09:58MusicmadSlasheri: works now
16:10:08preglowSlasheri: the pop at the start is probably just a step in the waveform caused by the last sample from the last track being held by the dac
16:10:30 Quit Tomhab (Read error: 60 (Operation timed out))
16:11:47MusicmadSlasheri: progress doesn't seem to be in sync with music after heavy forward seeking.
16:12:19SlasheriMusicmad: ok, that might be possible because the seeking is still experimental
16:12:30Musicmadok :)
16:13:00preglowso you should probably play some zero samples before a track change
16:14:15Slasheripreglow: Hmmm
16:14:33SlasheriThat should not be necessary (unless audio is stopped)
16:14:45preglowSlasheri: audio is stopped briefly when you switch tracks, isn't it?
16:15:00preglowSlasheri: like when i press 'next', not when the track is finished
16:15:13Slasherino, it's not
16:15:28Slasheriyes, it's when you press next
16:15:51preglowthere is no glitch when audio is continious
16:15:55preglowof course
16:16:52linuxstbB4gder: There is no need for WAV when you have FLAC - but I've been thinking about doing it. It should be almost trivial (until the bug reports come in with weird WAV files).
16:17:24B4gderwell, some people will have wav files they want to play without using flac
16:17:50amiconn...and probably the first shot on iriver recording will probably record to wav
16:18:16amiconnIt would be really odd when the iriver couldn't play its own recordings...
16:18:18linuxstbSomething else I want to do quite soon is a "wav2flac" plugin. I don't think real-time encoding is a good idea (for quality reasons), but on-iRiver encoders will be very useful.
16:18:43amiconnlinuxstb: lame !!
16:18:46preglowsomeone should also make a 2wav codec loader
16:18:52preglowso we can get rid of the xxx2wav fellas
16:19:06linuxstbI also think it's too risky to do real-time encoding - it will always be pushing the CPU to the limit.
16:19:34preglownot necessarily
16:20:37preglowa flac encode is a pretty speedy affair
16:21:16linuxstbNot as speedy as decoding, and we are still struggling to decode.
16:21:20dwihnoDepending on the CPU, perhaps variable compression?
16:22:20preglowwe'll see
16:22:24preglowi don't care much about recording
16:22:26linuxstbI see recording being used for three reasons - voice recording (i.e. low quality), live concert recording (i.e. best possible quality) and for people who want to convert CDs without using a PC.
16:22:29preglowwav will probably do for me
16:23:42linuxstbSorry, got to go...
16:23:44 Quit linuxstb ("Leaving")
16:25:11dwihnopreglow: neither do I. but then, there will always be
16:25:19bipakdoes the rockbox status on the page belong to the cvs status? :)
16:26:53Musicmadwish I had the skill to do som wma codec :)
16:27:02[solid]preglow: the h1x0 are a bootleggers dream actually if they wouldn't have the bugs/misfeatures the current fw has
16:27:26dwihnobooo! no wma!
16:27:33bipakdwihno: :)
16:27:35[solid]preglow: and i think encoding realtime could actually improve the runtime over wav...
16:28:00Musicmaddwihno: explicitly no? - why :) - it's the most used format for online music shops.
16:28:08Musicmadnon-ipod that is.
16:28:20dwihnoMusicmad: proprietary format!
16:28:35 Part asdsd____
16:28:47Musicmaddwihno: - well isn't there an open source floating point codec for it.
16:28:56preglowgo ahead
16:28:57preglowport away
16:29:05Musicmadno skillz :)
16:29:08dwihnoWhen we're speaking of it...
16:29:18dwihnoOnline stores should sell the music in a lossless format
16:29:58dwihnoOtherwise, it would not make sense paying such a relatively high price for something you don't know how well it's been compressed
16:30:19Musicmaddwihno: actually I think the price is fair. It's around ˝ price in Denmark.
16:30:21dwihnoI recall my pal telling me how great he thought the Xing encoder used to be, since it was such a fast encoder
16:30:23[solid]warps bleep store uses lame -ape :)
16:30:38dwihno-ape ? :)
16:30:39Musicmad[solid]: got a link?
16:30:48[solid]it's by far the best online shop i've seen
16:31:05[solid]and they've got all sorts of sold-out releases for sale there..
16:31:44dwihnoLossless to the people!
16:31:54[solid]and to quote them: "Secondly, Bleep music has no DRM or copy protection built in. We believe that most people like to be treated as customers and not potential criminals - DRM is easily circumvented and just puts obstacles in the way of enjoying music."
16:31:57*dwihno has started re-ripping all his CD's to FLAC :)
16:32:16Musicmad[solid]: excellent
16:32:18dwihno[solid]: a great quote indeed
16:32:44[solid]if i had the money, the bank account, and wasn't into lossless i'd surely buy there :)
16:32:48preglowbleep is the only place i've ever bought compressed music
16:32:54preglowand i've only bought the out of print releases
16:32:55 Join t0mas [0] (
16:33:10preglowbleep's even got some flac releases now
16:33:15SeedDRM is good. DRM is good. DRM is good.. damn, brainwashed again
16:33:25[solid]preglow: yeah, i've heard you could buy all autechre in flac
16:33:43SeedAutechre was made for lossless
16:33:52[solid]oh indeed
16:33:55[solid]how lovely
16:33:58preglowi've got most or all of their releases on cd
16:34:00[solid]and it's all of them
16:34:03preglowand a couple of 12"
16:34:03Seedone of the only bands that manages to beat MPC at −−extreme
16:34:09[solid]biggie up bleep :)
16:35:42[solid]dwihno: i was wrong, -aps, that's −−alt-preset standard (-ape is preset extreme)
16:36:02dwihno[solid]: ah, okay...
16:36:11preglowbut yeah
16:36:16preglowbuying lossless music sounds better
16:36:21dwihno[solid]: If I were to buy music online, I would not settle for less than lossless.
16:36:25preglowi'm very sceptical to buying lossy compressed music
16:36:39dapureplayais tehre a place whereu can buy lossess music?
16:36:39dwihnoIf I want the music in MP3, OGG or whatever, I can encode it myself
16:37:15Seedthe problem is how they rip, not encode
16:37:17bipakdwihno: non compressed format would cause higher traffic and then higher prices for the music, i think
16:37:24ashridahdapureplaya: itunes can supply aac lossless music iirc. of course, it's protected aac, afaik.
16:37:44dapureplayaso each song is like? 60-70mb?
16:38:02ashridahi have no idea
16:38:13ashridahi've never used itunes, being a linux user and all
16:38:19*t0mas has a new speaker set
16:38:24t0masCreative I-Trigue
16:38:37t0masman... there is a volume limit.... but it's not hardware... it are my ears
16:38:48bipaklol :>
16:38:49dapureplayat0mas: i was thinking about getting em for my PC. how do they go
16:38:58t0masone word: PERFECT
16:39:18t0masreally small speakers...
16:39:26t0maslook very nice next to a tft
16:39:38dwihnobipak: I'll test something. Brb.
16:39:39dapureplayahehe sounds nice. Yeah i got myself teh Klipsch GMX-A (2.1). They thump hard. can't use it atm. the house is sleeping heheh
16:40:04*t0mas has the basslevel at 33%
16:40:16Roriis that the 5.1 speakers?
16:40:20t0masbut still half of my rook is shaking
16:40:21Rorito t0mas
16:40:33t0masRori: nope, 2.1 for music
16:40:43t0masyou can better get a 2.1 at the price of a 5.1
16:40:49Roriwhich model?
16:41:04dwihnobipak: So the price is 100 SEK for an album (regular price in the store is approx 175 SEK)
16:41:05t0mas3450 black = 3500
16:41:09t0maslooked at the 3200 first
16:41:16t0masbut these sounded much better in store
16:41:24[solid]dwihno: i wouldn't buy music online at all i think, i like the physical part, the packaging... 90% of the cd's i bought in last two years are digipacks or similar :)
16:41:35t0masmy sister has 3200
16:41:37t0mastime for a test :P
16:41:41dwihno[solid]: there's more...
16:41:55Roribit [ricey
16:42:00bipakdwihno: for uncompressed files?
