June 24, 2008  
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30

[00:21:06] *** kb1ooo has quit IRC
[01:06:02] *** prophile has quit IRC
[01:21:02] *** barra__out has joined #openal
[01:21:03] *** ChanServ sets mode: +v barra__out
[01:37:32] *** kb1ooo has joined #openal
[01:37:32] *** ChanServ sets mode: +v kb1ooo
[01:38:12] *** barra_ has quit IRC
[01:44:16] *** barra__out has quit IRC
[01:51:51] *** Alam_DeathNote has left #openal
[02:29:09] * KittyCat is away: sleep
[02:31:08] <kb1ooo> KittyCat, what time is it for you?
[02:31:58] <KittyCat> 5:30pm, but I was up quite early today
[02:33:13] <kb1ooo> so ur on west coast us?
[02:33:22] <KittyCat> yeah
[02:34:06] <kb1ooo> ok figured that you were in europe or something :)
[02:34:16] <KittyCat> nah
[02:34:17] <kb1ooo> will let u go.
[04:02:02] *** kb1ooo has quit IRC
[04:19:14] *** jvalenzu has quit IRC
[04:41:26] *** jvalenzu has joined #openal
[04:41:27] *** ChanServ sets mode: +v jvalenzu
[06:09:04] *** jvalenzu has quit IRC
[07:27:02] *** jvalenzu has joined #openal
[07:27:02] *** ChanServ sets mode: +v jvalenzu
[07:49:08] *** jvalenzu has quit IRC
[08:59:39] *** Walt has quit IRC
[09:51:50] *** Walt has joined #openal
[09:51:50] *** ChanServ sets mode: +v Walt
[11:23:20] *** Stfu has joined #openal
[11:23:20] *** ChanServ sets mode: +v Stfu
[11:26:38] *** Stfu has left #openal
[15:41:39] *** juanmabc has quit IRC
[15:44:15] *** barraAway has joined #openal
[15:44:15] *** ChanServ sets mode: +v barraAway
[15:44:51] *** barraAway is now known as barra_
[16:08:25] *** jvalenzu has joined #openal
[16:08:25] *** ChanServ sets mode: +v jvalenzu
[16:56:31] * KittyCat is back.
[17:08:45] *** prophile has joined #openal
[17:08:45] *** ChanServ sets mode: +v prophile
[18:06:50] *** chieh has joined #openal
[18:06:50] *** ChanServ sets mode: +v chieh
[18:37:17] <angasule> hmm, I want an asus xonar :)
[18:38:44] <KittyCat> what's that?
[18:40:27] <angasule> it's a sound card with EAX 5.0 and linux drivers
[18:40:57] <KittyCat> true EAX 5.0, or just falsely reporting it?
[18:41:01] <KittyCat> and native OpenAL support?
[18:41:14] <angasule> not sure, I'm still reaidng
[18:41:16] <angasule> aghhh
[18:41:19] <angasule> konqueror crashed :()
[18:41:29] <angasule> and took the link with it, gimme a min
[18:41:43] <angasule> http://techgage.com/article/asus_xonar_d2_sound_card/
[18:41:53] <KittyCat> should be able to recover from crashes
[18:42:04] <KittyCat> Tools->Crashes
[18:42:35] <angasule> hmm, nothing there, but I'm using konqueror3 on kde4 so I guess some things only half-work
[18:42:46] <angasule> but konqueror4 is not ready for full time use, really
[18:43:17] <angasule> anyway, I have a live 5.1 card, so it doesn't take a lot to impress me :P
[18:43:34] <angasule> the x-fi hardware seems nice, too bad there are no open drives
[18:44:45] *** chieh has quit IRC
[18:45:31] <angasule> I've craved a real sound system for years, when I get a stable place I'll build an anechoic chamber with a 7.1 and then play 3D pong
[18:47:47] <KittyCat> hmm, the table only says EAX 1.0 and 2.0
[18:50:04] <angasule> the 5.0 I read yesterday on another site, and it said something about eax 5.0 having to be activated on each game
[18:50:22] <angasule> unfortunately it was about 3am so I don't remember well :)
[18:52:12] <KittyCat> yeah, I heard something about that too, which is why I initially asked
[18:52:40] <angasule> anyway, that is creative's doing, apparently (and according to what I already knew)
[18:52:45] <KittyCat> some games won't let you use EAX if it's not EAX 5. but the card only truelly does EAX 2.
[18:52:55] <KittyCat> or something like that
[18:53:19] <angasule> maybe the card is eax5 capable, but some patents by creative stop it from advertising as such or whatever
[18:53:31] <angasule> I don't really know what is in each eax version
[18:53:43] <angasule> but the sound quality on multiple channels alone would be nice
[18:54:05] <KittyCat> yeah. just hardware mixing and 3d would be nice
[18:54:29] <KittyCat> alsa and oss fail at that
[18:54:33] <angasule> btw, where is the EFX spec? I never found it
[18:55:07] <angasule> is alsa dying any time soon?
