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 2005-09-04

00:00:09Moosif you prefer the H3xx series do it :)
00:00:21MoosRockbox coming soon on it
00:00:24XavierGrOnly if Rockbox is ready for it;)
00:00:37]RowaN[im i right in thinking... h3xx vs h1xx = color screen vs optical out
00:00:40Moosbuy one before Rb for the price ;)
00:00:46XavierGrBut if Rockbox is ready for the H300 it would beat h100 for sure.
00:00:49Moosyou know Rb will be ported
00:00:57Moosnot sure ;)
00:01:03]RowaN[or is there more to it?
00:01:06Moosdepend of your choice
00:01:16muesli-Moos pray 4 it every day
00:01:19Moospersonnally I prefer a lot the H140
00:01:33XavierGrwhy ?
00:01:41XavierGrcause the optical in/out?
00:01:59]RowaN[i got a 30gig disk in my h120 now =]
00:02:11Moosyes+radio+I don't look at minus divx... bacckliht...
00:02:26MoosI musiv lover
00:02:44XavierGrradio is better on H300
00:02:49Moosand H140 is probably the best device for it, just it :)
00:02:58muesli-XavierGr really?
00:03:12]RowaN[with 30gigs of music... radio isnt something i care for hehe
00:03:18Moos+size :D
00:03:26XavierGrbetter... well I mean that it is meant to record without the little (very little) glitch
00:03:26muesli-size matters :D
00:03:36muesli-ah ok
00:03:38***Saving seen data "./dancer.seen"
00:03:40XavierGrI agree I don't care about radio too.
00:03:42]RowaN[are h1xx, h3xx different physical sizes?
00:03:50Moosyes it is
00:03:50XavierGrI just mentioned it because Moos said it.
00:03:55muesli-you can compare h140 and 320
00:03:58]RowaN[radio is kinda cool tho, like if u want to hear some news
00:04:08muesli-but the 340 is a brick stone
00:04:24]RowaN[h340 is fatter?
00:04:27XavierGrYou have a nice colour screen and USBOTG though
00:04:28muesli-hell fat!
00:04:35MoosXavier: yes I do the same just news or sport, for music my 40go collection :)
00:04:44muesli-but only usb1.1
00:04:48muesli-not very helpful
00:05:01Mooswhen the backlight turn of the screen apear very horrible for me
00:05:03XavierGrWell if you have a 128mb flash card...
00:05:04]RowaN[3xx only has only usb1.1 ???
00:05:15XavierGron USBOTG
00:05:18muesli-the usb2go port yes
00:05:48koniu_which totally sucks btw - dumping 512mb card takes forever
00:06:04muesli-guess would last ages
00:06:17]RowaN[i wonder if iriver will ever make another decent mp3 player hehe
00:06:39muesli-damn..have to go..g'night mates
00:06:47Moosgood night
00:07:47]RowaN[dont 4get to write
00:07:51XavierGrgood night from me too later all.
00:08:08Moosg'night too :)
00:08:33]RowaN[HOW much!!
00:09:12MoosBagder: are you here?
00:09:15]RowaN[woo proper usb charger for iriver
00:09:19]RowaN[didnt know u could get them
00:10:51 Quit lameD ("CGI:IRC (Ping timeout)")
00:21:06 Join Speedforneed [0] (
00:21:36 Join mrelwood [0] (
00:21:57mrelwoodrockbox is finding it's place with iRiver! Cool!
00:22:54SpeedforneedAnyone having a problem with the Sept. 2 build of the iRiver 120 port when it comes to WPS with bmps?
00:23:35mrelwoodwhat means "boot" and "sim" versions on the bleeding edge download page?
00:24:26 Nick Aison`zappelig is now known as Aison (
00:24:31dpassen1Speedforneed: the syntax for bmp WPS has changed
00:25:11SpeedforneedI know.
00:25:22SpeedforneedI fixed all the ones I had, but no graphics show.
00:25:26 Quit muesli- (Read error: 113 (No route to host))
00:25:40SpeedforneedI was looking at the CustomWPS page.
00:27:07 Quit dpassen1 ()
00:27:10SpeedforneedNo BMPs work at all.
00:28:31 Quit TCK- (Read error: 110 (Connection timed out))
00:28:38SpeedforneedWhen it says that the filename of the WPS can't be more than 24 characters, does that include the directory path too?
00:31:09]RowaN[speed yeah my bmps stopped working from these new builds, but it was just the syntax i had to update in my wps
00:37:37 Quit XavierGr (Read error: 113 (No route to host))
00:41:39 Join sango [0] (
00:41:39 Quit sango (Client Quit)
00:42:24 Join sango [0] (
00:43:24 Quit sango (Client Quit)
00:43:38 Part Moos
00:51:58 Quit koniu_ (Read error: 110 (Connection timed out))
00:55:13 Quit ender` (Read error: 110 (Connection timed out))
00:56:04 Nick paugh is now known as AliasCoffee (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
01:00:32 Quit Zagor (Client Quit)
01:10:05 Quit Aison ()
01:11:14 Join kurzhaarrocker [0] (
01:19:20 Join Wy [0] (
01:19:49Wyhey, anyone alive?
01:20:50 Quit matsl (Remote closed the connection)
01:33:28 Join TCK [0] (
01:34:18 Quit gromit` (Remote closed the connection)
01:36:36 Join solex_ [0] (
01:50:44 Quit solex (Read error: 110 (Connection timed out))
01:54:23 Join ashridah [0] (
02:03:42***Saving seen data "./dancer.seen"
02:09:58 Join Liehbeth [0] (
02:09:58 Quit mrelwood ("CGI:IRC (EOF)")
02:31:53kurzhaarrockerDoes anybody here know what the return value of main_menu means?
02:32:03 Part Liehbeth
02:43:49 Quit preglow ("leaving")
02:45:34kurzhaarrockerIn wps.c, line 650 the function wps_show() returns true even though the return value is declared as long. Is that the intension or just an old relict?
03:07:32 Quit ashridah (Read error: 110 (Connection timed out))
03:09:02 Quit AliasCoffee ("Leaving")
03:10:45 Quit kurzhaarrocker (Read error: 110 (Connection timed out))
03:40:47 Join ashridah [0] (
03:44:27WyDoes anyone know where the flash for the jukebox recorder is? Linky on the site is broken :(
03:55:26 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
03:56:37 Quit paugh (Client Quit)
03:57:51 Nick ashridah is now known as Lost-ash (
04:03:43***Saving seen data "./dancer.seen"
04:04:51 Quit merbanan ("Leaving")
04:06:11 Join QT [0] (i=as@madwifi/users/area51)
04:12:11 Quit cYmen ("zZz")
04:17:23 Quit QT_ (Read error: 110 (Connection timed out))
04:28:49 Quit tvelocity (Read error: 113 (No route to host))
04:46:32 Join random_man [0] (
04:56:24 Quit random_man ("CGI:IRC")
05:13:42 Join bagawk [0] (i=1000@unaffiliated/bagawk)
05:15:54 Join tvelocity [0] (
05:16:26 Quit bagawk (Client Quit)
05:32:28 Nick Lost-ash is now known as ashridah (
06:03:45***Saving seen data "./dancer.seen"
06:35:02 Quit webguest97 ("CGI:IRC (EOF)")
06:50:31 Join DMJC [0] (
07:23:01 Quit DMJC (Read error: 110 (Connection timed out))
07:27:29 Join Paul_The_Nerd [0] (
07:37:52 Quit ansivirus (Remote closed the connection)
07:44:30 Join ansivirus [0] (
08:03:48***Saving seen data "./dancer.seen"
08:15:16 Join knoppix [0] (
08:29:38 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.6/20050716]")
08:42:15 Quit thegeek (Read error: 110 (Connection timed out))
09:09:24 Quit tvelocity ("Leaving")
09:19:39 Join tvelocity [0] (
09:27:09SpeedforneedThat's funny, cause I have yet to go to bed.
09:27:27knoppixhere in good old germany its half past ten in the morning
09:27:38Speedforneed2:27am here.
09:27:59SpeedforneedI == dead.
09:29:05*amiconn is in germany too
09:29:12knoppixamiconn: do you kow the meaning of the return values of the boolean value the menus return? e.g. the retval of main_menu()
09:29:51amiconnknoppix: Yes, they return true when they leave because they detected an USB connection
09:29:53 Join HCl_ [0] (i=hcl@2001:610:1908:8000:290:27ff:feca:8029)
09:30:22knoppixhm. I'll probably abuse that mechanism then
09:31:02amiconnSee menu.c, lines 402ff
09:31:24 Quit HCl (Read error: 104 (Connection reset by peer))
09:32:20knoppixSo all the function that are called within a menu have to return true when they were in usb mode.
09:33:26knoppix(that's what I conclude from the default case in the function you mentioned)
09:33:57knoppixWhat I want to do: When the user quits the recording screen with the on button rockbox shall play the new recording instantly.
09:35:02knoppixunfortunately rockbox usually returns to the menu after the recording screen.
09:35:25 Quit ashridah ("Leaving")
09:35:38 Join ashridah [0] (
09:37:14knoppixSo I abuse the usb mode only to make rockbox skip the menu (and eventually the wps too) and return straight to tree. There I want to start playback of the new recording.
09:37:53 Nick knoppix is now known as kurzhaarrocker (
09:38:27 Join LaMeD [0] (
09:38:52LaMeDhello (and good morning to some)
09:39:02*kurzhaarrocker bows
09:39:19LaMeDa question: disk spindown means lowering the speed or stopping it at all?
09:39:46kurzhaarrockerThe disk stops spinning
09:40:26LaMeDYes! thank you. I knew my translation was more correct
09:42:17amiconnkurzhaarrocker: Imho it could be useful to extended the menu system's return values to be more flexible
09:43:40amiconnAfaiu the reason for having the return value is to let the tree know that there was usb, so the file system contents may have changed and needs to be re-read
09:44:52kurzhaarrockerI was in doubt wether it means "was in usb" or "file system changed"
09:45:08LaMeDhere's another one. why would in the previous version rockbox has asked "is battery low?" if saved settings has faild on recorder, and now it asks "no partition?"
09:45:16amiconnThe plugins do the same, and some of them (which create files) mis-use the "was in usb" return value
09:47:32amiconnLaMeD: This is historic; originally it was thought that the most likely cause for save settings failing would be low battery
09:47:35 Join Xanthos [0] (
09:48:01XanthosWow, full channel, anyone at keyboard?
09:49:33XanthosHey, I need some help with an ifp player
09:49:43XanthosWindows ate it
09:50:21*Xanthos offers free cookies :)
09:50:58solex_ifp == iriver?
09:51:16solex_disk broken?
09:51:20XanthosIt's an old on
09:52:22Xanthoskind of, it's the ifp100 series I've running firmware to make it a UMS and it works find, but XP crapped on the drive and when I tried to reformat it it failed
09:52:47XanthosNow the format is corrupt so I can't stick a hex file on it to flash the firmware
09:53:16solex_Have you tried repartitioning it?
09:53:22Xanthosthought about it
09:53:36Xanthosis there a hidden partition that stores all the iriver crap?
09:53:49Xanthoswas afraid of wiping that out
09:54:04solex_no, there is none (at least not on my h120
09:54:23Xanthosdisk manager isn't giving me an option to delete the partition
09:55:14Xanthosshows a 250meg healthy FAT partition (XP's not too swift is it)
09:55:51solex_whatever it means with "healthy"
09:56:14solex_and you really cannot delete it?
09:56:23solex_in other words: you know where the entry should be
09:56:23Xanthosany drive that's been attached to an XP os CAN'T be healthy
09:56:40Xanthosyeah, unless it's somewhere different for removable drives
09:56:41solex_Well, try Knoppix ;)
09:57:30Xanthosis it bootable?
09:57:43XanthosI mean, without a hard drive
09:58:16solex_Knoppix is a bootable Linux-CD-ROM
09:58:24solex_or even a DVD
09:58:27Xanthoslyke could I burn an iso and boot, format the piece of crap and then be happy?
09:58:54Xanthosthink it would see the usb ok without windows 500megs of drivers?
09:59:33solex_Knoppix? Sure. USB Mass Storage support is in all recent kernels.
09:59:51XanthosI need to learn linux anyway, I'm so friggin sick of M$ I could puke
10:00:50solex_if you've got the bandwidth:
10:01:06Xanthosso does linux support all the file formats? FAT/16/32 NTFS ?