16:42:01dwihno[solid]: you pay ~60% of the retail price for a download (which is lossy and you don't know how "well" it's encoded)
16:42:15t0masRori: same price as the 5.1 set I wanted
16:42:35Roriwhy didn't you get the 5.1 then?
16:42:39t0masand I decided to take these... and not 5.1 because I already have a decent 5.1 set for my dvd player
16:42:44dwihnobipak: So basically, the lossless should be set at that price and perhaps having mp3 or similar at 60% of the flac price
16:42:49t0masand these sound better
16:42:52[solid]dwihno: there is one lossless online store (not talking about allofmp3 which is illegal after all, and even if it's legal i bet the musicians don't get a dime from their sales)
16:42:57t0masand all my music is normal stereo on pc
16:43:02t0masso why take 5.1 for my pc
16:43:13[solid]dwihno: but it's average cost is higher than that of a cd bought in a store...
16:43:22bipakdwihno: yea that would be good :)
16:43:28RoriI feed my soundcard via SPDIF to a Sony top notch surround sound amp and some expensive speakers
16:44:06t0masmy dvd player is a B&o (don't know if they're international?)
16:44:25t0maswell... my parents dvd player ;)
16:44:38RoriJBL 15" 170 watts per channel fronts, 150 watt Paradigm center, 100 watts Wharfedale rears
16:44:40t0masbut I'm the only one who knows how it works
16:44:46RoriNo sub as I don't reall yneed it
16:44:47t0masRori: RMS?
16:45:11RoriNo idea :)
16:45:15Roriit sounds fucking mad
16:45:16preglow[solid]: allofmp3 isn't illegal
16:45:22Roriscuse the language
16:45:35t0masRori: I guess pmpo.... at 170 watts RMS is almost impossible...
16:45:51RoriPut some DTS title on like Lord of the Rings and the place shakes
16:46:13t0masI'm looking for some thx test movie on DVD
16:46:14RoriThe fronts are on proper pedestals
16:46:16t0masit's on some dvd's
16:46:40dapureplayat0mas: how many watts is the i-trigue?
16:46:44Roriyeah I have a few of those
16:47:11t0masdapureplaya: my sisters 3200 is 25 watt RMS\
16:47:14t0masand this one is 35
16:47:15t0masI guess
16:47:27t0maslet me check
16:47:27preglowsite says 48
16:47:49dwihno[solid]: hm. purchasing CDs is not that bad after all.
16:48:00t0mas30 Watts RMS
16:48:05dwihno[solid]: Although I FLAC encode them and store them on my "big" disk
16:48:18t0massays the manual
16:48:51dapureplayanice. My Klipsch are like 76 Watts RMS or something. It's disturbingly powerful for a 2.1 set. I was scared at first when i heard them/
16:49:06t0masyes, I have the bass at 1/3 now
16:49:14t0masand I guess the rest of the hous doesn't like this :P
16:49:37t0mashm... but still I think my sisters speakers look better :)
16:49:46t0masthe 3200
16:49:57dapureplayahaha yeah they don't like it, but they have to put up with it. It only thumps wduring the day and when people are up. I spent AUD$220 on them.
16:50:12dapureplayat0mas: you want power over looks tho.
16:50:45t0masthese were 70 euros
16:50:49t0masand I got them for free
16:51:15t0masbecause my tft has a dead pixel (bottom 3 pixels from the border, but still)
16:51:46t0masI payed the shop 30 euros for a pixel problem free tft... and they were unable to give me this type...
16:51:51dapureplayat0mas: free? Haha! No complaints then! I was expecting everything that i got for these speakers. Were they a present or something for you?
16:52:09t0masbut the shop had a big problem
16:52:11t0mashaving me there...
16:52:29dapureplayat0mas: oh. long story?
16:52:34t0masbecause they had no way to give me a new screen (Iiyama is in production trouble)
16:52:52t0masand they sold screens with pixel errors with a 70 euro discount
16:53:01dapureplayat0mas: oh fair enough.
16:53:06t0mas(but then you don't know where the pixel is... and how many you have)
16:53:34 Join DJ_Dooms_day [0] (
16:53:36t0masso I wanted 70 euros... and they started talking about the possibility that it was in the middle... and the discount screens might have 5 pixels... bla bla
16:53:52dapureplayaanyway speaking of creative products, anyone seen the Creative 5GB Neeon?
16:54:04t0masthen I said: "Ok, I give mine back... put it on that pile... and give me the money... then I bought it back with discount + my new speakers set :P
16:54:13dapureplayat0mas: nice
16:54:47DJ_Dooms_daySo how far along is the new rockbox now? I see it can play MP3, any bugs?
16:55:22t0masIf we would know we would fix DJ
16:55:30Roriah 160 watts not 170 watts
16:55:35t0masbut there are some bugs in the graphics patch
16:55:38t0masfixed in my local copy
16:55:44DJ_Dooms_dayNo doubt, not all bugs can be fixed straight away though
16:55:45RoriI have these. They are made by JBL so I read someplace
16:55:51t0masbut still working on some improvements
16:55:54RoriBut under the Realistic branding
16:56:01RoriThey are BIG!
16:56:01[solid]dwihno: i rip and store flac on dvd's, don't have the space on my disk... ogg -q6 is enough + it transcodes nicely directly from flac with oggenc and ogg is perfect for my h120 :)
16:56:02dapureplayaDJ_DOOMS_DAY: plays mp3, ogg and flac.
16:56:03t0masah and the next happy story :D
16:56:03DJ_Dooms_daygraphics patch?
16:56:07t0masI get a free archos
16:56:17t0masbroken... but it's fried with a wrong charger
16:56:21RoriRMS is 40 watts
16:56:22t0masso I guess I can re[air it
16:56:27t0masRori: that's a lot
16:56:30t0masmine is 30
16:56:33RoriThey kick booty ;)
16:56:44 Part Musicmad
16:56:48t0masDJ_Dooms_day: graphics on wps
16:56:58DJ_Dooms_dayAh, yes
16:57:01dwihno[solid]: ah. dvd's is also nice.
16:57:06Roritotal weight of these babies is 46.2 lbs. (21 kg)
16:57:14DJ_Dooms_dayGrey scale still AWOL?
16:57:14dwihno[solid]: disks are so cheap nowadays... dvd might be good for backup though
16:57:42[solid]dwihno: yeah i recently got 10 dvd's (44GB) for equvalent of $2.5
16:57:58[solid]dwihno: or exactly 2 euro with the todays exchange value :)
16:58:10dwihno[solid]: really, really nice!
16:59:00dapureplayatomas: <−−my babies ;-)
16:59:33t0masghe... they look cool
17:00:16*t0mas np: iRiver - iRiver, Catch the digital flow! (0:04) []
17:00:20t0masgood bass testing song
17:00:51dapureplayar u serious?
17:01:20RoriI can suggest better bass testing tracks
17:01:36t0masjup... some hardcore dance tracks
17:01:40t0masbut that iriver track is funny
17:01:49dapureplayaor any rnb/rap songs
17:01:50t0masdesigned for SRS on iriver I guess :)
17:02:07t0masor enya... to test overall quality
17:02:15t0masor the lotr music
17:02:16Rickdapureplaya: can you not speak in aolbonics? :/
17:02:27RoriGet the Junkie XL remix of BT's Somnambulist
17:02:31RoriThat will work the bass
17:02:35dapureplayaRick: aolbonics? sorry i'm not with teh lingo.
17:02:59Rorior the Burufunk remix if you want some kicking bass beat
17:04:15dapureplaya <−−-That just looks so cool. =(. Sucks.
17:05:36Roriyeah pisses me off the support iPod gets
17:05:44dapureplayai concur
17:06:07dapureplayai've seen my mate's Bose Sound-Dock. It is suprisingly powerful for such a tiny system.
17:07:24t0masBose always has these kind of tricks
17:07:29t0masgreat sound form a little box
17:11:17 Part LinusN
17:15:39*t0mas np: Usher - Yeah (Feat. Lil Jon And Ludacris) (0:52) []
17:15:51t0masman... this song sucks
17:16:11t0mas(this r&b is what you mean right?)