[18:55:11] <KittyCat> in the SDK. that's the only info I've found, anyway
[18:55:17] <KittyCat> I hope not
[18:55:29] <KittyCat> ALSA is nice. it just needs more hardware-oriented features
[18:55:35] <KittyCat> and better documentation
[18:55:36] <angasule> well, I meant dying as in getting replaced
[18:55:50] <angasule> ah, every time I had to deal with alsa it was painful
[18:56:00] <angasule> documentation and very odd config files
[18:56:02] <KittyCat> http://techgage.com/reviews/asus/xonar_sound_card/asus_xonar_19.jpg  that's neat
[18:56:39] <KittyCat> yeah. the API isn't too bad, though, and it'd be a shame to have to replace it just because documentation and hardware features are lacking
[18:56:51] <angasule> heh nice
[18:57:35] <angasule> is it too hard to document? or someone reading the source might do it?
[19:00:26] <angasule> I'm a bit puzzled by the card requiring an extra power connector
[19:00:43] <KittyCat> it can be difficult for someone not involved in the API design
[19:01:30] <KittyCat> because though you can see what the API does, it's ultimately up to the drivers as to what happens. and if there's no solid guidelines there, anything can happen
[19:03:36] <angasule> "Because the Xonar can record at almost exactly the same quality level with which it plays back sound, ALT allows high quality recordings to be made from the card's analog output. Asus advertises that this feature could be used to back up music files that are DRM-protected, allowing users to exercise their fair-use right to make a single archival copy of their digitally-stored, DRM-protected recordings."  <-- heh
[19:04:13] <KittyCat> heh
[19:05:11] <angasule> does alsa have a standard way of exposing the stuff needed for OpenAL?
[19:05:46] <KittyCat> nope. it doesn't even have the needed capabilities
[19:06:25] <angasule> eh, so how would the xonar go about doing a hardware openal?
[19:06:26] <KittyCat> the main thing there is that there's no per-output-speaker volume control.
[19:06:57] <KittyCat> like if you create a 1-channel PCM sound, you can only change the volume of that sound, not the individual output speakers its playing through
[19:07:31] <KittyCat> at most, you can emulate it by rechanneling it into a multi-channel buffer, but then you're wasting sources (one for each input channel)
[19:07:40] <angasule> aah
[19:08:37] <KittyCat> plus, you'd have to change and reload the buffer every time the volume changes
[19:08:51] <angasule> do cards usually detect the amount of speakers? or it's indicated by the user in some config file (on windows or whatver)?
[19:09:29] <KittyCat> I don't know, honestly, though I'd imagine the latter
[19:10:01] <KittyCat> the card gets the number of output speakers, and does the 3d->channel-colume conversions based on that
[19:10:11] <KittyCat> s/card/driver
[19:10:22] <KittyCat> *volume
[19:11:56] <KittyCat> the closest ALSA has to that functionality is the ttable. but that's likely software based since it's mainly a matrix. and it can't be changed in real-time (likely the runtime uses it when writing to the output buffer)
[19:12:36] <angasule> and the other features to implement openal in hardware are also missing?
[19:13:51] <KittyCat> the big one is the channel volume thing. past that, there could be issues with getting exact buffer sizes
[19:14:22] <KittyCat> though that can be fudged by doing some resampling as long as you can get a samplerate:buffersize match
[19:14:57] <KittyCat> and there's typically a give of 10hz or so between what's specified and what you get
[19:15:45] <angasule> I have no idea if that's much or not :)
[19:16:29] <KittyCat> I don't know either
[19:17:16] <KittyCat> I've not really looked into the buffer sizes you can get from ALSA vs the sample rates
[19:18:04] <KittyCat> "Of course, EAX is proprietary Creative Labs technology, so the very fact that we have to grade down the Xonar D2 because of its lacking EAX support goes to show just how tied-up the gaming industry is with Creative?s proprietary positional audio technology."
[19:18:12] <KittyCat> wonder if the same is true for EFX..