10:01:19solex_NTFS is read-only in Linux
10:01:28solex_but all the FAT formats are fine
10:01:54Xanthosthat's cool, NTFS in "what? where, I don't see another disk there" in DOS
10:02:16Xanthoscool, torrents :)
10:02:46solex_if it's too slow:
10:03:18solex_are you behind a router with dhcp?
10:03:49***Saving seen data "./dancer.seen"
10:04:25solex_good. then you will probably have an internet connection when running knoppix
10:04:32solex_there should be an irc client as well
10:04:41Xanthosthese are all dutch?
10:04:55Xanthosor is that german
10:05:06solex_there is a german and an english version
10:05:24Xanthosis 4.0 available on cd?
10:05:45solex_I don't think so, but I am not sure
10:05:58Xanthosoh duh, -EN and -DE isn't a dead giveaway on the lang
10:06:02solex_3.9 is still fine
10:07:10solex_when you boot from the cd, there will be a menu at the beginning
10:07:34LaMeDamiconn, are you there?
10:07:40solex_you have to press f1 or f2 to show some options
10:07:53solex_there you can select another keyboard layout
10:08:01solex_in case you don't want german or english
10:08:42Xanthosso it doesn't need any writable media at all to run?
10:08:43amiconnLaMeD: yup
10:09:08Xanthosthis would kick @$$ for a server box
10:09:22Maximemaybe but you can't write settings anywhere..
10:09:39Maximeso, when you reboot you have to reconfigure all
10:10:07XanthosI'll just use my 2 gig dos partition and store whatever it needs
10:10:18Xanthosdos is kidna useless now anyway
10:11:19Xanthosanyway, why do you think i wouldn't be able to delete the fat part on the iriver?
10:11:40 Quit tvelocity ("Leaving")
10:13:30solex_Xanthos: no, Knoppix doesn't need write access to any disks
10:13:48solex_but I have no idea, why windows doesn't allow you do delete the partition
10:14:46Xanthosman, that's a sweet torrent, I'm getting like 650kB a second
10:18:04*solex_ grumbles
10:19:28*Xanthos wonders if there's a way to flash the firmware from another drive
10:20:07 Join gromit` [0] (n=gromit`
10:20:09solex_what exactly do you mean?
10:20:25solex_the iriver firmware isn't on the hard disk, as far as i know.
10:20:46Xanthoslike, you have to have the .hex on the ums and flash from the menu in the device...
10:21:13Xanthosbut since the file system is corrupt I can't put the hex on the drive
10:22:31solex_I think the path to the hex file is hardwired into the original firmware, so i suspect the answer is 'no'
10:23:40Xanthosand I think the firmware files are encrypted too
10:25:08Xanthosok, but the original wasn't ums... if there's anything permanent in the device there should be a way to get to it right?
10:25:27XanthosI tried flashing an empty .hex tho... no good
10:26:22solex_btw, what is this "ums" stuff exactly?
10:26:33XanthosI've got a regular jump drive here, I can delete the partition on it...
10:26:39Xanthosmass storage device
10:26:45Xanthosnot sure
10:27:57Xanthosboth are seen as standard disk drives
10:28:27XanthosI'm thinking the only thing that could stop me deleting the partition is the driver, but it's a microsoft deal driver
10:28:56solex_another "feature" to protect you from yourself, I guess
10:29:23solex_linux allows everyone to shoot their feet as often as they want
10:29:28Xanthosok, I see one difference, it sees the iriver as a removable and the regular as a "basic"
10:29:46solex_which one is which?
10:29:51Xanthosregular usb jump drive
10:30:04solex_sorry, parse error on my side
10:30:15Xanthosthe iriver is seen as removable
10:30:16LaMeDcvs -z3 co rockbox < why is this wrong?
10:30:26solex_well, the iriver *is* removable and your other hard disks are probably not.
10:30:47LaMeD":no such repository"
10:30:50Xanthosno, I'm comparing it to a usb jump drive
10:31:38solex_and you (principally) could repartition the "basic" disks?
10:31:42Xanthoshmm, it doesn't have a logical drive letter in the properties... how is that possible
10:32:09solex_I suspect your ifp is not a regular mass storage device...
10:32:18amiconnLaMeD: cvs -z3 co rockbox-devel
10:32:23solex_...which would mean, you're out of luck with knoppix
10:32:29Xanthosit has no volume label either
10:32:40solex_but it's in the explorer?
10:32:43Xanthosbut it does...
10:32:50Xanthosyeah, it has a letter and label in explorer
10:32:56solex_Sounds like the behaviour of my mobile phone
10:33:17Xanthoserm, well it did lol maybe since I forgot I'm formatting it right now
10:33:46Xanthosit's being formatted right now, probably why it's unmounted :)
10:33:57LaMeDAmiconn, thanks. could you answer me in private window?
10:34:01solex_what are you formatting?
10:34:07Xanthosthe iriver, again
10:34:37Xanthosthis sucks
10:36:06Xanthosok, you can flash an optical drive with an exe and a hex file... couldn't you do that to a usb drive?
10:37:11solex_I think it depends on what the device supports
10:38:07Xanthosnero burns bootable roms right?
10:38:52solex_don't burn too fast, you'd overtake me. :)
10:38:59*solex_ is burning 4.2x
10:41:07solex_um, xanthos?
10:41:19solex_my iriver h120 has a menu option for formatting
10:41:28solex_menu -> general -> format
10:43:39solex_what about yours?
10:43:42kurzhaarrockera c question: I defined a variable outside of any function "static char path_buffer[MAX_PATH];" From a function within that file I want to "return path_buffer;" That should work, shouldn't it?
10:43:46Xanthosyeah, I looked for that, not in this one :)
10:45:21 Join ender` [0] (
10:45:37solex_would have been too easy
10:47:14LaMeDsorry for being so lame... how do I send a staffer
10:47:53kurzhaarrockerargh, I mixed up a function variable with a global variable. grrr.
10:48:24 Quit LaMeD ("CGI:IRC (EOF)")
10:48:55Xanthosnow all your variables will be globally non-functional... OMG you just wrote WinXP
10:49:05 Join Lshachar [0] (
10:53:56*solex_ is preparing breakfast
10:54:08*kurzhaarrocker steals solex_ a toast
10:54:25XanthosI think I did it
10:55:19XanthosI told it to mount it as if it were a 5 1/4 disk, and I have the irq and i/o for that drive disabled in bios, so it HAD to format it like a real volume
10:55:31Xanthosit's a 250 meg floppy disk now
10:55:42solex_on 5"1/4
10:55:48Xanthoslol yeah
10:55:51solex_nice! be sure not to bend it!
10:56:24Xanthoswell, if it works, at least I got a bootable copy of linux now :)
10:58:27Xanthoswoohoo! firmware is upgrading yay Thanks for your brainstorming help man
10:58:36solex_you're welcome
10:58:59solex_but you should really try knoppix, it's quite impressive what they get on a single cd
10:59:15Xanthoshehe, I was already checking prices on the H series I may get one anyway... I just love this little triangle tho
10:59:29XanthosYeah, I've got it in the drive, gonna check it out next time I reboot
11:00:03Xanthosgotta do some BF2 now tho... lol I was getting my mp3's together to play Batlefield like... 6 hours ago when Winblows decided to urinate on my player
11:00:11 Join tvelocity [0] (
11:00:12kurzhaarrockerHas anybody tried to use knoppix (dvd) as development environment for rockbox?
11:00:57solex_kurzhaarrocker: no. you would need some place to write
11:01:24solex_so you can as well go ahead and install any flavour of linux
11:01:25kurzhaarrockeryou can mount hd easily
11:01:26 Join bluebrother^ [0] (
11:01:42Xanthosis rockbox an os or a media manager?
11:01:53kurzhaarrockeror use an usb stick or even the jukebox itself as storage for your home dir
11:04:01kurzhaarrockerI haven't managed to build the cross compiler on knoppix yet, but I didn't try very hard.
11:05:40solex_Xanthos: an os
11:05:59solex_but it lives on the harddisk
11:06:04solex_unlike the iriver firmware
11:06:13solex_that way you can dual-boot
11:10:16Xanthoswhat are the advantages of it over other os's? it's small?
11:10:36 Join cYmen [0] (
11:11:41kurzhaarrockerrockbox is a open source firmware replacement for portable mp3 players
11:13:18kurzhaarrockerIt doesn't really make much sense to think of it as an os, its more.
11:14:04solex_imo, rockbox's best features are the customizable wps and on-the-fly playlists
11:14:22Xanthoswow guess I can't use it tho, at least till I buy a player with an actual hard drive
11:14:46 Nick HCl_ is now known as HCl (i=hcl@2001:610:1908:8000:290:27ff:feca:8029)
11:15:20Xanthosthis one's just 250 meg flash memory
11:15:21kurzhaarrockerOnly a few hand selected players are supported. See The very first sentence :)
11:15:55kurzhaarrockerAnd the ondi players are flash based, without hd ...
11:16:43 Join DangerousDan [0] (
11:17:27Xanthosman, that sounds awsome... I've gotta keep rockbox in mind when I pick which player I'm gonna buy then
11:17:55solex_i can recommend the old h1x0 series
11:18:01*kurzhaarrocker wonders how Xanthos found this channel without knowing what rockbox is about.
11:18:10solex_i bought one only a few weeks ago
11:19:04*solex_ too
11:19:06Xanthosthere's a few blogs on it
11:19:51XanthosI was planning on using rockbox to flash my firmware then I saw the irc channel on the site and came here first (before finishing reading the site) :)
11:20:18kurzhaarrocker... including the first sentence :) ...
11:21:13Xanthoswell actually I noticed my player wasn't supported which isn't anything new for me since it's like one of the first made lol
11:22:34solex_does anybody know why rockboy is not part of the iriver builds?
11:22:46Xanthosifp-190 doesn't get much attention from iRiver either, but judging from the fact that you guys have this wicked open source project going, I guess not many models get much attention from them
11:22:50 Quit Febs (Read error: 104 (Connection reset by peer))
11:23:33 Join Febs [0] (
11:24:35Lshachar<Solex_> maybe one of those?
11:24:37LshacharWe have a few other convenient aliases that gets several modules at once for you:
11:24:37Lshacharrockbox - gets everything you need to compile and build rockbox for target
11:24:37Lshacharrockbox-devel - like 'rockbox' but also includes simulators and gdb code
11:24:37DBUGEnqueued KICK Lshachar
11:24:37Lshacharrockbox-all - gets everything there is in CVS, all modules
11:24:37Lshacharwebsite - gets the www and docs modules
11:26:12Lshacharit's present in the Devel package. i've just checked.
11:26:38solex_hm, that might be a reason to set up a devel-environment...
11:26:44 Join webguest06 [0] (
11:26:49webguest06lshachar 123
11:27:47 Quit webguest06 (Client Quit)
11:29:00kurzhaarrockerthose irivers don't feature digital in, do they?
11:29:10amiconnThey do
11:29:20amiconnoptical in, optical out
11:29:37HClh3x0 doesn't
11:29:40HClbut h1x0 does
11:29:47Xanthoswhat's the max disk size on the new irivers?
11:29:57kurzhaarrockerah, thanx, amiconn
11:31:13HCli think 40gb, but last time i checked they're upgradable
11:31:45Xanthosare they using a third party for the drives? (like western digital)
11:32:41HClmst upgradable mods are done by users
11:32:49HClnot supported by iriver themselves
11:33:16Xanthosbut there's space in them for an industry standard thumb drive?
11:33:38HCli'm not sure what you mean with thumb drive
11:33:44HClthe iriver uses 1.8" drives
11:34:18Xanthosthat's probably not the right word.. I don't know much about recent hardware, havn't been keeping up with it
11:34:54Xanthosbut they're like an industry standard tho right? not something proprietary to Iriver
11:35:34HClthey are
11:35:52Xanthosman, I gotta get a player
11:36:16Xanthosyou know what kind of ram they're using in them?
11:36:26HClnot really
11:36:46Xanthosprobably all soldered in anyway
11:37:25Xanthosbut you could really mod the crap out of them if you could get the parts
11:37:40HCli guess. irivers have 32mb of ram by default
11:37:52HCli can't really see how having more would help
11:38:23Xanthoswell, if they have a decent processor in them, you'd could make it closer to a pda than a portable player
11:38:43HClthe cpu in it is only a coldfire 140mhz
11:38:51HClrunning at 120mhz max
11:39:07Xanthosguess they wouldn't need much more for audio processing
11:39:39solex_Xanthos: and for a pda you would need a decent keyboard
11:40:02Xanthosor a touch screen
11:40:04HClfor a pda, i just use my pda, lol.