17:16:27dapureplayauh. there are other songs which you can better for bass
17:16:40t0masor, and maybe better for my mood
17:16:44t0masI hate this...
17:17:03dapureplayalet's c......J-Kwon - Tipsy. I think that's a good tester.
17:17:20t0masis that the fucking song about a girl getting tipsy in a bar?
17:17:24DJ_Dooms_dayGet 'The Chemical Brothers - Bass Test'
17:17:39dapureplayahahaha...i dunno
17:17:47DJ_Dooms_dayHas a nice big fat bass wind down
17:17:59*t0mas doesn't like R&B at all...
17:18:08t0masmore rock, dance... and even classic
17:18:09dapureplayaor any of teh old prodigy stuff.
17:18:10*preglow Calyx feat. Gridlok - Dead Ringer
17:18:11t0masbut no r&b
17:18:28dapureplayawhat about funky house?
17:18:35Roriremember those bass test CD;s you used to get from car stereo fitting places? :)
17:18:39DJ_Dooms_dayAnyway, bed time. Catch yaz all later
17:18:45*t0mas np: Scooter - One (Always Hardcore) (0:04) []
17:18:52t0masApres ski cd
17:18:54 Nick DJ_Dooms_day is now known as DJDD_SLEEP (
17:18:55RoriCD's with names like 'Ultimate bass' etc ;)
17:19:20t0masyeah... everybody here knows them...
17:19:29t0masbut it's a "black" thing here
17:20:06t0masmostly jokes about black people having a more expensive radio than the car it's mounted in
17:20:30preglowlike it should be
17:20:31RoriI put some new speakers in my car recently. Sounds amazing. Only snag is the car body vibrates/buzzes lol
17:20:47bill20r3too much cowbell.
17:21:08RoriThese speakers literally kick you
17:21:25RoriYou feel it more than hear it :)
17:21:26*preglow recommends drum and bass for speaker testing
17:21:36Roriaye or breaks
17:21:48t0masa jumping car :P
17:21:49RoriNice bit of breakbeat
17:22:12dapureplayanothing beats my stereo systems in my car. Stock stadnard. hahah
17:22:33RoriI have a line in on my car stereo front for my iRiver ;)
17:22:42t0masthat rocks
17:22:44RoriI use a mobile phone thing to hold it
17:22:57Rorion the air vent
17:23:07bill20r3I finally ordered my first mp3 player last night, an H340
17:23:14*t0mas still hopes it will work for VW's audio system someday
17:23:30dapureplayadamn. I wish i could have that. ATM i'm using crappy FM Trans. Not to mention is a Belkin.
17:23:38t0masbill20r3: and now the waiting for rockbox on H3xx starts ;)
17:23:39Rorithey don't do a CD to line convertor yet. Only the tape cassette ones
17:23:58dapureplayabill20r3 gratz! you rgoing to love video playback on it. the screen on it is super dopper clear.
17:23:59RoriI've had the line in stereo for years and I won't part with it because of it
17:23:59t0masRori: cd's almost impossible
17:24:07bill20r3yeah t0mas, hopefully soon.
17:24:10RoriAwai branded
17:24:15bill20r3anyone here have a 320 or 340?
17:24:21Roriremovable face too
17:24:24bill20r3you dapure, which do you have?
17:24:30RoriI was thinkl
17:24:34RoriI was thinking ahead
17:24:38Roriwhen I bought it years ago
17:24:43dapureplayabill23or3: i have an H140. but i have used my cousin's H340 very extensively.
17:24:58bill20r3how do you like it?
17:25:03bill20r3also, do I need a remote?
17:25:04RoriI knew I would need it for some kinda DAP eventually
17:25:08bill20r3or do I just want one?
17:25:18dapureplayabill20or3: It's great. personally i'm not a rmote user.
17:25:26Roriwhenever I change cars I take it out ;)
17:25:57dapureplayabill20r3: the only downfall is that navigation scrolling thru songs is slightly slower than H140. but that didn't get to me very much, especially when i taught my cousin how to organise his music very well.
17:26:14bill20r3yeah, I'm pretty fussy about organizing stuff, so I should be ok.
17:26:35bill20r3too bad the 3x0 remotes are so much
17:26:50dapureplayaif u had an intl model, you would have had a remote incl.
17:27:07bill20r3I got a US model.
17:27:10dapureplayafiletree is a great way to organise music. i don't like it how ipod organises things for you it's annoying
17:27:20bill20r3what format video does it play?
17:27:51Rorifiletree rules
17:27:59RoriI know where everything is
17:28:14yngwifiltree is cool
17:28:18Roriby category and name and album
17:28:23yngwinever gonna need a database
17:28:31dapureplayabill20r3:avi's. btw, you can only play videos if u flash the intl. firmware. once you get rid of your US firmnware, say bye bye to the DRM capabilities. So if you plan to buy online, you won't have vid playback
17:28:51bill20r3I dont care about drm
17:29:02DJDD_SLEEPfiletree is actually pretty shit when you have around 250 albums.
17:29:03dapureplayagood! all teh more reason to flash 1.28K :)
17:29:25RoriI sort by Genre/Artist/Album/file
17:29:27DJDD_SLEEPMainly because of the slow scrolling speed of the iriver
17:29:38yngwiDJDD thats not true, i have about 300... and it rocks :-)
17:29:42bill20r31.28 Korean is the good firmware?
17:29:45DJDD_SLEEPTHats no help either, i have around 1800 artists
17:29:55bill20r3you can do sub-folders right?
17:30:05bill20r3like \music\A\Artist ?
17:30:11yngwi1800 artists, thats much
17:30:18bill20r3so as long as you knwo where everything is, you wont have speed problems
17:30:28DJDD_SLEEPYes you will
17:30:42DJDD_SLEEPand its hard memorizing 4500 songs
17:30:48dapureplayabill20r3: yeah any intl version of H300 firmware will work. not t he US one.
17:31:06DJDD_SLEEPAt 250 folders, getting to the 150th folder takes a good few minutes
17:31:08B4gderI have >6000 songs in a nicely sorted filetree
17:31:12bill20r3does theusbtogo work with any usb mass-storage device?
17:31:16yngwihmm i haven't had any problems with 3600 songs until now
17:31:25bill20r3ie: could I copy stuff straight off another H340?
17:31:34DJDD_SLEEPDoes the rockbox have a good scrolling thingy?
17:31:37RoriI like being able to switch back to iRiver fw. Means I can still play audio books while the Rockbox speed issue is resolved
17:31:42bill20r3(assuming I modify my US version so the usbtogo actually works)
17:32:11CoCoLUSdjdd, the scrolling accelerates if you scroll longer, so yes, its "good scrolling"
17:32:12RoriiRiver has no bookmark feature though :P
17:32:12dapureplayabill20r3: not too sure how it works on the US modell, but i transferred files fro my H140 to my cousin's H340 vpretty successfully. keep in mind it was USB 1.1 so super slow.
17:32:17 Join ]RowaN[ [0] (a2b0y@
17:32:20Rorior resume when switching on
17:32:33bill20r3what would be *WAY* cool is a rockbox feature for "copy from ipod"
17:32:34bill20r3heh heh
17:32:53Roristrips the DRM as it copies ;)
17:32:54dapureplayairiver is too good for aac :p
17:32:56]RowaN[have u guys thought about having a setting to enable crossfading on next/prev track skip? so when you skip to the next song it crossfades there too?
17:33:09DJDD_SLEEPRori the iriver firmware DOES resume on playback.
17:33:12preglowthat's a bit harder to do
17:33:16RoriRowaN great idea
17:33:16bill20r3if I'm skipping tracks I want it to skip *now*
17:33:33DJDD_SLEEPI agree with Bill
17:33:35RoriI'd like the option :)
17:33:48]RowaN[bill: thats why i say make it an option =]
17:34:02]RowaN[as long as the crossfade is short it wont be annoying, just pleasant
17:34:23RoriGimmie some generic gapless (Something close is all I ask) for any MP3 and I will be happy though
17:34:26bill20r3are there any US sources for the H3x0 remote?
17:34:29bill20r3or just ebay?