[19:19:15] <angasule> have you seen the EFX docs? I haven't yet
[19:19:30] <angasule> the EAX SDK has some NDA before downloading, so I don't have it
[19:19:32] <KittyCat> just what's in the OpenAL SDK
[19:19:36] <angasule> I have the EAX 2.0 SDK
[19:19:42] <angasule> which didn't require an NDA
[19:19:58] <KittyCat> the OpenAL SDK doesn't require an NDA, and has a PDF detailing the EFX API
[19:20:19] <angasule> ah, getting it now, then
[19:20:21] <KittyCat> which is all I've used in my engineering of OpenAL Soft's EFX support
[19:21:34] <angasule> yay downloading
[19:21:35] <KittyCat> it's not very detailed in specific information, though. eg. it doesn't say just how the low pass filter is supposed to work
[19:21:50] <KittyCat> all I could do though is play it by ear and compare it to what creative's drivers output
[19:21:58] <angasule> yes, OpenAL is nowhere near as well documented as OpenGL
[19:23:19] <KittyCat> the base OpenAL documentation is pretty good. al extensions have always been horribly documentated, though. even the loki-based ones
[19:23:54] <KittyCat> the info provided for EFX was refreshingly well documented by comparison
[19:24:12] <KittyCat> and a bit surprising given Creative's history, especially wrt EAX
[19:26:04] <angasule> hmm, the SDK is an .exe
[19:26:21] <KittyCat> it's an installer, yeah
[19:26:33] <angasule> wine depends on mozilla?
[19:26:53] <KittyCat> for internet explorer/html rendering
[19:27:29] <angasule> any chance of using another engine? like konqueror or webkit? :)
[19:28:06] <KittyCat> it uses win32 gecko
[19:28:28] <KittyCat> webkit may be able to work, but someone will have to do it, obviously
[19:28:32] <KittyCat> and it needs to be C
[19:28:59] <KittyCat> "DSP is not hardware-accelerated"  :(
[19:33:43] <angasule> win32 gecko? so if I download win32 firefox and install it it'll have html?
[19:33:51] <angasule> KittyCat: for xonar? :/
[19:34:28] <KittyCat> angasule, it should do it automatically
[19:34:29] <KittyCat> and yeah
[19:34:47] <KittyCat> if it doesn't, try running 'wine iexplore.exe'
[19:35:58] <angasule> very cool, installing now
[20:38:54] *** Alam_DeathNote has joined #openal
[20:38:55] *** ChanServ sets mode: +v Alam_DeathNote
[21:39:29] <angasule> KittyCat: I'm reading the EFX docs, and it has meters per unit
[21:40:19] <KittyCat> yeah, though it only applies to the effects
[21:40:26] <angasule> KittyCat: that, coupled with knowledge of the position of each speaker would allow placing sounds inside the speakers
[21:40:35] <angasule> oh
[21:40:48] <KittyCat> base openal only takes units
[21:42:37] <angasule> hmm
[21:43:39] <angasule> wouldn't it be possible to do that if meters per unit is specified?
[21:45:02] <KittyCat> well, I suppose it could be used in conjunction with other things to get the speaker spacing
[21:46:09] <angasule> imagine a conference room with speakers on the walls
[21:46:28] <KittyCat> but as I said, base openal only takes units. so if you set the min distance to 1, the max distance to 11, and the source is at a distance of 6, it'll be ttenuated the same regardless of what the meters per unit is
[21:46:41] <angasule> of course, most people wouldn't like it on account of promoting even more meetings :P
[21:47:07] <angasule> oh, yes, I meant it to place the speakers only
[21:47:27] <KittyCat> but then, what's the benefit?
[21:47:46] <angasule> you can have sounds inside the room, rather than coming from outside
[21:47:56] <KittyCat> hmm..
[21:48:07] <KittyCat> it would need an extension
[21:48:21] <KittyCat> a context attribute, most likely
[21:48:48] <KittyCat> but context attributes are integers, not floats
[21:48:52] <angasule> what for? :?
[21:49:39] <KittyCat> just the way it's defined. the attribute list is a key, val, key, val, etc array
[21:49:43] <angasule> the position of the speakers would be told to OpenAL through the driver (in configuration software), the app wouldn't have a say on it
[21:49:43] <KittyCat> of ALint
[21:49:50] <KittyCat> and *(ALint*)&somefloat; is ugly
[21:50:09] <KittyCat> ah, true
[21:50:23] <angasule> settings meters per unit would activate the 'inside sound' mode/path
[21:51:40] <KittyCat> that sounds ugly, though
[21:53:11] <angasule> maybe have a boolean to turn it on, then?
[21:55:50] <KittyCat> I could just use it regardless
[21:57:51] <KittyCat> I'd just need to do Distance *= SpeakerRadiusInMeters/MetersPerUnit;
[21:58:22] <KittyCat> that assumes the app sets the correct MetersPerUnit, though
[22:01:33] *** angasule has quit IRC
[22:05:03] *** angasule has joined #openal
[22:05:04] *** ChanServ sets mode: +v angasule
[22:05:48] <angasule> friggin' memory leaks
[22:05:50] <angasule> good thing I always work in screen
[22:18:01] <KittyCat> heh
[23:21:09] *** jvalenzu has quit IRC
[23:50:15] *** barra_ has quit IRC

top