11:40:15Xanthosgood point
11:40:26solex_maybe someone could hack in a touchscreen... :)
11:40:39Xanthossounds kind of expensive :)
11:40:49XanthosI need an lcd for my laptop
11:41:04Xanthosturns out HP's are incompatible with asphalt
11:41:26kurzhaarrockerplug'n break
11:41:55Xanthosyeah dam samsonite suitcase is to blame tho... told it to stay and it rolled
11:42:12Xanthosthat or the screen wasn't strong enough, I just know I'm not to blame :P
11:43:49Xanthosmulti-handicapped mainstream?
11:45:18Xanthosvideo playback with sound? awsome
11:48:02kurzhaarrockeron the archos it's a funny gimmick. It works but you can't really watch movies on 112x64 pixel.
11:48:13Xanthosomfh... 8 minutes to transfer 150 megs, I really do need a new player
11:48:23HClagain something for which i use my pda... heh..
11:48:36HCldivx 320x240 16bits color..
11:48:39Xanthoslol it's a cool feature to have tho
11:48:48HCli dunno
11:48:57kurzhaarrockerI use my mobile phone for that :)
11:48:57HClsome features i don't see the usage of
11:49:02Xanthoswell if the screens get better, rockbox will be ready
11:49:08HClbut i guess other people have the same thing with features i add.
11:49:23HClthe problem is that though the h3x0 has a colorscreen
11:49:28HClthe cpu is still the same one
11:49:42HClpeople expect to get color playback and such and color gameboy on h3x0
11:49:44Xanthosso more of a slide show than a video
11:49:55HClbut they fail to realize that the cpu isn't upgraded at all and still very slow
11:50:18HCli believe the iriverfirmware actually had a decent framerate on h3x0
11:50:27amiconnkurzhaarrocker: Believe me, you can watch movies. I did this back when playing around with the video plugin
11:50:48amiconnThe white backlight mod is helpful...
11:51:55amiconnIn general it's not really suited for watching movies, but for the occasional music video it is sufficient...
11:52:00Xanthosany news on the next series to come out?
11:52:04kurzhaarrockeryes you can watch them but can you really enjoy them? Don't get me wrong. I love that geek feature.
11:52:56amiconnAt least the archos lcd is faster than the iriver H1x0 lcd
11:53:02Xanthosmusic video's might be something iriver has on the board
11:54:05kurzhaarrockerXanthos: Info on the next version of rockbox:
11:55:41XanthosCool, I was talking about the next series of players, but Rockbox looks like they're already thinking about the future
11:57:28Xanthosgonna go do some battlefield, I'll drop back in this channel for some advice b4 I buy my new player tho
11:58:21 Quit Xanthos ()
11:59:04 Quit Lshachar ("CGI:IRC")
12:02:24 Join webguest84 [0] (
12:02:32 Quit tvelocity (Read error: 104 (Connection reset by peer))
12:03:51***Saving seen data "./dancer.seen"
12:05:50webguest84amiconn, I'm having difficulties...
12:06:03webguest84few more minuts (lshachar)
12:06:08 Quit webguest84 (Client Quit)
12:06:26 Join Lshachar [0] (
12:07:04kurzhaarrockergoto 1
12:07:26kurzhaarrocker*stack overflow error*
12:07:26Lshacharloop until i = 25
12:07:54Lshachardebug <kurzhaarrocker>
12:08:07kurzhaarrockerNo soap please!
12:09:40 Join webguest94 [0] (
12:10:01 Join Moos [0] (
12:10:33MoosHello guys!
12:10:43 Quit webguest94 (Client Quit)
12:17:47Slasherihi :)
12:23:07CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
12:23:07*HCl yawns
12:23:43*kurzhaarrocker splashes a bucket of cold water over HCl to wake him up
12:24:48HClthat doesn't work on me
12:36:31LshacharDAAAMMMNNN..... alll thhat wastedd tiimmeee...
12:50:09 Join Mxm`Pas`Bien [0] (
12:50:09 Quit Maxime (Read error: 104 (Connection reset by peer))
12:50:55 Nick Mxm`Pas`Bien is now known as Maxime (
12:54:30 Quit kurzhaarrocker (Remote closed the connection)
12:54:33 Join hicks [0] (
13:01:42HClLshachar: ?
13:02:25solex_damn, my new wps file is 33 characters too big
13:04:26 Join gulugulu [0] (n=HRCLITE6@
13:04:30solex_is anyone here able/willing to increase the maximum wps file size?
13:04:50solex_the new conditionals (esp. for codec) can become quite big
13:04:59 Join hebele [0] (n=HRCLITE6@
13:06:48hebelehi yall I have a question. I know this is not your business but I have to ask it and u re my last chance. I need answers key of this test : . I m a learner of English. Please help me :)
13:07:48HClyou won't learn english by cheating a test.
13:08:25hebeleI did it but I dont have answers key of the test
13:09:13hebelepardon I made mistake sorry
13:09:57hebeleI did it but I havent answers of the test. I hope this is true :)
13:10:50hebelecan u help me. I suppose not. Thanx yall. Have a nice day
13:11:00Lshacharhcl yeah?
13:11:32HCli was confused about what you meant with wasted time?
13:11:38HClhebele: i can't even access that site
13:11:42HCl :/
13:12:55Lshacharhcl I wasn't identified and tried sending amiconn messages for half an houre
13:13:31hebeleanyway thx have a nice day
13:13:50 Part hebele
13:17:21HCl :X
13:23:33 Quit gulugulu (Read error: 110 (Connection timed out))
13:26:01 Part t0mas
13:26:01 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
13:26:25Mode"#RockBox +o t0mas " by ChanServ (ChanServ@services.)
13:26:29Mode"#RockBox -o t0mas " by t0mas (n=Tomas@unaffiliated/t0mas)
13:26:31t0masok, works
13:26:47t0masis some-one with an archos device around?
13:26:56t0masid: LANG_F2_MODE
13:26:56t0masdesc: in wps F2 pressed
13:26:56DBUGEnqueued KICK t0mas
13:26:56t0maseng: "Mode:"
13:27:01t0masin the language file...
13:27:05t0maswhat "mode" is that?
13:27:11t0masPlayback mode?
13:33:15LshacharI don't have time right now to build an archos sim, but on the iriver (which is very similar in that menu ) a long click on the a-b button - the string "mode:" appears twice: shuffle mode and repeat mode. this is very similar in archos.
13:33:32 Join Paul_The_Nerd [0] (
13:34:38 Quit Paul_The_Nerd (Client Quit)
13:36:09t0masI've build an archos sim
13:36:17t0masit's playback mode
13:36:22 Join tvelocity [0] (
13:39:18 Join Lear [0] (
13:39:46LearAnyone else with simulator build problems?
13:39:58t0maswich one?
13:40:36Leaririver, though I wonder if that matters...
13:40:48t0maswell... I just did an archos build, and it worked
13:40:57t0masdoing Iriver now
13:41:50t0maserror's in dsp.c ?
13:41:54LearIt works fine until GENLANG, then I only get the following message: gcc: 1: No such file or directory
13:41:58LearNot very helpful...
13:42:27t0masno... not really...
13:42:36t0mascheck apps/lang
13:42:41t0masare the .lang files there?
13:42:46t0masand have you done make in tools?
13:42:58LearSure is, I mean, lang.c is generated properly...
13:43:32t0masno, in the apps/lang directory... are the files there?
13:43:40t0masand have you run make in the toold dir?
13:43:50 Quit Lshachar ("CGI:IRC")
13:44:07LearI have tools; did a target build yesterday...
13:46:58LearSeems like it is one of the defines on the command line; I removed them, then it worked...
13:48:10LearHm.. -DIRIVER_H120 1 lacks an equal sign. tools/configure time I guess...
13:48:49 Join matsl [0] (
13:48:53t0masah, can you try if it works with a clean checkout?
13:49:27SlasheriLear: make clean and reconfigure
13:49:28Learlets try the re-configure first...
13:49:35Slasherithen it works
13:50:41LearAh, yes, that "broken sim build" commit...
13:56:59solex_for those interestet: i finished my wps with a custom status bar
13:57:13solex_it uses images for play status, repeat mode and codec:
13:58:08solex_it's just a pity that i hit the maximum file size limit for the wps
13:58:24solex_the declaration for the codec bitmaps is a little bit greedy
13:59:06amiconnSlasheri: Your updated finnish.lang has some problems.
13:59:47amiconn(1) it contains many of the ### todo comments. These are supposed to be removed when done, not to be committed
14:00:32amiconn(2) The deprecations aren't done
14:01:22LearFilesize could be increased a bit on iriver, I guess.. But shouldn't one of the images read "VBR"?
14:01:54 Join Zagor [0] (i=foobar@pdpc/supporter/sustaining/Zagor)
14:03:55***Saving seen data "./dancer.seen"
14:03:55amiconnImho it would be useful to support bitmap clipping, especially since we have conditional image support.
14:04:23amiconnCurrently you have to use one .bmp per image, not very practical if you want to have dozens of them
14:04:45amiconnWith clipping support, all images used in a .wps design could reside in one .bmp file
14:07:11solex_amiconn: how does it work? is it like a gif file with several images inside?
14:07:20t0masit's 1 bmp file
14:07:28t0mascontaining all images next to eachother
14:07:38t0mas*each other
14:07:59t0masand I think it's a good idea
14:08:13t0masthen you load just 1 bmp file... and display all images with separate tags
14:08:32solex_and you have to specify a rectangle to display?
14:10:08solex_Lear: no, the "VRB" is for Vorbis, bot for VBR
14:10:44solex_VBR is not displayed - not enough space
14:11:34 Quit edx (Read error: 110 (Connection timed out))
14:11:57LearI see. :)
14:13:47solex_if you have a better 3-letter abbreviation...
14:14:08solex_I didn't want to say "OGG", cause it'sjust the container
14:14:26crwli guess "Ogg" is a lot more informative, though not really correct...
14:15:26Slasheriamiconn: oh, i will fix that
14:15:39Slasheriamiconn: can i completely remove the deprecated strings?
14:18:11amiconnInteresting effect of moving the credits to a plugin: The Ondio is the only target where you don't notice it at all
14:18:15Slasherihmm.. i think i will hack the uplang script to do that
14:18:32amiconnInfo->Version is instantaneous, as before.. :)
14:21:58solex_I updated the vorbis.bmp
14:22:12solex_it now shows 'OGG'
14:22:54solex_and the zipfile now includes the .wps
14:23:12solex_btw, does anybody know why the wps gallery page is so wide?
14:23:52solex_ah, i found a very long line in a <verbatim> environment
14:26:12solex_i suggest the verbatim-thing should have the css-attribute "overflow: auto;"
14:26:30solex_that way you would get scrollbars just for the verbatim part
14:26:42Zagorsolex_: good idea. i'll try that.
14:27:38 Join muesli- [0] (
14:27:58CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
14:27:58*Zagor learned something new today
14:28:28solex_it's a little bit annoying that (at least firefox) sends the scroll wheel event to the <pre> part
14:28:42solex_but it's definitely better that way
14:29:10 Join edx [0] (
14:30:07solex_and maybe it would look a little better if the verbatim part had a different backgound-color
14:30:26solex_but i don't know where exactly it is used
14:30:35solex_i mean, on which pages
14:30:40solex_and whether it is appropriate
14:30:53Zagorit doesn't have its' own class yet, so I simply changed the <pre> tag. so I don't want to change it too much... :-)
14:31:08solex_it's ok
14:31:54solex_the cool thing is, that you don't have to see it all just to select, copy and paste it
14:34:07muesli-moin jochen ;)
14:34:52amiconnZagor: This overflow:auto is looking a bit strange, at least in firefox
14:35:29amiconnIt still shows the vertical scrollbar even if there is nothing to scroll
14:36:11solex_amiconn: not with my firefox (1.0.6 on Linux)
14:36:12amiconnHmm, at least it does that for *some* of these areas
14:36:23Zagorthat's odd
14:36:31solex_oh no, you're right
14:36:36amiconnIt definitely does that here with
14:37:07solex_it is because they have spaces at the end of the lines
14:37:07amiconnThese don't need any scrolling here
14:37:19amiconnFirefox 1.0.6 on WinXP
14:37:24amiconn1400x1050 screen
14:37:24solex_try selecting the text
14:37:27Zagordoesn't do that for me
14:37:45solex_i'll try to edit them
14:38:01Zagoroh, found one. NicolasGif has one for me. very odd indeed.