17:34:35]RowaN[not even crossfade.. just fade the 1st track out for a second and have the next one start at full volume over the top
17:34:51bill20r3I think that'd be of minimal use.
17:35:04RoriLame gapless will be a start
17:35:07dapureplaya were selling them but they sold out in a snap. you can try asking the store if they plan to stock anymore.
17:35:18preglowlame gapless will be a start?
17:35:30preglowi don't want any more than lame gapless
17:35:34Roriyeah once Slasheri or someone figures it out
17:35:40bill20r3whic model is this remote from?
17:35:53RoriWell I never rule out an option if it's possible :)
17:36:04dapureplayabill20r3: H100 remote.
17:36:16***Saving seen data "./dancer.seen"
17:36:19ashridahyeah, H1x0 remote
17:36:36Roribut not going to get into the gapless argument again...for now ;)
17:37:01bill20r3dapure, do you know which remotes work w/ the 340?
17:37:02*t0mas np: Freestylers - Push Up (0:34) []
17:37:06]RowaN[how does crossfading work atm? i havent managed to hear it yet (yes ive turned it on). Does the next song fade in over the top, or does it just start at normal volume over the last track (which is fading)?
17:37:13t0masRori, dapureplaya: that's a really good one :P
17:37:17dapureplayabill20r3: you can also use it for H300 aswell. but there is a special H300 remote which has a joystick on it.
17:37:32Roriwhat is dapureplaya?
17:37:42 Quit B4gder ("Lämnar")
17:37:56bill20r3of course I'd rather have the 'proper' remote, but the 100 ones seem a lot easier to find
17:37:59dapureplayaoh yeah. Push up is a good one heheh
17:38:10Rorioh lol
17:38:23dapureplayabill20r3: good luck if u can find the H300 one.
17:38:25Rorisilly me I thought you was referring to some plugin :D
17:38:31dapureplayabill20r3: the H100 remote works fine
17:38:56RoriFreestylers are cool BTW
17:39:08dapureplayai only ever heard one of their songs and that was it hehe
17:39:52preglow]RowaN[: the first track fades out, the second track fades in
17:39:54]RowaN[bill20r3: i think it would be a nice touch - atm i use dare i say it, winamp, and with directaudio plugin, when you hit next track, just for not even a second, the current track fades out instead of abruptly stopping, and the next on starts with no delay
17:40:07]RowaN[preg: at the same time?
17:40:17]RowaN[over what period of time?
17:40:19preglowit wouldn't be crossfading if not
17:40:28preglowwhy not just try it?
17:40:41]RowaN[like i said, i have
17:40:49]RowaN[and cant manage to hear 2 tracks playing at once at any point
17:41:25 Join Musicmad_ [0] (
17:42:00]RowaN[the wps screen changes to the next mp3 a little early, but the next track seems to just start as normal.. maybe its just me, i'll have to try some other mp3s tonite
17:42:30preglowhave you enabled it?
17:42:38 Quit Musicmad_ (Client Quit)
17:42:45 Join Cassandra [0] (
17:42:52]RowaN[since it became on option yesterday or whenever it was
17:43:09]RowaN[ah maybe i need to reenable it again coz i updated rockbox a few times in the last few days
17:43:58CassandraI'm getting a bug in iriver playback where occasionally Rockbox drops back to the WPS at the start of a track. Is this known about?
17:44:23Cassandra(On the positive side, recent commits have fixed all the other flakyness)
17:44:24]RowaN[but that aside, what do you think about the next/prev cross.. well not crossfading, just current track fading out (for like half a sec) and the next one starts with no delay (so you hear it immediately, as the 1st track is still fading)
17:45:20CassandraThis is with shuffle on. I don't know if that makes a difference.
17:45:33bill20r3jesus those 340 remotes are expensive
17:45:38]RowaN[it mite sound weird but its much easier on the ears to have a track not end abruptly when you skip to the next, trust me
17:45:42 Join MO-Pantsu [0] (
17:46:01preglowCassandra: can't say i've noticed it
17:46:05]RowaN[infact a friend of mine said "wow" when he heard my winamp do it =p
17:46:08 Join webguest43 [0] (
17:47:12 Quit Rori (Read error: 104 (Connection reset by peer))
17:47:38bill20r3is there a side-by-side comparison of the different remotes anywhere?
17:47:50CassandraI seem to be able to reliably reproduce it. Make a root.m3u, enable shuffle, start it off, and press the next button a few times. Eventually you will drop backt to the WPS.
17:48:55CassandraPersonally I hate crossfading. If you have a DJ selecting tracks it works, but some tracks do not crossfade well into each other.
17:48:57]RowaN[hmm still cant seem to experience crossfade here.. contineous mix cd in lame mp3 format listening to i am (whoops got all yodaey)
17:49:31]RowaN[yes but for a 1 sec crossfade, it doesnt need a dj so much
17:49:47preglowi'll never use crossfading myself
17:50:03webguest43there should be plugins for sound adjustment on iriverbox
17:50:05]RowaN[im against crossfading... just fade the old track OUT, no need to fade the new one in, just have it start full volume
17:50:06Cassandra(Oh, as another aside, gapless works fine.)
17:50:18preglowCassandra: not for mp3s, no?
17:50:37]RowaN[gapless seems find (dont remember enabling it, maybe its on by default) but still cant hear crossfading. any ideas preglow?
17:50:46CassandraFunny I could have sworn it was working for my copy of The Wall.
17:51:24 Nick Lynx_ is now known as Lynx_awy (Lynx@
17:51:38preglowi think it's a coincidental thing at the moment
17:51:40CassandraI mean obviously I ripped it splitting on frame boundaries.
17:51:53preglowahh, −−nogap ?=
17:52:01CassandraWouldn't expect it to work otherwise.
17:52:04preglowthen it should work just dandy, yes
17:52:28preglowthe more proper method using the LAME info header doesn't work yet
17:52:44CassandraAh - didn't know about that.
17:53:11preglowyou don't need to pad out the last frame with next track data then
17:53:38CassandraThat'd be kind of cool. Saves having to rip things in a special way.
17:54:02preglowand keeps track transitions the way they're supposed to be
17:54:24CassandraSo, can anyone else reproduce this dropping to WPS bug?
17:56:21CassandraHmm. I'm also getting a bug where sometimes, the track buffers correctly but freezes at 0:00.
17:56:31preglowthere are still bugs left no doubt
17:56:51preglowgotta go
17:56:56 Part preglow
17:57:08]RowaN[is anything supposed to happen when you click the iriver joystick during playback/wps?
17:57:27]RowaN[in original firmware it goes back to the file list
17:57:52dapureplayahey is there a way to show which songs are of what filetypes while in browser mode?
17:58:03yngwiIt should do that with rockbox too, at least it does it for me
17:58:50]RowaN[ah yep, working for me now too
17:59:04CassandraAh, the second one is a side effect of the first. After it's dropped back to the WPS, playback always halts at 0:00
18:05:02]RowaN[after i press stop, and the song fades out, there is then a little blip of sound, does everyone else get that too?
18:07:55 Join linuxstb [0] (
18:08:03yngwiRowan: do you mean the song gets played shortly or do you mean a noise?
18:15:15 Quit webguest43 ("CGI:IRC (Ping timeout)")
18:27:20]RowaN[blip of the song u were just heading fade, but at normal volume
18:28:32yngwido you use the bleeding edge build?
18:28:58 Join Tomhab [0] (~tom@
18:29:48 Quit ]RowaN[ ("back in 2 hours")
18:30:30yngwihmm i had that too, but not anymore i thought, but i'll try
18:32:21[solid]i had that for ages
18:32:22yngwihmm no, that doesn't happen anymore, do you have crossfade on?
18:32:38[solid]but i didn't test explicitely for it with the latest build, no
18:33:37yngwii remember, that i had it and i don't have it anymore, so it must be the latest build or some settings, and i remember turning crossfade of so it was on by default
18:34:19 Quit ashridah (Read error: 110 (Connection timed out))
18:35:20yngwino, even with crossfade on, it is gone,... so i don't have a clue
18:35:54amiconnHmm, "fun" :/
18:36:11 Quit Cassandra (Read error: 60 (Operation timed out))
18:39:27amiconnThe rockbox scheduler does only work on archos because there is the pr register.