14:38:07 Quit Lear (Excess Flood)
14:39:13solex_TorstenT should be better now
14:39:28 Join preglow [0] (
14:39:30solex_it's the spaces at the end
14:39:40solex_probably happened while copy pasting from a terminal
14:39:55amiconnAll gone now...
14:40:07solex_all of them!?
14:40:44ZagorI can still trigger NicolasGif if I resize the window out and in
14:40:50 Join Lear [0] (
14:40:52Zagorlooks like a Firefox bug
14:42:04solex_agree. misread vertical/horizontal
14:43:07amiconnNow I could trigger it by resizing for JoerchB and CarP again
14:43:19amiconnReloading the page fixes it
14:43:34amiconnThe strangeness of css...
14:44:00preglowthe strangeness of css implementations
14:45:58LearBtw, regarding the playback menu, shouldn't at least optical out and maybe anti-skip buffer be moved to the system menu?
14:46:43LearAnd tag priority?
14:49:58solex_imo optical and anti-skip should be moved to system
14:50:09solex_and tag priority to display
14:50:21 Quit hicks (Read error: 110 (Connection timed out))
14:51:01Learno, not display, I'd say...
14:53:15LearAnother thing I thought about recently... To support chained ogg streams (and cue sheets, I guess), the track metadata should be stored in the playback buffer, alongside the file data (like a new codec is).
14:54:27LearThat way, a more or less arbitrary number of streams (with corresponding tag info) can be stored, and support for custom vorbis and ape tags can be added without making the current metadata array too large...
14:54:52muesli-moving "playback to the uppest level and "sleep timer" to playback is my wish :o
14:55:28LearCould be a bit tricky to handle the situation where a chained stream is (much) larger than the buffer though...
14:58:11preglowlear: yes, and where the current track is too large to fit in the buffer, there'd be no next track info, then
14:59:47Learwell, it wouldn't be impossible to reserve a little space for that, I guess, but it might be tricky to do...
15:00:28amiconnpreglow: The next track info problem also happens right now if a track is too large to fit the buffer
15:01:09Leardoesn't the iriver playback code always load the next track info, even if nothing of that track is buffered?
15:01:51amiconnI don't know. The archos playback code doesn't do that
15:02:06amiconnImho it doesn't make sense to do that
15:02:31amiconnIt costs additional battery power. I consider the next track info a little extra.
15:03:13preglowi consider it quite useless, hehe
15:03:20amiconnIf it is easily made available since we are already holding (a part of) the next track in the buffer, we display it, otherwise we don't bother
15:04:22amiconnKISS, you know...
15:05:54preglowi wish someone would start adapting the codec api for non-streaming formats soon :)
15:06:48amiconnpreglow: Any news on optimised codecs, profiting from the extra iram, and possibly less emac inits?
15:07:37Learwhat do you mean by non-streaming?
15:07:48preglowamiconn: none other than that i'm working on it, i've got a million other things that require my attention atm, i'm hoping to get some work done today
15:07:55preglowLear: like mods, midi
15:08:06preglowLear: codecs that require the entire file to be loaded into the buffer at once
15:08:22 Quit matsl (Read error: 110 (Connection timed out))
15:09:05Slasheripreglow: Hmm, that should be simple to make. We would just need to guarantee that the entire file remains in the buffer all the time
15:09:10preglowLear: apart from the buffer handling, these codecs will also very probably need to load their own data
15:09:16preglowSlasheri: read last line
15:09:38preglowSlasheri: the loading part will mean we need another function in the codec plugins, and i don't know how to handle that
15:10:29preglowSlasheri: take for example a .mod file, you can't just read it verbatim into the buffer and expect the playback engine to have an easy time, it might have to decompress pattern data, decompress samples, etc
15:10:35Slasheriwhat about if we just add a configuration option like CODEC_STREAMING (true/false)?
15:10:39amiconnImho it would be easier to support non-streaming formats via plugins first.
15:11:03Slasherithen the codec could tell the core if it requires access to the entire file constantly
15:11:15 Join [1]Moos [0] (
15:11:16preglowSlasheri: what's more, the loading function should be able to be invoked prior to load the actual playback plugin
15:11:21amiconnWhile that doesn't allow them to be part of a playlist (yet), it would certainly help to see what functions/ accesses are really needed
15:11:30 Quit Moos (Read error: 104 (Connection reset by peer))
15:11:30 Nick [1]Moos is now known as Moos (
15:11:32amiconn...and it would also help optimising them
15:11:34preglowamiconn: certainly, i expect this to be what will happen
15:12:03preglowbut the things i've mentioned thus far, i'm very certain will be needed
15:12:09 Quit edx ()
15:12:54amiconnI think that a non-streaming codec requires flushing the audio buffer at start, and the buffer would be controlled by the codec while it is playing
15:13:10preglowwhy require flushing?
15:13:27preglowi don't think that will be required in the majority of cases
15:13:30amiconnFor at least 2 reasons
15:14:06amiconn(1) The codec will probably need the data in contiguous form. Ordinary buffering may wrap the data around at the end of the buffer
15:14:35preglowwell, sure, it can't do that if it detects a non-streaming codec
15:14:46preglowif the file is too large, it wont be loaded until the buffer is empty
15:14:49amiconn(2) The codec will possibly need extra ram. Much extra ram.
15:14:51amiconn(a) The data might be compressed
15:15:19amiconn(b) The codec might need to load largeish extra files for playback, like the instrument set for midi etc
15:15:40Slasheriyes, we could give non-streamings codecs a direct access to the file buffer
15:15:46preglowwe have top
15:16:03preglowbut it still doesn't mean the data can't coexist with what's already there
15:16:17preglowthe codec just needs a means of telling the buffering layer that what's left of space isn't enough
15:16:22amiconnYes, and it implies a short pause at the start of such a non-streaming track
15:16:23preglowand that buffering should wait until more is available
15:17:00amiconnThe problem isn't to ensure enough memory is available, but to ensure it is contiguous
15:17:54preglowbut that's the same problem, really, afaik, the buffer doesn't contain gaps of notable size, and the only non-contiguous portion is at the end
15:18:14amiconnNo, it isn't
15:18:31amiconnWith ordinary buffering, the code always buffers as many tracks as fit the buffer
15:19:02amiconnThat means there will already be some more tracks buffered after e.g. a midi track (*very* likely with midi)
15:19:11amiconnThese need to be flushed to make room
15:20:12amiconnResetting the whole buffer is the simplest solution here
15:20:23preglowsimplest, but incredibly wasteful
15:20:32preglowi don't want to spin the disk up every time i skip to a new mod
15:20:36amiconnEverything before the non-streaming track is already decoded, and everything after it must be flushed anyway
15:20:37preglowor spc, for example
15:20:42preglowthese files are 64k big
15:21:35amiconnOtherwise we would need a means for the codecs to tell the buffering system to leave extra room *at buffering time*
15:21:49preglownot if we allow the codecs themselves to load their own data
15:21:53amiconn...probably doing a part of the decoding at load time
15:22:03preglowthey'll know exactly how much they need
15:22:17preglowa mod loader will decompress pattern data and samples, and place them where they can be used when playing backj
15:22:34preglowno further space will be required
15:22:56amiconnThat's what I'm trying to say all the time
15:23:14amiconnA non-streaming codec will need to do some operations *at buffering time*
15:23:18preglowoh, yes
15:23:24preglowthat's what i've been saying all the time as well :P
15:23:27preglowok, we agree then
15:23:27 Join edx [0] (
15:24:01amiconn...because at start-playback time it is too late. Subsequent memory is already taken by the next track(s)
15:24:25 Quit muesli- (Read error: 104 (Connection reset by peer))
15:24:29amiconnThe mod codec needs to decompress on-the-fly in case of compressed mods
15:24:47amiconnThe midi codec needs to buffer the instrument set
15:25:06Slasheriperhaps the metadata system could handle that. If it detecs for example a mod file being loaded, it could calculate the amount of free space we would need on the buffer
15:25:19preglowthis is how most libs work anyway, they have a load function that takes a mod and prepares some internal representation of it
15:25:31preglowSlasheri: you probably can't do that without actually loading it
15:25:45Slasheripreglow: hmm, that's true too..
15:25:49Learbut you might need pretty much all of the codec then, so metadata loading should perhaps be done by the codec...
15:25:56Learand we have codec swapping now... :)
15:25:59preglowLear: it should, if you ask me
15:26:14Slasherihehe :D
15:26:45Learneed a new way to yield though, so that the currently playing codec gets a chance to do that...
15:26:56amiconnThat's not possible
15:27:16amiconnThe playing codec might be entirely different from the one that tries to prepare its data
15:27:28preglowwe would need to split the codecs in two parts, yes
15:27:37preglowand preload all of the loader/metadata parts...
15:27:42amiconnHehe, just wanted to say the same :)
15:28:32amiconnI didn't try how well gcc handles it, but the loader/metadata part could be built as pic
15:28:41preglowi think gcc handles it quite well
15:28:50preglowdepends, of course
15:28:57preglowif you want pic with or without fixups
15:29:07amiconnThis way the whole metadata handling could be made fully dynamic
15:29:33amiconnIdeally, addition of a codec doesn't require a change in th ecore
15:29:37amiconn*the core
15:29:43*preglow laments the fact that the best spc playback engine is written in x86 asm :/
15:29:59preglowamiconn: that's what my vision of a good system would be as well
15:30:00amiconnUse an x86 emulator
15:30:17preglowbut we'd need a more complex plugin format for that
15:30:19amiconn...with dynarec ;)
15:30:21preglowwhich isn't _that_ hard to do
15:30:25*preglow summons hcl!
15:30:31*amiconn looks at HCl
15:30:57preglowi've gotta go help a friend for a while
15:31:11amiconnpreglow: We only need a more complex format if both parts should reside in one file
15:31:34amiconn(which I would consider a good thing)
15:32:15amiconnDoesn't need to be very complex, just a little header telling the position of the 2 parts
15:32:49preglowyes, indeed, and it would need to be relocatable
15:33:01preglowif this can't be done properly without a relocation table, we'll need that too
15:33:08preglowbut now i gotta go, brb
15:33:21amiconnI would try to use full pic, without any necessary relocation
15:33:37amiconn...for the metadata part
15:35:44 Quit ashridah (Remote closed the connection)
15:35:48 Join Lost-ash [0] (
15:38:51 Nick Lost-ash is now known as ashridah (
15:40:17 Join hicks [0] (
15:43:25 Join matsl [0] (
15:50:02 Join linuxstb [0] (
15:50:29 Quit edx (Read error: 110 (Connection timed out))
15:52:26linuxstbI'm not saying it's definitely the best approach, but another approach would be to unify the metadata code even more, and have all the demuxing/container format code core to Rockbox, and the codecs would only actually decode the audio frames.
15:53:04linuxstbI don't see any problems with having to modify the core Rockbox in order to add support for new formats - we have already almost run out of new formats to add.
15:54:29linuxstbAlso, if there is no metadata parsing in Rockbox itself, how do we know which codec to load? (the Ogg/M4A problem).
15:59:33 Join edx [0] (
16:03:38]RowaN[has there been any progress on midi since i last asked.. erm, days and days ago =p
16:03:59***Saving seen data "./dancer.seen"
16:08:03 Join kurzhaarrocker [0] (
16:09:51 Join DMJC [0] (
16:11:53 Join edx__ [0] (
16:15:27 Quit ashridah ("Leaving")
16:28:18 Quit edx (Read error: 110 (Connection timed out))
16:28:55 Nick edx__ is now known as edx (
16:46:35 Quit Febs (" Want to be different? HydraIRC -> <-")
16:46:45amiconnlinuxstb: The core could "ask" each codec's metadata parser whether it recognises the format
16:47:15amiconnThis approach is in fact old; e.g. good old DeliTracker on Amiga uses it
16:49:26amiconnIf all metadata parsing is done in the core, adding a new format aways requires extending the core
16:50:08kurzhaarrockeryay! I finally managed to make the on button quit the recording screen and play the new recording.
16:50:18amiconnThen the whole modular codec architecture doesn't make much sense anymore
16:51:40LearDoesn't make much sense to load "all" codecs (or at least the metadata part) for each file either...
16:51:50preglowthat's what we're doing now
16:51:54preglowyou don't load it for each file
16:51:59preglowyou load all of them at the start
16:52:03preglowand never unload them
16:52:29Learso they're for all practical purposes part of the core... :)
16:52:50preglowwell, no
16:52:52preglowyou can remove them as you please
16:52:54amiconnThe metadata part needs to be present at all times
16:52:56preglowand add new ones as you please all metadata parser parts of all codecs need to be loaded at startup
16:53:17preglowafter they're loaded, then yes, it's almost the same, but i still think it's a nice distinction
16:53:29amiconnThis is still different from having them built into the core
16:54:25amiconnFor instance, if someone e.g. only uses mp3 and ogg vorbis, he could remove flac, wavpack etc codecs
16:54:50amiconn...and afterwards this installation wouldn't even show flac, wavpack etc files as supported
16:55:36preglowaltogether preferable
16:56:06preglowbut still, i wonder if the metadata parts might have to be totally distinct from the codecs, after all, several codecs share the same metadata format
17:06:18linuxstbIMO, there is too much similarity between all the metadata parsers to make a modular approach sensible.
17:07:15amiconnThere is *currently* much similarity, but this will most likely change with non-streaming formats
17:08:11amiconnIt may also change if we want to support container formats
17:08:18linuxstbThe streaming formats will *always* share much similarity - even if we add non-streaming formats.
17:08:43linuxstbThat's my argument in favour of keeping the metadata code unified - the metadata code handles the container format, and the codecs handle the actual decoding.
17:09:34amiconnThen we could simply link all codecs to the core as well.
17:09:56amiconnKeeping them modular is pointless without modular metadata parsing, imho
17:10:48linuxstbThe reason to keep the codecs modular is that they have relatively high memory requirements, and we don't need them all at the same time.
17:11:11linuxstbBut we do need all the metadata code available at the same time - to probe for file formats and decide which codec to load.
17:12:10preglowthis we'll find out after a while anyway
17:13:13preglowpersonally i'd really like if everything codec specific was placed in plugins
17:14:29kurzhaarrockerI want to break a gui design rule
17:15:30kurzhaarrockerWhen the user returns from adjusting a setting he normally expects to return to the menu where he entered chose the parameter from
17:15:59linuxstbpreglow: It depends how you define "codec specific". Are ID3 tags codec specific? Or parsing of an Ogg container?
17:16:49kurzhaarrockerBut not when he is in the recording screen. Quitting the recording screen shouldn't return to the menu but to the browser.
17:17:15kurzhaarrockerOr maybe to the wps screen, when he has resume enabled.
17:17:30amiconnkurzhaarrocker: I disagree here
17:18:01amiconnThe recording screen is entered from the menu. Everything else than exiting back to the menu would be really confusing
17:18:04kurzhaarrockerIt is pain in the a** to have to press n buttons to finally find the new recording.
17:18:18preglowlinuxstb: i meant more like anything relating to codecs, if a user doesn't ever use ogg files, it would be neat if he could just not include the ogg parser
17:19:40kurzhaarrockerreturning to the menu seems more logical, that's true, but it is not ergonomical.
17:20:00amiconnUnlogic behaviour isn't ergonomical either
17:21:01amiconnThe recording screen could provide a method to start playback of the latest recording, but leaving the ordinary way should always bring you back where you came from
17:21:08kurzhaarrockerI think that it isn't _that_ confusing because a recording takes a while. When I press stop after a recording I hardly know where in the menu I really was.
17:21:42amiconnRecording doesn't always take long, e.g. when recording voice notes
17:21:52kurzhaarrockerThat's the way I have implemented it right now, but it feels very complicated.
17:21:57amiconnBtw, that reminds me of that missing trigger mode...
17:22:37preglowbtw, am i the only one thinking the buffering algo should be configurable? i generally want rockbox to load the entire buffer fully when i press afile, not just a tiny bit, before resuming later
17:22:42amiconnYou could call the wps from the recording screen. Stopping would bring you back to the recording screen then
17:23:43kurzhaarrockerDirectly calling the wps is problematic as it could result in stack problems when you go wps -> menu -> recording -> wps -> menu -> recording ...
17:24:11amiconnOf course such circles need to be prevented
17:24:58amiconnPerhaps each screen should have a status variable, stating whether this screen is already in use
17:25:14kurzhaarrockerThus I made the on button _return_ straight down to the browser from the recording screen. (works even if you were in wps meanwhile)
17:25:20amiconnCalling it a second time would immediately return with an error messange (splash)
17:26:04kurzhaarrockerI abused the usb mode for that.
17:26:11amiconnThis way you could e.g. do wps->menu->recording->wps?MEEEP!
17:26:43 Quit hicks (Read error: 110 (Connection timed out))
17:26:43amiconnOr browser->menu->recording->wps->menu->recording?MEEP!
17:26:57amiconnNote: menu isn't a screen in this sense
17:27:54kurzhaarrockerNo when the menu->recording was chosen from wps it returns to wps - which immediately returns to the browser because of usb mode.
17:28:04Learamiconn: the context menu breaks that rule. Go to sound menu (from wps), then when you exit that, you drop back to the wps, and not the context menu.
17:28:56amiconnYes. The context menu is a bit different, in that selecting something in a submenu drops you back out.
17:29:09preglowmust one reboot for voice ui to take effect?
17:29:29amiconnEven that is slightly confusing, but acceptable
17:29:50preglowman, voice ui clips like mad on iriver
17:30:12 Quit tvelocity (Read error: 113 (No route to host))
17:30:43WyHrm. Hey, I'm kinda stupid, but anyone know where the firmware for the jukebox recorder is?
17:30:52WyLink on the site is broken
17:30:55 Join muesli- [0] (
17:31:15kurzhaarrockerLet's delete the link - who cares about archos firmwares :)
17:31:38linuxstbpreglow: I'm not sure what a user would gain by (e.g.) removing Ogg support.
17:32:18WyHeh, I finally found the time to dump my 80gb drive into my recorder
17:32:34WyBtw, anyone also know of a good way to, say ... synch the recorder to itunes?
17:33:32preglowlinuxstb: more space for other things than functions that will never be used?
17:33:44preglowlinuxstb: less time used in format checking?
17:34:50Slasheriif we don't have any metadata parsing inside the core, then we would need to scan and load all plugins to retrieve some info from them. And that will add very much complexity to the code
17:34:59linuxstbBoth are insignificant IMO. Are we short of space on the iriver?
17:35:20preglowSlasheri: i think 'very much' is exagerating
17:35:33Slasherii think it's not..
17:35:51Slasheriwe have to deal with all performance problems etc. also
17:36:05preglowit's not any worse than just including a few fixed size fields at the start of a plugin
17:36:27preglowor a struct
17:36:37preglowgood old linker magic
17:36:41kurzhaarrockerWe cannot play a song without it being in a playlist, can we?
17:37:05preglowbut anyway, this isn't very important right now, i suggest we take this up again once we have more of the code we're discussing
17:37:07Slasheriwe would need to run two plugins at the same time while parsing metadata and playing. If we also want to use voice ui, that will be complex enough
17:37:39preglowSlasheri: we were talking about making the parser code position independent and loading all of them at startup
17:38:06preglowSlasheri: the parsing would be pretty much seperate from the main codec action
17:38:15Slasheriah, that sounds better
17:39:12preglowstill, like i said, there are more important things to be done now, whether this is done one way or the other, it doesn't much touch how the user perceives rockbox
17:39:21preglowit's a code thing
17:39:48 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
17:40:02linuxstbSo what is more important at the moment? (I'm asking seriously - I'm looking for jobs).
17:40:14amiconnSlasheri: We don't need to run 2 plugins at once. This would indeed add complexity
17:40:21Slasherilinuxstb: Hmm, remote support? :)
17:40:29linuxstbI've lost my remote... :(
17:40:49Slasheriamiconn: true if we handle the metadata parsing outside from codecs (as it's now)
17:40:55amiconnIf we split the codec plugins in 2 (still within one file), the metadata parsers of all codecs would be loaded into the "metadata parser code buffer" at startup
17:41:32amiconnThen the core would just ask each parser if he recognises a certain format (simple loop across all parsers)
17:41:59preglowthe only thing we'd need to keep in the core is a set of codec identifiers to be passed around
17:42:23amiconnEach codec can be assigned a unique identifier at build time
17:42:30preglowa string would be nice, actually
17:42:54preglowso the different parts at least stand a chance to work across different compiles of rockbox
17:42:55amiconnI think a 4-char identifier would be sufficient
17:42:58preglowyup, agreed
17:43:11amiconnIt can be passed in a long, and still recognisable as a name
17:43:25preglowhmm, what asm restraint can i use for 4 byte constant?
17:43:30preglowconstraint, that is
17:43:56amiconnWhat do you mean?
17:44:21preglowi'm adding a flags parameter to mc5249_init_mac
17:44:45amiconnWhere do I find that?
17:44:47preglowwhich is inline, so i want the asm statement to put a constant where %0 is
17:44:51*amiconn is trying to multitask...
17:45:25Learpreglow: voice ui probably clips becase the samples are multiplied by 4...
17:45:28preglowlike, where #0x20 is now, i want %0 to be, which i want to be replaced by a constant, not a register, to make the tightest code. this might ofcourse not be possible :P
17:45:32preglowLear: yes, i know
17:45:49preglowLear: doesn't sound too clever ;)
17:46:00amiconnpreglow: Ah, you're looking for the 32 immediate constraint. Hmm.
17:46:32amiconnLear: Why the heck is this??
17:46:48*amiconn tries to find out from gcc source
17:46:51 Quit muesli- (Read error: 104 (Connection reset by peer))
17:46:55Learamiconn: ask slasheri. :) apparantly the volume was quite low otherwise...
17:46:56preglowamiconn: i think slasher said it sounded really weak without
17:47:32amiconnThe voice files aren't full volume, for a reason
17:47:32Slasheriyep, that was the problem
17:47:45amiconnHowever, multiplying by 4 will make them clip
17:47:50Slasheriwith some songs it was hard to hear the voice ui while playing
17:48:11amiconnAll voice files generated by me are scaled to 60% level (linear)
17:48:16preglowSlasheri: i think the music should be enveloped a bit while playing
17:48:21Slasherimaybe even multiplying with 2 would be enough
17:48:21preglowSlasheri: playing voice clips, that is
17:48:44preglowat least, that'd be a nice option
17:48:57Slasherihmm, decreasing music volume while playing voice samples? that is a great idea :)
17:49:35amiconnOriginally the voice files were generated at full volume, but for Bagder's sensitive ears ;-) (or was it Zagor?) the voice was too loud compared to the music, so we lowered it
17:50:01amiconnThat was on archos of course, where voice + music is impossible
17:50:31preglowSlasheri: a nice option would be how many decibels to lower music when voice is playing
17:50:43 Join webguest51 [0] (
17:50:46amiconnI think multiplying by 2 plus lowering music volume should be enough. Better no multiplication when no music is playing
17:50:50Slasheripreglow: true, that should be implemented :)
17:51:16preglowthink i'll try the i constraint
17:51:59kurzhaarrockerloud voices are good for commercials. "IM JAMBA MONATSPAKET!!" :)
17:54:58preglowno constraint match, hrmf
17:55:19preglowperhaps i should just make it a macro instead of an inline function
17:55:22preglowthat ought to make it easier
17:57:00preglowyup, that worked, i assume you guys don't frown on macros? ;)
17:58:17preglowSlasheri: there bug with a track change skipping back and forth between the previous track and the new track is still here, any idea why?
17:58:23amiconnInline function is preferred
17:58:39amiconnTry K or M constraint
17:59:11amiconnLook for "Immediate integer operand constraints"
17:59:14preglowamiconn: M wont do
17:59:19preglowamiconn: it's for 5 bit numbers
17:59:55preglow Integer in the range 0 to 32
18:00:25amiconnThe m68k machine description says different...
18:00:38preglowshit, i'm looking at the wrong arch
18:01:25preglowboth K and M set requirements to how large the number is
18:01:27preglowi can't use that
18:02:36preglowesp since the flags will always be lower than 0xf0
18:02:45preglowmight even be 0
18:03:16preglowdoesn't work anyway, i think it's gccs processing of inline functions that are causing the grief
18:03:24preglow#define coldfire_init_emac(flags) \
18:03:24preglow asm volatile ("move.l %0, %%macsr" \
18:03:24preglow : : "i" (flags));
18:03:28preglowthat works perfectly
18:03:43preglowif the inline equivalent doesn't work, it's not the constraints fault
18:04:01***Saving seen data "./dancer.seen"
18:04:40amiconnYou used a lowercase i ??