18:42:42amiconn...wrong. There's a specially crafted thing in load_context()
18:43:10thegeekI got that song blip too
18:43:15thegeekusing a build from late last night
18:43:21thegeekit only happens sometimes
18:43:32thegeekdid not spend time finding the trigger
18:43:40 Quit thegeek (Read error: 54 (Connection reset by peer))
18:45:15 Quit dapureplaya (Read error: 110 (Connection timed out))
18:52:00linuxstbI'm in the process of writing codecwav - does anyone know how much data I should be passing to the audiobuffer_insert() function in one call?
18:52:48 Quit DMJC ("Leaving")
18:53:16 Quit zezayer (Remote closed the connection)
18:53:48 Quit Tomhab (Read error: 60 (Operation timed out))
18:53:53yngwiRowan: I've got it, the "Fade on Stop/Pause" is the problem, something there is not working, try switching it off
19:04:16 Quit [solid] ("Lost terminal")
19:07:29 Quit tedboer ("Leaving")
19:08:01 Join DMJC-L [0] (
19:10:07 Quit DMJC-L (Client Quit)
19:33:48Slasherilinuxstb: There is no limit
19:35:51Slasherilinuxstb: So feel free to pass as much data as you need to. But maybe you should call yield sometimes, so try to limit the data to a few kilobytes. You should experiment which values gives the best ui and overall performance
19:36:18***Saving seen data "./dancer.seen"
19:36:55SlasheriNow iriver seems to be able to play even 500 kbit/s vorbis flawlessly
19:40:21linuxstbSlasheri: How many bytes would you suggest my WAV codec reads/writes in one go?
19:40:41linuxstbIs 4096 bytes sensible?
19:41:48 Join DrMouss [0] (
19:42:04 Join thegeek [0] (
19:44:47 Quit Chamois (" Try HydraIRC -> <-")
19:50:59 Quit MoosCamaro (Read error: 60 (Operation timed out))
19:55:28Slasherilinuxstb: yes, it's quite good i think. And btw, there is limit but it's very big
19:56:24amiconnYay! The modified scheduler for iriver is working, including stack check :-)
19:56:26linuxstbI tried 4096 bytes, but it wasn't sounding correct - increasing to 16K fixed the problem.
20:00:49Slasheriamiconn: oh, sounds good :)
20:04:05linuxstbSlasheri: Should I copy your two new ci->configure lines from codecflac into codecwav? Do you think I should increase the values?
20:04:54Slasherilinuxstb: You should determine the best values by testing :)
20:05:09linuxstb:-) I'll just copy the FLAC values for now then...
20:08:50linuxstbOK, I've just committed codecwav.c to CVS.
20:09:42bill20r3what format video do the iRivers play?
20:09:55Slasherilinuxstb: great :)
20:10:03 Nick DrMouss is now known as MoosCamaro (
20:13:04linuxstbbill20r3: Rockbox doesn't play videos on the iRivers
20:14:04bill20r3I mean on the iriver firmware, for the H3x0's
20:14:27linuxstbSorry, I have no idea.
20:16:13bill20r3I just ordered an H340
20:17:42Slasherilinuxstb: btw, instead of request_filebuf you should use request_buffer so there is no need to copy data twice
20:17:43 Join zezayer [0] (~chatzilla@
20:18:10nobbehi have a tag related w
20:18:27nobbehany change of ape tags on mp3s being supported?
20:20:08linuxstbSlasheri: OK, I'll change that.
20:22:01Slasherilinuxstb: Then you could also use short int as pointer type so byteswapping would be easier (and probably faster). #define BITSWAP(x) (((x>>8) & 0xff) | ((x<<8) & 0xff00)
20:22:52linuxstbDo I need to call advance_buffer as well?
20:23:05 Join elinenbe [0] (~elinenbe_@
20:23:26Slasheriyes you need, after you have done with the data
20:24:51linuxstbI'm looking at codecmpa - I don't understand how it works if the MP3 file is larger than the memory buffer.
20:25:28SlasheriIt requests small chunks same way read_filebuf does
20:25:38linuxstbBut it doesn't advance the pointer?
20:25:56Slasheriit will, at the end of the loop
20:26:27Slasheriit uses advance_buffer_loc but you should use advance_buffer(amount)
20:26:27linuxstbOK - I missed that.
20:27:54Slasheriafter calling something like that: short int pointer = ci->request_buffer(&size, CHUNK_SIZE); the size will tell you have many bytes was read
20:28:08linuxstbSo it's just request_buffer, byte-swap, audiobuffer_insert, advance_buffer
20:28:11Slasheriif size is zero (or return value NULL), it's end of file
20:28:16Slasheriyep :)
20:30:37amiconnPlease don't define the byte swap as BITSWAP, this will be *very* confusing in rockbox
20:30:38Slasheriand remember not to request too large buffer sizes (maximum allowed amount is GUARD_BUFSIZE defined in playback.c, 8 KiB, but you should use lower values still
20:31:19Slasheriamiconn: That was only a silly example, i meant BYTESWAP or something like that =)
20:32:55linuxstbSlasheri: I have problems processing chunks smaller than 16K - try changing the size of wavbuf in my current codecwav to (e.g.) 4K, and listen to a WAV file. There is very bad stuttering.
20:33:20Slasherilinuxstb: Hmm, maybe you have too many yields then
20:33:25SlasheriTry decreasing those
20:33:44linuxstbOK. Too many changes to do at once....
20:33:48Slasherior add a counter if necessary
20:38:14*t0mas np: Sound Effects - Thx - Dolby Dts - Ultimate Bass Test(Techno 320) (0:04) []
20:39:10zezayerdoes anyone else find teh latest bleeding edge doesnt play mp3's
20:41:03Slasherilinuxstb: You have forgot to update SOURCES, your codec wont compile..
20:43:01amiconnlinuxstb: It seems the .wav bitrate is still a bit off. Afaik the bitrate is given in kBit/s, not KBit/s... it should be /(1000/8) not /(1024/8)
20:43:36linuxstbSlasheri: Sorry, I'll commit it now.
20:43:42Slasherilinuxstb: great :)
20:43:53Slasheriamiconn: KBit.. what's that? :)
20:44:02linuxstbamiconn: Who made that rule? :-)
20:44:03SlasheriKelvin Bits.. sorry :D
20:44:38linuxstbSOURCES is updated now.
20:44:43amiconnlinuxstb: I don't know, but it's the way it works with mpeg audio
20:45:28amiconnSlasheri: K == 1024, k == 1000
20:45:35linuxstbOK, I'll change the bitrate. Bloody computer scientists.
20:45:55Slasheriamiconn: Hmm, in europe we use SI-units. So k = 1000, Ki = 1024 and K = Kelvin =)
20:46:20amiconnSlasheri: I know that you use Ki for that, however, your commits were the first thing I've seen using that
20:46:30Slasherioh :)
20:46:41amiconn...and I *am* in europe :)
20:47:31Slasheriok, that Ki, Mi etc. are still quite rarely used
20:47:33linuxstbI can't believe I'm sitting here optimising the WAV codec....
20:48:15Slasheriamiconn: This site has more explanation on that:
20:48:25amiconnSlasheri: ...and btw, those k, M, Ki, Mi etc aren't units, but prefixes
20:48:37Slasheriah, yes.. that's true :)
20:49:03amiconnOr would you think 'mm' means 'meter meter'? ;-)
20:50:02amiconnThose Ki, Mi etc look a bit strange imho
20:50:18amiconnKi might be okay, but Mi? 'Migabyte'??
20:50:48HCllinuxstb: does it run at 11mhz yet?
20:51:28yngwiI'm in europe too, have been going to a technical school and have never heard of Ki, ect... :-)
20:51:47yngwibut I'm not what you'd call a good student :-)
20:51:50Slasheriyngwi: Hmm, really weird :)
20:52:51linuxstbHCl: I don't know yet - it's not working reliably ATM.