18:05:42preglowthe uppercase wont sure as hell wont work for m68k
18:05:46preglowthat's a number from 1 to 8
18:07:03preglowhmm, seems like it works if i make the argument const
18:07:30amiconnAh yes, and yes
18:08:29Slasheripreglow: Hmm, i will try that soon (no ideas yet)
18:10:25preglowSlasheri: i switch a track, i hear about a second of new audio, then the old one glitches in again, then i get the new one, and it stays
18:11:35 Join amiconn_ [0] (
18:11:49 Quit amiconn (Nick collision from services.)
18:11:50 Nick amiconn_ is now known as amiconn (
18:12:35 Nick paugh is now known as AliasCoffee (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
18:13:17Slasheripreglow: ah, that might be a performance problem. I think you have crossfade disabled
18:14:14preglowSlasheri: no, i haven't
18:14:25preglowSlasheri: i've never had it enabled, and i've pretty much always had this bug
18:15:03preglowSlasheri: ehh, disabled, yes, i have it disabled :P
18:15:11preglowi should actually read what people say to me
18:15:28preglowbut how can _disabled_ crossfade give a performance problem?
18:16:26CoCoLUS<fn~preglow> Slasheri: i switch a track, i hear about a second of new audio, then the old one glitches in again, then i get the new one, and it stays
18:16:30CoCoLUSi get that, sometimes, too
18:16:43Slasherithat is a problem with audiobuffer flushing (it tries to provide continuous playback always when possible). If there is no new data coming fast enough, part of previous data could be heard
18:16:49CoCoLUScrossfade not enabled
18:16:59Slasherii think we could tune that flushing a bit to get rid of the problem
18:18:24preglowthat would rock
18:18:34CoCoLUSwhat bothers me the most about the current playback system is that it's impossible to skip a few tracks instantly
18:19:41CoCoLUSrepeatedly pressing next/previous track, that is
18:20:20CoCoLUSit works perfect for the next one or two tracks, but after that, it loads a few seconds for every track, and it won't accept any new button presses
18:20:31preglowamiconn: there, i've removed all but one emac init in each codec, seems to work just fine and dandy
18:26:31amiconnNice :)
18:26:39amiconnNow for the iram usage part... ;)
18:34:42 Join tvelocity [0] (
18:40:49kurzhaarrockerfunny. It seems to be a bad idea do declare a variable in the *.h file as "extern char* filename;" and then define it in the *.c file as "char filename[MAX_PATH];"
18:41:04kurzhaarrockerI always thought that both were the same.
18:41:46amiconnThey aren't
18:42:21amiconnThe former is a pointer, the latter is an array
18:44:03kurzhaarrockerI thought that char arrays internally were the same as a pointer to a char. I thought wrong
18:44:39amiconnThere are several differences
18:45:14preglowamiconn: i already did some layer12 work, but i've got a hard time measuring the improvements, need to code a quick realtime usage indicator first
18:45:54amiconnA pointer to a char always takes 4 bytes, which store the address of a string. No space is reserved for the string by the pointer declaration itself
18:46:02 Quit edx ()
18:46:46amiconnA char array takes as many bytes as you give it within the []. The address itself isn't stored, it is implicit
18:47:22amiconn(known symbol at compile time, resolved at link time, fixed at runtime)
18:48:03kurzhaarrockerSo it is the compiler that automatically converts when you assign a char array to a char pointer?
18:48:18amiconnSo char hello[] = "Hello" takes 6 bytes, while char *hello = "Hello" takes 10
18:48:48amiconnIt doesn't convert, it just writes the (fixed) array address into the pointer variable
18:50:21kurzhaarrockerSo char *hello = "Hello" does both: generate an anonymous array and a named char pointer.
18:50:34amiconnIn my 2 examples, the former declares the array 'hello' whose address is fixed, while the latter declares the pointer 'hello' which initially points to the string "Hello"
18:51:22 Quit webguest51 ("CGI:IRC")
18:54:09amiconnBleh, "panic: rec upd: -7" :-(
18:54:27amiconnObviously there's something wrong with my new xing header creation code...
18:55:53preglowi'm including system.h in codec.h
18:56:52preglowamiconn: is the voice ui any closer to archos behaviour with the new iram stack, btw?
18:56:54amiconnMrf, this stand under not do I
18:59:22*amiconn is silly
18:59:53amiconnOf course this can't work... I try updating the new file (after the split) which does not yet exists at that point....
19:00:43*preglow gets scared by the static
19:01:37amiconnopen() return value -7 means ENOENT...
19:02:11*preglow loves bugs that appear ,then disappear
19:02:44 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
19:02:46amiconnBlinkenbugs... I know this evil type :/
19:03:33preglowsomething's wrong here
19:04:12 Quit tvelocity (Read error: 113 (No route to host))
19:09:03kurzhaarrockerENO ENT? Is that a brother of Treebeard?
19:09:22amiconnHaha, no. ENOENT means Error NO ENTry
19:09:27amiconn(in the directory)
19:11:23 Quit cYmen ("wusch")
19:14:07 Join [1]Moos [0] (
19:14:08 Quit Moos (Read error: 104 (Connection reset by peer))
19:14:11 Nick [1]Moos is now known as Moos (
19:15:38 Join nc [0] (
19:16:11nchi all.
19:16:52ncis the 2.5 rockbox to be released tomorrow? does anyone know?
19:18:29 Join wubbla [0] (
19:18:34wubblaahoi! :D
19:18:53kurzhaarrockerPersonally I'd like linus (or someone else) to apply a patch that fixes a ui problem first:
19:21:46nci thought they did not tale any more requests since the 2.5 freeze.
19:22:21kurzhaarrockerLinus agreed that this specific patch is more a bug fix than a new feature.
19:22:25preglowSlasheri: do both codecs run in seperate threads?
19:22:32 Join OnkelJonas [0] (
19:22:34wubblaare there any new successes concerning the h300 port?
19:22:58ncanyone here tried any previos versions of rockbox? I am new here and own a iHP-140 and wonder whether i should give 2.5 a go.
19:23:13preglownc: that wont work on iriver
19:23:34OnkelJonasjust get a "daily build"
19:23:38kurzhaarrockerrockbox v2.5 won't be an official releas for iriver
19:23:42amiconnpreglow: Yes (your qn to Slasheri)
19:24:01preglowamiconn: i'm getting audio corruptions sometimes now
19:24:06preglowbut only with voice ui
19:24:11ncoh, just for Achos? and then they might make it for iRiver as well?
19:24:25OnkelJonasnc: <−−- that will work on the iriver (h140 too)
19:25:26preglowamiconn: ignore me, it's probably my fault
19:25:27OnkelJonasyou need to install the firmware first though - looking for the link
19:25:42kurzhaarrockernc: rockbox for iriver still is considered 'alpha'
19:26:02 Quit matsl (Read error: 110 (Connection timed out))
19:26:09ncyou mean version 2.4 is alpha?
19:26:52OnkelJonasthe iriver port is alpha
19:26:54kurzhaarrocker2.4 is a release for the _archos_ players. The port to the iriver players is still in progress.
19:27:45OnkelJonasbut very stable, and no dead players yet afaik
19:28:07amiconnNah, not *very* stable, at least not for me
19:28:29amiconnIt's somewhat stable, but I know numerous ways to make it freeze, reproducably
19:28:45ncdid it cause any trouble to your player amiconn?
19:29:03amiconnYou can always reset
19:29:15ncso u recommend trying it out?
19:29:38amiconnIn fact I help developing it :)
19:29:49ncok. and i can always go back to iriver's firmware, right?
19:29:57OnkelJonasyou can easily boot to iriver firmware, and in my experience "normal" usage rarely crash it now
19:30:32dpassen1i haven't booted up the iriver firmware in months
19:30:40 Join cYmen [0] (
19:30:48ncthat well,e? :)
19:31:26OnkelJonasi boot it occasionally for the 1 album i have in wma format, and for radio
19:31:43ncso tomorrow we will have the 2,5 firmware. does anyone know when the iriver port will be ready for 2.5?
19:31:45OnkelJonasbut there is even radio in rockbox now, although still pretty basic
19:33:11ncIf i get this right there is a global firmware and specific ports for each player? (archos,iriver)?
19:34:18OnkelJonasa lot of the code is shared across players, but since they have different hardware it is not compatible
19:34:30kurzhaarrockerall those players have completely different hardware that must be adressed differently.
19:34:31OnkelJonasso you download for a specific player
19:35:55ncok. so there 's no need to wait for 2.5 for iriver since it would take a while for a port to be available, and i should go for 2.4.
19:36:16kurzhaarrockerThere's no V2.4 for iriver
19:36:32ncerrrr. i am a little confused
19:36:49kurzhaarrockerThere's no official version for the iriver at all.
19:36:54OnkelJonasthere is no "V XXX" for iriver since it is still in development
19:36:56kurzhaarrockerDevelopment is in progress
19:37:11OnkelJonasso you should get a "daily build"
19:37:37kurzhaarrockerWhen rockbox V2.6 is released it may be for the iriver too.
19:38:17preglowi wonder why vorbis has suddenly started crapping its pants
19:38:45ncoh i see. so i am just an average user. do u recommend i try the daily use and keep in touch w/ he project?
19:39:02ncsorry for all the questions
19:39:20nci am trying to understand what is going on.
19:39:53OnkelJonassure... personally I check the page for recent development ~once a week, and get a new build
19:40:13kurzhaarrockerYou can try the daily builds, but expect errors and make sure you backup important data on your iriver first.
19:41:03 Quit cYmen ("leaving")
19:41:06ncthe worst that can happen (from user experience ) is data loss only? and as i was told the reset button help out.
19:41:35ncif that's the worse that's happened so far i am willing to test it
19:42:07OnkelJonasevery report of dead players I've seen so far turned out to be solvable by simpæly using reset
19:42:09amiconnAhahaha, discovered an old bug in mpeg.c: a missing break at the end of a case:
19:42:24RickDon'tcha love those?
19:42:46amiconnIntroduced by me (tm) 3 months ago...
19:43:13ncthanx guys i am going now to try rockbox. :)
19:43:18RickHeh heh heh
19:43:21Ricknc: Good luck ;)
19:45:10OnkelJonasso when do we tell him that rockbox will fry his player at first boot?
19:45:22OnkelJonas:P i keed i keed
19:46:36kurzhaarrockernc: make sure you wear safety glasses and gloves when you try rockbox first.
19:47:19ncyou guys
19:47:44ncdoes is support any gnuboy games btw?
19:47:51preglowamiconn: i don't get this, i actually need to have more than just one emac init in vorbisfile.c for it to work
19:49:22OnkelJonasit supports gameboy games is all i know, but dont expect all of them to run at full speed
19:50:29nctetris and snake is all i need :)
19:50:38nc(for beginners)
19:51:03 Join tvelocity [0] (
19:51:24HCli need to do some optimizations for gnuboy
19:51:26kurzhaarrockerpah. snake. Play wormlet! More worms! Artificial stupidity! Aarghs!
19:51:28HClbut i'm not sure how :/
19:51:52preglowhcl: optimised cpu core, find variables to iram
19:51:54amiconnpreglow: That's strange. Is there any codec code that is called outside the codec thread (or voice thread, for voice codec)?
19:52:18HClpreglow: first i'd like to add asm functions of certain optimizable functions
19:52:22amiconnkurzhaarrocker: wormlet still runs on recorder only
19:52:22preglowamiconn: seems its only needed one place, i placed it in the first function i saw vorbis.c call, but that didn't cut it, it _needs_ to be placed in ov_read_fixed
19:52:27HCli believe amiconn already did this for archos
19:52:42kurzhaarrockerWhat a shame, amiconn.
19:52:57kurzhaarrocker<- can't fix that, doesn't have any irivers
19:53:23amiconnI was unable to adapt wormlet to variable button assignment. Wormlet's button handling is scattered all over the code...
19:53:45amiconnIt would probably be easier to rewrite wormlet than to adapt it...
19:54:35OnkelJonasis bmp support broken in recent builds?
19:54:35OnkelJonasi just installed the latest, and the wps's show up without bmp's
19:54:59solex_no, works here
19:55:01dpassen1the syntax has changed
19:55:36OnkelJonasprobably that... is it in the wiki?
19:55:44linuxstbAnyone know if there are byte-swap macros centrally defined in rockbox - e.g. to convert a 32-bit int from big-endian to host-endian?
19:56:13preglowoh yes
19:56:30preglowin system.h, afaik
19:56:49preglowSWAB32, for example
19:56:57amiconnYes and no
19:57:24amiconnThe byte swap macros in system.h are for converting little-endian to host-endian
19:57:38amiconnThere are no macros to convert from big-endian to host-endian
19:58:20linuxstbI need them for my .m4a code (alac today and maybe aac in the future). Would it make sense to add them?