20:53:21yngwihehe taking out a pin to reset... :-)
20:54:03SlasheriHmm, you should use some paperclip.. pins are known to be harmful to the reset button
20:54:39yngwioh, haven't thought of that, thanks
20:55:57yngwia good tip, its also working better... again, thanks
20:55:58SlasheriWhat i have read from misticforums, some people did use pins and had broken reset button into pieces after that.
20:56:03Slasherigood :)
20:56:18yngwihmm, thats bad
20:56:31yngwidon't want to hurt my baby..
20:56:53zezayerhave done that personally
20:57:00zezayermakes it a pain 2 reset now
20:58:45yngwithe progress is awesome
20:58:54yngwiyou are doing a great job
20:59:21thegeekhow can anyone actually put the sharp end of a pin in there?
20:59:22thegeekI mean
20:59:37thegeekit's pretty obvious it's not good sticking an object that sharp into there
20:59:51yngwiyeah thats true, shame on me :-(
21:04:39 Quit xen` ()
21:06:49 Join ghostiger [0] (~ghostiger@2a4bd8d6dfc3474b.session.tor)
21:10:11yngwiis there a way to turn off gapless playback for mp3??
21:10:25yngwihaven't found it in the menu
21:13:10CoCoLUSwhy would you want that?
21:13:32yngwibecaus it seems to shave of the end of many songs...
21:13:37linuxstbOK, I've updated codecwav.c to use ci->request_buffer, but I can't prevent it stuttering unless I process chunks of 16K at a time.
21:13:42linuxstbCan someone look at it?
21:18:42Slasherilinuxstb: i will :)
21:19:26 Quit Hadaka ("poof")
21:21:39yngwiind i can't remember that I have any album that needs gapless (except NOFX-Ribbed :-) )
21:23:07linuxstbyngwi: "gapless" (at the moment) just means that Rockbox doesn't insert any silence between tracks. So if you are losing the end of a track, it's a bug.
21:24:43yngwii could only remember having read about cutting parts from the end of the file, but i guess that's where my memory starts to cheat on me... :-)
21:25:17yngwiI'm gonna try this out some more..
21:25:54linuxstbMake sure you're using the very latest "bleeding edge" build - there's been lots of bugfixing today and last night.
21:26:39Slasherilinuxstb: Check my commit :)
21:28:06Slasherilinuxstb: btw, thanks to your "shuttering" code i might be able to locate one bug that will cause a crash :)
21:28:20yngwiI'm "always" using the newest build, i'ts very exciting :-) in fact, I sat the whole day in front of the PC pretending to study Latin but checking the IRC and build Page very often....
21:28:33SlasheriHehe :D
21:28:49yngwiits thrilling... what will they improve with the next build.... *sigh*
21:29:22yngwimaybe I'm kinda weird...
21:29:29linuxstbSlasheri: I don't understand your changes, but it's working perfectly. Glad I could help!
21:29:42Slasherilinuxstb: I think the filebuffer chunk size was too small
21:29:49SlasheriI increased it and it's now working :)
21:30:31linuxstbSlasheri: What are you planning on doing next? What about seeking in FLAC/OGG/WAV ?
21:30:45 Join Naked [0] (
21:30:53Slasherilinuxstb: I want to find the remaining bugs that will cause rockbox to crash..
21:31:08linuxstbOf course!
21:31:40amiconnSlasheri: Crashes might very well be caused by stack overflows
21:31:48linuxstbI may regret it, but I'll try to look inside libFLAC and see if I can make some more use of the IRAM to speed it up.
21:31:51Slasheriamiconn: Hmm, indeed
21:31:59amiconnStack usage is at 80% just by playing mp3s
21:32:51amiconnThat's why I want to have the stack overflow check in the scheduler working
21:33:22amiconnA stack overflow will then trigger a panic on the next context switch, and rockbox will print the offending thread's name
21:33:41amiconnI already have it working, just need to do some more tests
21:33:49Slasheriah, thats really good :)
21:36:19***Saving seen data "./dancer.seen"
21:40:52 Nick Naked is now known as Hadaka (
21:43:33 Quit zezayer ("Chatzilla [SUSE 1.0.4-1.1/20050511]")
21:48:42 Join Tangleding [0] (
21:49:08TangledingNice to see latest progress on wave too...
21:49:43MoosCamarohi Tang
21:49:49TangledingHi Moos
21:49:59amiconnSlasheri: mp3 playback seems to need the largest stack space
21:49:59MoosCamarolinuxstb work on it ;)
21:50:08amiconnStill no overflow though
21:50:10Tangledinglittle thing to edit in the wiki codec status
21:50:17Tangleding" WAV Plays from the debug menu (first 8mb of /sample.wav)"
21:50:29Tangledingthe prenthesis seems useless now no?
21:50:56elinenbeamiconn: do you have an iriver now?
21:51:00amiconnEverything after 'Play' should be useless
21:51:05amiconnelinenbe: yup
21:51:38Tangledinglol indeed
21:51:51yngwisorry to be off topic, but do you (wherever you are) also have stupid movie translators?? (feel free to tell me if you don't want that kind of talk on the channel)
21:51:52Tangledinggood shot amiconnn
21:52:02Tangledingmikaa seems to edit actualy
21:52:34amiconnSlasheri: flac needs ~12% of the codec stack, vorbis ~17%. mp3 needs ~80%
21:53:01amiconnNow I need to forge a stack overflow condition...
22:00:48Slasherilinuxstb: your code was great, i was able to find one crashing bug using it :)
22:02:26linuxstbDo you know how many are left?
22:03:09Slasherino i don't but i believe this will fix many problems
22:03:30Slasheri(i think there is second bug somewhere too, now searching it)
22:09:16 Join bipak_ [0] (
22:09:30*amiconn scratches head
22:10:28*t0mas has a new way to clean his desk:
22:10:30*t0mas np: Sound Effects - Thx - Dolby Dts - Ultimate Bass Test(Techno 320) (0:30) []
22:10:46 Join Rori [0] (
22:11:31 Quit MO-Pantsu (Read error: 104 (Connection reset by peer))
22:11:54Tangledinganyone has some news of BobTHC?
22:14:47Tangledingstange bug with latest building edge
22:15:07Tangledingthe playback was cut
22:15:27nobbehany chance of ape tags on mp3s being supported?
22:15:29 Quit bipak_ (Read error: 54 (Connection reset by peer))
22:15:58Tangledingsince RG info are stored in ape tag
22:16:07Tangledingand RG are planned to be supported
22:16:15Tangledingi think the answer is YES, no?
22:16:21 Join bipak_ [0] (
22:16:32linuxstbThe isn't no.
22:16:41linuxstb^The answer isn't no.
22:17:45MoosCamarodepand if someone work on it ;)
22:18:11Tangledingmaybe nobbeh is volunteer? :D
22:18:21HClwhats rg info?
22:18:21Tangleding(just joking)
22:18:26linuxstbI could work on it, but I never tag any my files...
22:18:39Tangledingi think it work from APe tage
22:18:53Tangledinggonna check if i didn't mistaken
22:20:24amiconn.wav shouldn't need to boost the cpu at all, correct?
22:23:37linuxstbamiconn: I would hope not. All it does is byte-swap the data.
22:24:11amiconnHmm. Playing a .wav still boosts the cpu while reading from disk
22:24:20 Quit bipak (Read error: 110 (Connection timed out))
22:24:43linuxstbHow can I check the CPU status?
22:24:59 Join preglow [0] (
22:25:39Slashericurrent implementation always boosts cpu while reading from disk..
22:25:47Slasherilinuxstb: debug menu :)
22:25:57preglowboosts cpu while reading from disk???
22:26:05preglowhave anyone thought about getting ata dma to work?
22:26:31Slasheripreglow: yep, that's mainly to speed-up disk reading
22:26:58Slasherihowever, dynamic cpu_boost could do it also (haven't tested yet)
22:28:59amiconnpreglow: mp3 uses the most of all codecs by far
22:29:51Slashericvs updated, at least one crashing problem fixed!
22:31:38preglowamiconn: not more than tremor, i'd wager
22:31:49amiconn*way* more
22:31:49preglowamiconn: or what do you mean?
22:31:53preglowmore cou?
22:31:54preglowcpu <-
22:31:59amiconnStack space
22:32:02preglowahh, yes
22:32:03preglowthat i know
22:32:10preglowtake a look at III_decode
22:32:18preglowit uses about 8k of stack space for temporary samples
22:32:28amiconnHmz, the most important word was missing in my sentence :-/
22:32:59preglowmno, 4k
22:33:08preglowhow much stack space usage have you seen?
22:33:23amiconn[21:52:11] <amiconn> Slasheri: flac needs ~12% of the codec stack, vorbis ~17%. mp3 needs ~80%
22:33:32amiconnI can add mp2: ~19 % now
22:33:44 Quit bipak_ ("Lost terminal")
22:34:24amiconnCodec stack size is 10.25 KB
22:34:41preglowbut anywho
22:34:45preglowit's needed for performance
22:34:50preglowso might as well let it use that stack
22:35:58amiconnThe stkov check is working now (not yet committed), but for obvious reasons it can't catch all cases
22:36:21amiconnIf some crucial code gets overwritten by the stack overflow....
22:36:22 Join Jason [0] (
22:37:02amiconnI was able to cause a 'stkov audio' but not a 'stkov codec'. Rockbox simply hung
22:37:12amiconn...probably because the codec stack is in iram
22:37:30JasonIm just trying out Crossfade
22:38:04preglowamiconn: exactly how does stkov work?
22:38:07SlasheriJason: currently it's broken
22:38:24Slasheri(at least i think it is)
22:38:25Jasonwell its doing something on my iRiver
22:38:33Slasheriah, ok :)
22:38:33Jasonwhat would be good is a variable crossfader setting
22:38:54elinenbeSlasheri: you haven't heard that one yet <grin>
22:39:17SlasheriHmm :D
22:39:17amiconnpreglow: switch_thread() checks whether the first longword is != 0xdeadbeef. If it is -> panicf()
22:39:18Jasonthe only thing iRiver Rockbox doesn't have is sound adjustment
22:39:43amiconnJason: There are many more things missing
22:39:49Jasonas well as a clock
22:39:56amiconnIt's far from complete
22:40:21Jasonwell i now use Rockbox simply because the shuffle is a million times better than iRiver
22:40:42*bill20r3 shops for a remote
22:41:02Jasonmiss that a lot decent shuffle
22:41:39 Join matsl [0] (
22:41:41TangledingIt's not finalised thought
22:41:52Jasonwell its somethign
22:41:56Tangledingbut too i use Rbx since i have some mp3 taht iRiver fw can't play
22:41:58bill20r3is there just the 2 models, the common 1x0 remote, and the unobtainable H3x0 remote?
22:42:01Tangledingand Rockbox does
22:42:13Tangledingyes but
22:42:22Tangledingfor H3xx only the two work
22:42:32 Quit ghostiger (Remote closed the connection)
22:42:34Tangledingfor H1xx the H3xx remote won't work fine (no buttons)
22:42:41Jasonbut I must admit iRiverbox looks promising
22:42:50bill20r3so which ones work on the H3x0?
22:43:16JasonI reckon the H1x0 series are gonna be very collectable
22:43:50 Join ehntoo [0] (
22:44:20Tangledingboth H1xx and H3xx work for H3xx
22:45:25amiconnMrf, I don't get this
22:45:59bill20r3are there different remotes within the H1xx models, or are they all identical?
22:46:09bill20r3seems like some pictures are chrome, and some are black.
22:46:11Tangledingal; identical
22:46:15Tangledingand you can even buy
22:46:19TangledingiMP550 remote
22:46:31Tangledingit's the same only the names of the button change
22:46:52Tangledingi own an iMP550 remote in replacement of my iHP1xx old remote that i broke
22:46:54bill20r3is there any reason to get the iMP500 remote over one of the H1xx remotes?
22:46:57Tangledingit work fine
22:47:02Tangleding(on my iHP1xx)
22:47:13bill20r3are they different colors or anything?
22:47:22bill20r3cause I'd like one that's black, if possible.
22:47:29Tangledingno more reason but sometimes iHP1xx remote aren't available and iMP550 are
22:48:02JasonIt would be awesome if iRiverbox supports the FM tuner and enables recording from it
22:48:03amiconnIt might be that there are H-1x0 remotes with different colours. Mine is black
22:48:40Tangledingthere is black for US H120 and iHP140
22:48:53Tangledingand champaign for UE iHP120
22:49:06Tangledingbut also silver for iMP550 and black for iMP550 too
22:49:20Tangledinganyway the only différence is the button labeling
22:50:35MoosCamaroand black for h140 too ;)
22:55:22preglowamiconn: but how does it check stack usage? writes 0xdeadbeef as far as it can and sees how far it's been erased? :P
22:55:49amiconncreate_thread fills the whole stack of the new thread with 0xdeadbeef
22:56:07preglowahh, yes
22:56:14amiconnThe check in switch_thread is a quick check for overflow
22:56:33 Quit silencer_ ("Lost terminal")
22:56:36preglowbut yeah, no surprise libmad needs stack, slasheris first attempts at audio crashed because of this
22:56:44amiconnThe values in the debug menu are calculated by thread_stack_usage(),
22:57:05preglowamiconn: are you gonna commit the working stack check soon? it's a nice tool to have
22:57:17amiconn..which will scan each thread's stack until it finds something else than 0xdeadbeef
22:58:14amiconnI would have done that already if there wouldn't be a problem
22:58:54amiconnI've unified the scheduler as much as possible for coldfire and sh1.
22:59:19amiconnNow I'm trying to trigger stkov on both. see whether it displays the correct thread name
22:59:47amiconnOn archos, I did not yet manage to do that
22:59:50 Join telliott [0] (
23:00:12amiconnI might have to do with build dependencies; trying a full rebuild now
23:00:34yngwiamiconn: my remote is also black
23:03:59amiconnSomething is seriously wrong here
23:04:04*amiconn is puzzled
23:04:37 Part telliott
23:12:48Rickwhat do you mean
23:13:32 Join road_runner [0] (
23:14:31road_runnerI've just posted a question at the website, now this is kinda silly right? I can ask you guys... i'm pasteing it.
23:15:04road_runnerHello. In a few minuts i'm going to (finally) install rockbox as now it plays music and I must try it...
23:15:05road_runnerI'm using hebrew on the original iriver (1.65u) firmware - by altering the cyrillic font by h3mod to the hebrew letters. this has worked o.k. for me - surely with the mirrored problem.
23:15:05road_runnerIn order to install the bootloader I would like to use the fwpatcher. And so I get a checksum error if I use it on an already moded .hex firmware file.
23:15:09road_runnerI got away with this by firstly patching the firmware with the bootloader and only then h3mod'ing it to include the hebrew charset.
23:15:14road_runnerWayyyy too afraid of trying this myself, I have made a resolution to wait until u guys will tell me if this is safe or not...
23:15:19road_runnerp/s, if the fwpatcher program wouldn't do a checksum check, (would descrambling-merging-rescrambling it myself = wouldn't produce the same error) would lead to a corrupted firmware? b.t.w. I guess not.
23:15:22road_runner(hey it worked)
23:15:53BagderI'm not sure what you're asking
23:16:01road_runner(and if someone has a bdm interface I would love to hand him my firmware so he could check if it really works...)
23:16:08linuxstbHas anyone done any work on the AC3/A52 decoder recently? I've just tested it, and it's decoding at 130% realtime on one test file, and 115% on the other.
23:16:18preglowroad_runner: you can only patch an unmodified firmware
23:16:34preglowroad_runner: with fwpatcher, that is
23:16:50preglowlinuxstb: no work whatsoever
23:16:58preglowlinuxstb: since i have little or no interest in the format
23:17:15linuxstbIt seems no work is necessary.
23:17:28nobbehroad_runner:theoretically, if only the fonts are replaced, you should be ok.
23:17:33preglownecessary, no, but still needs to be done, heh
23:17:41linuxstbThe only use for AC3 is for digital TV or DVD rips.