19:58:34amiconn...and the SWA??? macros work on variables only, no unaligned read etc
19:58:53 Join cYmen [0] (
19:59:00ncdo i get the latest build for iriver here: ?
19:59:20linuxstbMy code currently reads 4 bytes from a file into an integer, and then byte-swaps it if necessary (which is only necessary on little-endian simulators)
20:00:33 Quit OnkelJonas (Read error: 104 (Connection reset by peer))
20:01:20kurzhaarrockeryes, nc
20:01:42linuxstbBut the macros seem to be assember-optimised. Does that mean they wouldn't work in a Simulator?
20:02:04 Join OnkelJonas [0] (
20:02:40Learpreglow: dsp uses emac too, probably uses a different mode.
20:03:25Learlinuxstb: a simulator version should be around.
20:03:29amiconnLear: Yes, but dsp does run in a different thread, doesn't it?
20:03:37kurzhaarrockerDo patches only appear on the rockbox site when you do not forget to specify the category?
20:03:50Learnope, codec calls code to insert in pcmbuf, no context switches there...
20:03:54preglowLear: well, macsr is now saved in the thread context
20:04:05preglowLear: and dsp runs in a different thread than the codec, no?
20:04:06***Saving seen data "./dancer.seen"
20:04:10Learbut dsp is run in codec context...
20:04:25 Join [MuIL]SpongeBoB [0] (
20:04:32preglowwhy would dsp need anything other than frac mode?
20:04:45Leardidn't you write the resampler?
20:04:50preglowi did
20:04:53preglowand it uses frac mode
20:04:56[MuIL]SpongeBoBhey sorry for the dummy question. I have Iriver H10 what version of Rockbox do I need?
20:05:04preglow[MuIL]SpongeBoB: not supported
20:05:06Learwell, it does use saturation mode, iirc...
20:05:14preglowLear: as does all other codecs now
20:05:22preglowLear: i see no reason for not using saturation
20:05:25LearI see...
20:05:31preglowonly thing varying is rounding
20:05:43preglowwhich tremor can't use because of MULT31_SHIFT15
20:06:01preglowhalf the reason for the macsr context save vanished
20:06:04Leardsp uses #b0 mode...
20:06:18preglowb0 is frac and round
20:07:12Learstandard mac init (in system.h) is 0x20, dsp uses 0xb0...
20:07:15 Join muesli- [0] (
20:07:32Learyou know what works in dsp.c, I guess... :)
20:08:08linuxstb[MuIL]SpongeBoB: Rockbox doesn't support the H10. Only H120/H140 at the moment, with work just starting on H300
20:08:43 Quit [MuIL]SpongeBoB (Client Quit)
20:09:41linuxstbLear: I can't find any simulator versions of SWABNN - I think they are only used in the low-level firmware code, which isn't in the sims.
20:11:00LearI'm pretty sure it is, at least for little endian platforms (i.e., x86)...
20:11:07nci just installed Rockbox on my iHP-140!!! :) :) and i like the interface a lot!!!! thumbs up for developers and everyone that helps.
20:11:36linuxstbFor little-endian, it's just #defined as SWAB32(x)=x
20:12:18linuxstbHas anyone ever compiled a simulator on a big-endian machine?
20:12:36Learperhaps not... :)
20:12:40preglowtea everywhere
20:13:11Learlinuxstb: you on ppc?
20:13:55linuxstbSometimes. I sometimes use an iBook running Mac OS X.
20:14:21Learwell, should be easy enough to convert the iriver swab32 code to C...
20:15:34linuxstbThere is already a C definition, but #ifdef'ed for the TCC730 processor.
20:15:42 Join Mark_ [0] (
20:16:12linuxstbBut "swab" (swap adjacent bytes) doesn't seem to be the correct name for the function - if it only works for little-endian to host-endian conversions.
20:16:19kurzhaarrockerYour iriver didn't explode, nc?
20:20:35 Join matsl [0] (
20:20:46 Part OnkelJonas
20:20:59 Join OnkelJonas [0] (
20:22:39amiconnWee, enhanced xing header generation already works quite well. Some minor quirks still to be fixed...
20:25:02preglowwhy does dsp run in codec context anyway?
20:25:27muesli-amiconn dont you watch the chancellor duell? ;)=
20:27:09amiconnmuesli-: Why should I?
20:27:13 Join silencer_ [0] (
20:27:33muesli-interesting if you ask me
20:28:08muesli-wont change my dicision
20:28:33muesli-but am curious ;)
20:29:09Learwhy do a context switch just to write data to pcmbuf?
20:29:54Learso why not let dsp run in codec context (apart from the macsr thing)?
20:33:19preglowit would make more sense having it in the audio thread
20:33:41preglowSlasheri: got anything to say on this?
20:35:01Learwouldn't that just complicate things, and add processing (first copy data to pcmbuf, then run dsp over it, keeping track of what is processed and what isn't)?
20:35:11SlasheriLear: Hmm, we are making a context switch to write data to pcmbuf?
20:35:36Learslasheri: no, were're not, and that's what preglow is "complaining" about...
20:35:43Slasherioh, ok
20:36:39Slasherii think the current aproach is quite performance effective
20:42:55 Quit kurzhaarrocker ()
20:42:55 Quit AliasCoffee ("bbl")
20:44:27 Quit nc (Read error: 113 (No route to host))
20:46:07 Join Mxm`Pas`Bien [0] (
20:46:08 Quit Maxime (Read error: 104 (Connection reset by peer))
20:51:11preglowSlasheri: hmm, yes, but it kind of defeats what we've done with the macsr context save
20:51:30preglowdsp should at least save it so the codecs don't have to worry about it being clobbered
20:52:18Slasheriyes, that's true
20:55:03 Quit muesli- (Read error: 113 (No route to host))
21:00:09linuxstbDoes anyone understand this bug (from ReleaseTodo): "The %F WPS conditionals don't work as expected since the enum change"
21:01:32 Join [1]Moos [0] (
21:01:53 Quit Moos (Read error: 104 (Connection reset by peer))
21:01:53 Nick [1]Moos is now known as Moos (
21:03:14 Quit ansivirus (Remote closed the connection)
21:07:06linuxstbSlasheri: In the read_next_metadata() function in playback.c, the codectype is set back to 0 (after a call to probe_file_format sets the value). Do you know why that is?
21:10:45dpassen1linuxstb: i understand it, i believe
21:11:30linuxstbExplain :)
21:11:44dpassen1there was a CVS change, i'm looking for it
21:11:49 Join webguest28 [0] (
21:11:59dpassen1and since, the next file (capital F) in the WPS does not work as it should
21:13:24linuxstbYes - the %Fc tag no longer works for the next track info - because the codectype has been set to zero.
21:13:59linuxstbI think the playback code uses zero to indicate "codec not loaded", but the rest of rockbox uses it to mean unknown codec.
21:16:16Learmaybe wps-display should use id3->codectype + 1 for the codectype intval then...
21:17:52linuxstbI can't think of any alternatives.
21:18:20 Quit Sucka (Read error: 110 (Connection timed out))
21:19:21linuxstbThe relevant change in playback.c is here:
21:19:22amiconnlinuxstb: %fc does work?
21:19:51linuxstbYes - %fc (current track) seems to work for me, but %Fc doesn't - it always displays "???".
21:20:04amiconnAh, yes
21:20:31amiconnThis must be caused by the conversion to multi-conditionals (a la switch())
21:20:54linuxstbThe "bug" was introduced with the change I've just linked to.
21:21:11amiconn...and by the fact that one field in the trackinfo is used for 2 purposes
21:21:31amiconncodectype has nothing to do with the load status, imho
21:21:36linuxstbI agree.
21:22:11linuxstbBut I'm curious if there is another bug as well - why did Linus put this issue on the ReleaseTodo page?
21:22:12Slasherilinuxstb: hmm, i think that zero as codectype means that file is not yet buffered at all
21:22:24Slasheriso it's important to keep that assigment
21:22:25amiconnAlso, the codec type is metadata supplied by the parser, hence to be stored in tracks[trackno].id3.codectype (like it is)
21:22:43linuxstbamiconn: Exactly.
21:22:48amiconn...but the codec load status is playback engine data and has nothing to do with the id3 sub-structure
21:23:37amiconnShould be something like tracks[trackno].codec_loaded or so, as a boolean
21:23:57Slasherithat was also why codec type zero meant "ERR" but now it's just "unknown codec"
21:24:56*tvelocity eimai gay
21:24:58amiconnEven with this assignment %Fc wouldn't work right
21:25:39linuxstbIn what way?
21:26:12*tvelocity never leave your keyboard exposed in public...
21:26:13Learno, get_tag in wps-display should return null for codectype 0, I think...
21:26:32 Join San [0] (n=Test@
21:26:34amiconnThis would be weird
21:26:47amiconn...and %Fc still wouldn't work
21:27:25amiconnAt least it wouldn't work at the same time as the other next_track tags
21:27:31Learwould display nothing rather than "???" at least. :)
21:27:46linuxstbIt's impossible for a track being played to have an unknown codec.... It's more like "Not yet known", and only applicable to the next track.
21:28:04linuxstbSo %fc will always have a valid value.
21:28:22amiconnI'd rather change it the way I suggested earlier
21:28:39amiconnThe current method mixes playback engine status info with metadata info
21:28:47linuxstbamiconn: I agree 100% with that.
21:29:37LearAh, playback.c sets codectype to 0 just after reading the info for the next track, wonder why...
21:30:05linuxstbbecause codectype is used for multiple reasons...
21:31:10Learyes, as an indicator that the codec is loaded...
21:33:58 Quit DangerousDan (Read error: 104 (Connection reset by peer))
21:36:31 Join silencer [0] (
21:36:35 Quit silencer (Client Quit)
21:37:41preglowi think i'll merge upsample and downsample
21:38:21preglowmight one of the arrays to be resampled be in iram?
21:39:55Learyes, resample buf is, buffer from codec is likely to be...
21:40:17Lear(unless it has been converted, in case it is known to be in iram)
21:40:18preglowso there might be some small point in letting downsampling be in-place
21:41:02LearI don't see it... :)
21:41:08preglowiram accesses are fast
21:42:43Learand the resample buf is in iram, so you don't gain anything...
21:43:07linuxstbI'm not sure how many codecs put the pcm data in IRAM. FLAC does, but what about mpa and vorbis?
21:44:28linuxstbAC3/A52 has it in IRAM. WAV doesn't (but could easy be changed to).
21:45:01linuxstbWAVPACK does.
21:45:26Learmpa and vorbis too.
21:45:53amiconnSomeone should check out the extended wav codec patch and commit it
21:46:08LearI plan to, after feature freeze.
21:46:13amiconn(Someone with some test data and/or some time to do it)
21:46:23Lear(Because then I plan to change metadata.c a lot)
21:46:35linuxstbI have some time, but no test data. But I've read the patch, and it seems to be good.
21:46:35LearAnd I have created a bunch of test files using sox...
21:47:22linuxstbpreglow: It looks like the answer is that the PCM buffer passed to pcm_insert() is always in IRAM.
21:48:00linuxstbWith the exception of WAV (at the moment), and my ALAC decoder - but I may be able to fix that.
21:48:59 Quit Moos (Read error: 104 (Connection reset by peer))
21:49:10 Join Moos [0] (
21:52:55 Join XavierGr [0] (
21:55:48 Join DangerousDan [0] (
22:00:52 Quit matsl (Remote closed the connection)
22:01:22preglowperhaps i should try making downsample in place, then
22:01:48preglowbut anywho, need to finish this first
22:03:08 Join TCK- [0] (
22:03:21 Quit Hadaka ("great things are afoot.")
22:04:07***Saving seen data "./dancer.seen"
22:07:02 Quit XavierGr ()
22:09:01 Join [1]Moos [0] (
22:09:01 Quit Moos (Read error: 104 (Connection reset by peer))
22:09:04 Nick [1]Moos is now known as Moos (
22:19:31 Quit TCK (Read error: 110 (Connection timed out))
22:29:08 Quit San ()
22:37:01 Quit dpassen1 ()
22:37:03 Join muesli- [0] (
22:38:25linuxstbpreglow: Do you have any way of measuring codec performance?
22:40:49LearI have a hack for vorbis... :/
22:41:10linuxstbTo do what?
22:41:22Learmeasure codec performance
22:42:21linuxstbHow does it work?