23:18:31road_runnerpreglow: how can I be sure, i already crashed the player 5 (that is FIVE!!!) times until warrenty voides
23:18:38 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
23:18:51preglowyou can't be sure
23:19:04preglowbut the bootloader doesn't touch those parts of the firmware, I THINK
23:19:17preglowbut no guarantees
23:19:36 Join yngwi [0] (
23:20:12road_runnerwell I DO KNOW that using h3mod would produce good altered firmwares... question is if this is the right way, or scrambling it myself :?
23:20:34Bagderyou'll get the same end result I guess
23:20:49road_runnerthe key word is "guess"
23:20:56preglowbut hell, why do you need iriver firmware when you've got rockbox anyway? :>
23:22:19road_runnerhehe.. because everyone warned me I would have to start carring pins with me everywhere
23:22:49Bagderthe pin is very light ;-)
23:22:53road_runnerheeee... lol
23:22:55preglowhardly takes any space at all
23:23:10Bagderhad to use mine on the train back from work today ;-)
23:23:12road_runneru r the 1st guys how acctually told me to try it
23:24:18road_runneranyways... can someone acctually try C if my firmware works 4 him or would this be 2 much 2 ask 4?
23:24:45Tangledingwhat the wquestion?
23:25:17TangledingYou want to know if you can change the fw bitmaps with H3mod?
23:25:17Bagderso you're asking someone else to risk what you don't dare?
23:25:25road_runnercan someone try my patched 1.65U firmware with hebrew charset and bootloader... ?
23:26:09TangledingAsk Linus if he has BDM
23:26:09road_runnerI'm asking if someone in this forum has a bdm inteface - > can risk his own player because he can always reflash it!!!
23:26:24Tangledinganyway you know
23:26:27road_runnerLinus surely has bdm
23:26:29Slasherii have changed bitmaps from my 1.65E firmware, patched it with bootloader after that and flashed. Works fine
23:26:29BagderLinus is the only one to has one
23:26:34Bagderwho has
23:26:38Tangledingmaybe Linus has betetr to do with Rockbox
23:26:44Tangledingis this very important?
23:27:01BagderI say run without the font patch
23:27:19Bagderor cross your fingers and run for it
23:27:32road_runnerslasheri: modding before patching? didn't it give you a checksum error? have you used the fwpatecher? probebly not.
23:27:49Slasheriroad_runner: i patched it on linux using command line tools
23:28:09Tangledingif you want to know if we can mod a iHP fw with bootloader using H3mod
23:28:12Tangledingthe answer is YES
23:28:25Tangledingi've done it for myself
23:28:33road_runnertangleding: yeah very. half of my playlist is in hebrew and i don't like reflashing, crashed way too much for me
23:28:37Tangledingi just do'nt understand what's the matter
23:29:06amiconnUgh, severe bug found
23:29:08road_runnerI'll re explain myslef
23:29:15Tangledingin private
23:29:24Slasheriamiconn: Hmm, from cvs code?
23:29:24amiconnI wonder how this could have worked...
23:29:36amiconnSlasheri: Nope, in my reworked scheduler
23:29:39Slasheriah :)
23:29:59amiconnI was mixing up some registers in the context load on archos
23:30:15Slashericvs has still one unsolved crash problem (I04: something)
23:30:21amiconnIt still worked, just not always
23:30:35 Join Nuxator [0] (
23:30:53amiconnI04 is illegal instruction
23:31:04Nuxatorany dev here? i've got a problem to compil rockbox
23:31:22 Join pbvas [0] (
23:31:24Nuxatorsrc/it/itread.c: In function `it_load_sigdata'
23:31:39Nuxatorit seems that dumb doesn't compile
23:31:57Bagderyour first build?
23:32:11BagderI mean, did it ever work?
23:32:21Nuxatori build rockbox ong ago without dumb in cbvs
23:32:36Bagderwhat gcc version?
23:32:38Nuxatorand under cigwin nom i'm in linux
23:32:50Nuxatorbut i've got the same error in cygwin and linux
23:33:09Bagderlatest cvs?
23:33:11crwlhm, queuing doesn't seem to work very well currently, does it? :)
23:33:12Nuxatorgcc from linux or gcc m68k-elf
23:33:32Bagder"gcc −−version" shows
23:33:34Nuxatoryes latest cvs
23:34:16Bagderwell, it builds fine for me and numerous others
23:34:17crwllooks like if i play something and then queue, say, an album full of songs, then rockbox will still play all the songs it had buffered before moving on to the queued tunes
23:34:37BagderNuxator: did you create a new build dir and ran configure in there now with the recent code?
23:34:39yngwicrwl: i noticed that too
23:34:44crwland when it's playing those buffered songs, the wps correctly says what it's playing, but the playlist window (which now works) claims it's playing the queued song it's supposed to play
23:34:54Nuxatordumb doesn't compile: itread.c has errors
23:35:06NuxatorBadger : yes
23:35:09amiconnAhahhhh! 'Stkov mpeg'
23:35:14Bagderit compiles for me
23:35:19amiconnIt works on archos as well :-)
23:35:30crwlSlasheri, have you noticed what i reported?
23:35:36Nuxatorerror: insn does not satisfy its constraints
23:35:57BagderNuxator: what binutils version?
23:36:04Nuxatori wanted just to fix fasts forwrd rewind for remote
23:36:10Slashericrwl: Hmm, i haven't tried that
23:36:21Nuxatorbinutils how i know version?
23:36:23***Saving seen data "./dancer.seen"
23:36:40crwlSlasheri, it seems to be easily... öö, toistettavissa ;D
23:36:56BagderNuxator: how did you install binutils for m68k-elf then?
23:37:01amiconnNuxator: m68k-elf-as −−version
23:37:07Nuxatorfrom documezntation
23:37:08Slashericrwl: hehe =)
23:37:19Nuxatorsame version has on webbbbbsite
23:37:21Slashericrwl: Okay, i will try that tomorrow :)
23:37:25crwlSlasheri, ok :)
23:37:35 Join Zoom2 [0] (
23:37:53Nuxatorand m68k-gcc-elf 3.3.4
23:37:58Bagderthat error sounds like a bad binutils version error to me
23:38:27amiconnNuxator: Try to build or install gcc 3.4.x for m68k
23:38:46 Join Shari [0] (
23:38:52 Part Shari
23:39:50Nuxatorpfff just all this workd just to add 2 define in one file
23:40:57Nuxatorwich version do you have?
23:41:14Bagderm68k-elf-gcc (GCC) 3.4.3
23:41:34amiconn3.4.4 here
23:41:37BagderGNU assembler 2.15.94 20050202 (with coldfire patch)
23:41:57amiconn..and binutils 2.16
23:43:10Nuxatorit's strange that i had same prob under cygwin
23:43:29amiconnDid you use the precompiled m68k-elf-gcc?
23:44:20Nuxatori followed the page on rockbox website
23:44:25Nuxatorha for cygwin
23:44:48Nuxatori intalled from the url in setup so i guess it's precomplied
23:45:22amiconnMy guess is that gcc 3.3.x won't work for iriver builds
23:45:35Nuxatorha ok
23:45:48SlasheriAah, found probably the last bug!
23:45:57Slasheri(major crash bug)
23:46:02Nuxatori'll try with 3.4.3
23:46:04amiconnSlasheri: There is no such thing as a last bug ;)
23:46:14Slasheriyep, i corrected that ;)
23:46:16NuxatorLast bug found
23:46:26Nuxatoruntil next
23:47:05Slasherigoodnite :)
23:51:27yngwilike i told some time ago, i encountered a bug with mp3 playback
23:51:45yngwion some files, the end gets cut of
23:52:22preglowwe know about that
23:52:32 Quit thegeek (Read error: 54 (Connection reset by peer))
23:52:58yngwii just tried a few songs... and just had to admit, that i could not find a system
23:53:08yngwifor the error
23:53:49yngwii tried songs from a single album which i ripped with -aps and no other modifications, and on some it skips and on others not
23:57:38 Quit ehntoo (Read error: 113 (No route to host))
23:57:53preglowanyone know if [solid] did some vorbis battery tests?
23:58:20 Join ehntoo [0] (
23:58:55yngwii read the whole irc-log and at least I haven't seen any results posted by him, though he said he would do some tests today

Previous day | Next day