22:43:45LearIt's based on the old xxx2wav files, so I just took the vorbis codec, made it a plugin, added some stuff to time things and removed the stuff to write the data to the pcm buffer.
22:45:20linuxstbThat's what I was thinking. But there must be a better way.
22:46:30 Quit Lear ("Chatzilla [Firefox 1.0+/undefined]")
22:46:59 Join bagawk [0] (i=1000@unaffiliated/bagawk)
22:53:37muesli-hi t0mas
23:13:48 Quit Zagor ("Client exiting")
23:14:05]RowaN[guys is it possible in theory to build a kinda of connector into an h120 so that you can replace the hd with an sd card, or smaller form factor hd?
23:14:16t0masgood night :)
23:14:37t0mas]RowaN[: I think there is... but I don't know if the controller can handle it
23:15:02]RowaN[the controller is some kinda of hardware controlled device?
23:15:13muesli-rowan..there are max 4gb of sc-card space possible..
23:15:21muesli-dunno if that will make sense
23:15:41]RowaN[what do you mean possible, what is limiting it?
23:16:14muesli-reckon there no bigger than 4gb on the market
23:16:36t0mas]RowaN[: jup, the controller is a hardware device
23:16:38]RowaN[eh? ive seen 12gig sd cards, well read about them
23:16:40t0masit "controls" the disk
23:17:05t0masand it should be able to handle some memorycards if they're plugged in a memcard->IDE adapter
23:17:28crwlCF cards should work without problems, no?
23:17:47]RowaN[in a few years time i will replace the hd with a 100gb card then hehe, will be nice and cheap
23:17:48crwlbecause their interface is essentially ATA
23:18:03]RowaN[i'd like to think a better mp3 player would be out by then, but its doubtful, the way things are going
23:18:41muesli-12gig sd lol..are you donald trump? ;)
23:19:00]RowaN[i didnt say i'd buy one!
23:20:02]RowaN[not sure the date of that article tho, should be cheaper now
23:20:08muesli-maybe i could get discount if i take 3 ^^
23:20:22]RowaN[ah March 18, 2004
23:21:38]RowaN[whats the name of that theory thing which predicts the rate at which processor speed will increase over time
23:21:46]RowaN[some guys name
23:21:52zemoore isn't it?
23:21:56]RowaN[ah yea
23:21:57zemoore's law i think
23:22:08zewhich isn't a law, just an observation
23:22:15]RowaN[u know in a few years time you'll be going.. 12gig flash player? thats crap, ive got a 40tb hd player
23:22:38zeand many think its actually doing more to dictate the rate these days, than really predicting it
23:22:44]RowaN[its a prediction, based on an observation
23:24:27 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:24:49]RowaN[bah stupid cpu overheating
23:24:58]RowaN[WHY did i change a perfercly good fan/heasink!!
23:25:27wubblaare there any new successes concerning the h300 port?
23:25:54]RowaN[3ghz pentium now idles at 60.C/ (150 F), and shoots up to 74.C when i run someone cpu intensive
23:26:47OnkelJonasif music is still released in the same 'quality' as today, having a 40TB player is a joke
23:26:59muesli-moore's law its about the development of computing speed and not about prices :p
23:27:15]RowaN[a 20tb partition would be used for porn =p
23:27:27]RowaN[i didnt say it was about prices
23:27:48]RowaN[it just popped into my head when i was thinking about prices
23:27:52OnkelJonassure ... but then it would have to be a video player to make sense
23:28:05]RowaN[it would have a projector lens in it =]
23:28:13]RowaN[theres probably a similar law about prices
23:28:13OnkelJonasunless you just have to carry 20TB of pr0n on you at all times
23:29:08bagawkpike, Hello :)
23:29:09OnkelJonasdepending on your school of thought, prices are either determined mainly by supply/demand, or by monopoly pricing schemes...
23:29:23]RowaN[is there a "codec" so u can play sacd rips on a computer?
23:29:36pikebagawk ;) it's based on libmad
23:29:39]RowaN[i call it, ROWANS LAW =]
23:29:42OnkelJonaswhen there are monopolies, manufacturing costs doesnt matter wrt product pricing :(
23:30:10OnkelJonas"the stuff ]RowaN[ wants has to be cheap" :P
23:30:24]RowaN[if only that was law
23:31:43]RowaN[<muesli-> moore's law its about the development of computing speed and not about prices :p
23:31:44OnkelJonasROFL... 40TB is more than 50.000 albums without any compression
23:31:45]RowaN[yes it is...
23:31:52]RowaN[Moore's law is the empirical observation that at our rate of technological development, the complexity of an integrated circuit, with respect to minimum component cost will double in about 18 months.
23:32:38]RowaN[you mean in wav format? future music will take up more space, it'll be like, HD, 20 channel surround
23:32:45OnkelJonasi dont think thats the cost in $ - but the "cost" in no. of components/squareinch
23:33:11]RowaN[yes, cost
23:33:18OnkelJonasyou will still be dreaming that sweat dream when you sit deaf in your retirement home
23:33:42]RowaN[sweet dreams for me
23:34:31preglowlinuxstb: no way yet, no, i'm kind of busy rightnow
23:35:14]RowaN[so anyone know about SACD.. is there a codec, does anyone release rips? not interesting in sacd warez just interested to know hehe
23:35:25linuxstbpreglow: No problem. Are you planning something?
23:35:58OnkelJonasspeaking of SACD (which afaik is dying slowly but surely)... does any of the big companies have plans for future audioformats atm?
23:36:00linuxstb]RowaN[: AFAIK, computer CD/DVD drives can't read SACD data.
23:36:28]RowaN[yup they cant, but surely optical output of a sacd player could go to pc
23:36:35linuxstbDVD-Audio can now be ripped (using a hacked version of a Windows player).
23:36:57zeyeah when your music is lossless 32bit 192khz 16-channel 3D surround, thats a lot less than 50000 albums
23:36:59linuxstb]RowaN[: Optical outputs of high-resolution audio players (SACD/DVD-Audio) are encrypted.
23:37:03OnkelJonasdoes dvd-audio have a future (besides its niche appearance)
23:37:28linuxstbYou tell me :). I personally like it because I can put 4.5GB of 44.1KHz WAV files on a single disc.
23:37:28]RowaN[since when did encryption ever stop anyone?
23:37:29OnkelJonasbut who will make those recordings ze?
23:37:37zeOnkelJonas: who won't?
23:37:53linuxstbOnkelJonas: They have already been made - on high resolution analogue tapes.
23:38:04ze24bit 96khz/192khz is already the internal digital standard in the industry
23:38:23linuxstbBut a lot of people argue that 44.1KHz/16-bit is already good enough for almost all humans.
23:38:29OnkelJonassure... but do you seriously see them being published outside some rare niche stuff any time soon?
23:38:36zeand yeah analog recordings can always be digititized to any depth and rate
23:39:01zelinuxstb: there's well known advantages to 24bit
23:39:08zevs 16
23:39:16linuxstbThe consumer seems happy with MP3/AAC at the moment, so it seems unlikely.
23:39:16zehigher sampling rates are more arguable
23:39:49 Join [1]Moos [0] (
23:39:59zebut lots of people argued that nobody needed more than 16bit color too
23:40:01OnkelJonasbesides... considering the "quality" of audio equipment in the average home its unlikely to make much of a difference anyway
23:40:07zebut just try looking at 16bit gradients
23:40:14 Quit Moos (Read error: 104 (Connection reset by peer))
23:40:14 Nick [1]Moos is now known as Moos (
23:40:30]RowaN[u guys heard of XRCD ( its not a new format at all, just a well mastered CD.. kinda like the CD equivalent of a "SuperBit" DVD movie
23:40:39]RowaN[average homes? yuk!
23:40:56]RowaN[average people shouldnt have a say in the future of music quality =p
23:41:32zei'd hope the quality of equipment improves by the time we reach 40TB
23:41:38]RowaN[well, you have to assume its well mastered.. u pay your money then find out
23:41:48OnkelJonashahaha... XRCD = teh funny
23:42:07zemusic should be well mastered whatever format its on
23:42:18linuxstbI'm sure the recording industry hate CDs because they are not "secure". So they are happy to push new formats - as are hardware manufacturers.
23:42:23]RowaN[should, i love that word
23:42:39OnkelJonasyou gotta admire an industry that can make crappy mastering the standard, and then release properly mastered CDs for a premium :D
23:42:58]RowaN[thems the breakz kids
23:43:27zeall the while telling you how important it is to get your stuff mastered by a professional
23:43:38zeand how virtually impossible it is for you to do a decent job of it yourself
23:44:30]RowaN[cds used to say AAA, AAD, DDD or whatever on them to indicate a snippet of the mastering process.. dont tend to see that these days
23:44:50]RowaN[or is it just on classical stuff?
23:45:00]RowaN[wow heh
23:45:25 Quit Moos (" HydraIRC -> <- Try something fresh")
23:45:56zeyeah most cd's released in the 80's had that
23:46:40]RowaN[maybe at some board meeting they had the idea that telling the consumer how sloppy the mastering was might adversly affect sales
23:47:37OnkelJonascan you get lossless in the itunes store now?
23:47:55]RowaN[i think so, not that i'd ever use itunes
23:48:12]RowaN[apple have their own lossless format dont they
23:48:25zeit wasn't about sloppyness, it was just indicating what domain each part of the process took place in
23:48:49]RowaN[AAA is pretty sloppy aint it
23:49:01]RowaN[compared to DDD
23:49:06OnkelJonasif itms starts pushing lossless (hopefully at a minimal premium) more people will start caring about it
23:49:06zewhat're th 3 parts?
23:49:10zerecording, mastering, and ?
23:49:15zeor is it recording, mixing, and mastering?
23:49:21]RowaN[no idea
23:49:25crwlrecording, mixing and mastering
23:49:36zelots of people still do it all analog
23:49:50]RowaN[jesus, we're not missing out on anything with XRCD, unless you like tony bennet or obscure jazz
23:50:03crwl(never seen AAA cd's though - is it possible to do a CD mastering process in an completely analog way?)
23:50:03zeamong audiophiles and purists (which make up a lot of the recording industry), "digital sucks"
23:50:15OnkelJonasnonono... its about the mastering process: Aspirin, Aspirin, Doughnut
23:50:22zecrwl: sure
23:50:27OnkelJonasmore As = more problems
23:50:47zeanalog isn't inherently problematic, and digital isn't without its issues
23:51:18OnkelJonasthere must be a digital stage at the end...
23:51:41zemastering is just adjusting levels, frequency content, and misc stuff like that, of a final mix, to try to get it to be as consistent as possible on a wide range of playback equipment
23:51:52zeor to maximize volume, or really numerous project-specific goals
23:52:34zefor mainstream music, there's lots of dynamics compression, etc... for classical, very little
23:52:46OnkelJonastrue - but there will still be a degradation of quality (/amount of audio information left from original) when moving through several analog stages, something that doesnt have to be the case with digital
23:52:47zebut it can quite easily be done in the analog domain
23:53:17zeand then the results can be digitally sampled
23:53:28zeOnkelJonas: not necessarily, and it depends on the equipment
23:53:31zeand the processes
23:53:31 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
23:54:00OnkelJonashow are you going to move an analog signal through a cable without loosing information?
23:54:26zeshort cables, proper impedance matching
23:55:09zeand keep in mind the goal of music isn't ultimate fidelity, which is often undesired as it sounds rather sterile, but being humanly percieved as sounding good musically
23:55:13OnkelJonasthere will still be some loss - of course it can become an academic issue since it can be inaudible amounts, but still...
23:55:20OnkelJonasagreed :)
23:55:39zea lot of people prefer analog precisely for the ways in which it distorts the signal
23:55:40]RowaN[lets not fight lets all be friends =p
23:55:54zethey call it 'warmth'
23:56:06]RowaN[a dsp could add that warmth heh
23:56:17zeand there's a lot of people doing a lot in the analog domain before transferring to digital just to get that warmth
23:56:50zethere's a lot of things that digital processing isn't very good at that's trivial in an analog circuit
23:57:13zeits not that easy to digitally emulate the analog warmth of a particular circuit
23:57:23zeyou have to account for a lot of factors
23:57:24 Join TCK [0] (
23:57:26OnkelJonasIMO "mastering" is all about changing the audio to whatever the endproduct should be like... so if that means feeding it through a dead crocodiles hindlegs I dont give a s###
23:58:08zeOnkelJonas: right
23:58:31zeyou have to also realize, the more processing you do to a digital signal, the more you risk aliasing and various digital oddities like that

Previous day | Next day