| gnarface | slim chance of this but can anyone sanity check my ~/.asoundrc? the 8-channel default stopped working, i assume due to a regression from a kernel update, but maybe they actually fixed something broken and my config has been wrong all along: https://paste.debian.net/1253356/ | 03:02 |
|---|---|---|
| gnarface | (interestingly, if i play directly to pcm.output or pcm.duplex all 8 channels are still properly recognized, the issue is that "!default" will only use the front two and marks all the others as unused) | 03:03 |
| gnarface | suggestions for workarounds also welcome | 03:05 |
| gnarface | i already tried forcing the model as a module parameter again, no effect | 03:05 |
| gnarface | maybe there was a change to how "type plug" inherits values from the slaves or something? | 03:05 |
| joerg | gnarface: default is "type plug" right? Isn't this an auto-adapting thing? | 03:13 |
| gnarface | joerg: yes, i've set default to type plug in this config, and i do think you're correct that it also does format conversions but i am not clear on what else i could use in this context without sabotaging dmix/dsnoop | 03:15 |
| joerg | while the "duplex" combines the 8 out and 2 in, and plug trying to open it might only be able to auto-adapt to the common denominator which is two channels | 03:16 |
| gnarface | it used to work but various regressions with this driver have been common over the years, especially with regards to 5.1/7.1 channel sound | 03:16 |
| gnarface | hmm, interesting hypothesis.... | 03:16 |
| joerg | it's a while since I last time looked into alsa plugin definitions | 03:18 |
| gnarface | it's kinda vague to me but i was under the impression that plug has to be somewhere in the chain ahead of dmix for dmix to work | 03:19 |
| gnarface | but i don't have all the types memorized so maybe something else would be more appropriate | 03:19 |
| joerg | there's one that allows splitting / routing /mixing a two real PCM into/from 8 virtual channels I think. Or you kick the pcm.duplex completely out | 03:20 |
| joerg | ooh, dmix/dsnoop are a tad "special" too iirc. plug might not be mandatory for dmix to work, but it helps a lot | 03:23 |
| joerg | I seem to recall a `aplay -l` or `aplay -L` (also arecord) | 03:27 |
| gnarface | hmm, well i liked your theory about the channel difference in input and output but commenting out 'capture.pcm "input"' from pcm.duplex has no effect on the output | 03:27 |
| joerg | helped me a lot | 03:28 |
| gnarface | to confirm commenting is working at all i also tried a test where i commented out the playback.pcm and speaker-test just errors then | 03:29 |
| joerg | probably that's the reason why it looks like this, the soundcard has only 2 ADCs aka record PCMs? | 03:29 |
| gnarface | (so "type plug" is still suspect but maybe not the channel difference in input and output) | 03:30 |
| gnarface | uh, how would i tell? | 03:30 |
| gnarface | oh, yes, "aplay -l" lists 2 analog inputs | 03:31 |
| gnarface | i mean "arecord -l" | 03:31 |
| joerg | :-) | 03:31 |
| joerg | that's beyond my paygrade | 03:31 |
| gnarface | not clear on the significance of what you're saying, but you're right, 2 analog inputs, one analog output, and one digital output | 03:31 |
| gnarface | hmm, i wonder if i'm supposed to be using device 2 for the input instead of using device 0 for both input and output? | 03:33 |
| gnarface | would it be likely for both inputs to be wired to the same analog jack? | 03:33 |
| gnarface | device 0 is listed as both an input and an output, device 1 is the digital output and device 2 is "Alt Analog" input | 03:35 |
| gnarface | i had assumed i think by testing that the "alt analog" input is something on the motherboard that just isn't wired to anything | 03:36 |
| gnarface | but if these inputs and outputs could share jacks, maybe they're both listening there | 03:36 |
| gnarface | ? | 03:37 |
| gnarface | dunno | 03:37 |
| gnarface | it's probably relevant that it doesn't work right without the config either | 03:38 |
| joerg | aplay/arecord --verbose : Show PCM structure and setup | 03:39 |
| gnarface | default is clearly 2 channels as far as the driver is concerned, unless a config says otherwise | 03:39 |
| joerg | will give a hint what the plug pcm does | 03:40 |
| gnarface | it doesn't seem to have any effect on the output | 03:40 |
| gnarface | oh, or do you mean while actually playing or recording something? | 03:41 |
| joerg | yes | 03:41 |
| gnarface | --verbose doesn't change the -l/-L output | 03:41 |
| joerg | no, it's complementary to -l/-L | 03:41 |
| joerg | you should try aplay --channels=8 | 03:43 |
| gnarface | hmmm | 03:44 |
| gnarface | maybe the bug is just in speaker-test somehow | 03:44 |
| gnarface | channels : 8 | 03:45 |
| gnarface | this shows up in the last block with or without "--channels=8" on the command-line | 03:45 |
| gnarface | even though the wav i'm sending in is only 1 channel | 03:45 |
| gnarface | not sure exactly how to read this but it seems like it's doing what i'd expect correctly https://paste.debian.net/1253363/ | 03:48 |
| gnarface | not sure how to make a 8-channel wav easily | 03:49 |
| joerg | line 29: Transformation table: | 03:50 |
| gnarface | i have ttable in the ~/.asoundrc just to lower the center channel volume because the driver combines center/LFE wrong | 03:51 |
| gnarface | i'm not clear on if it's saying it's obedient there or not | 03:52 |
| gnarface | should be no changes except 50% volume on center and lfe | 03:52 |
| gnarface | otherwise the center speaker is double the volume of all the surrounds | 03:53 |
| gnarface | or at least was, last time this config was working | 03:53 |
| gnarface | haven't gotten as far as testing that part yet | 03:53 |
| gnarface | because as of right now it doesn't play anything through that speaker | 03:53 |
| gnarface | you think this could be breaking something? | 03:54 |
| joerg | no, should output *something* | 03:54 |
| gnarface | well it's not the problem because there's no change to this state without any config | 03:55 |
| joerg | channel 0 (left?) from .wav to out chan 0, 2, 4*50%; 6, | 03:56 |
| gnarface | oh, shit you're onto something here | 03:57 |
| gnarface | commenting out all the ttable lines from the config, "speaker-test -c 8" now works as expected | 03:57 |
| gnarface | thanks for that, that's progress | 03:58 |
| gnarface | i still don't know how to apply my ttable correction without sabotaging the channel count though | 03:58 |
| gnarface | that used to work right | 03:59 |
| gnarface | i'll have to do a test to see if i even need it i suppose, but i'm assuming from past behavior that i will | 03:59 |
| joerg | try line 43: slave.pcm "output" | 04:13 |
| joerg | instead of "duplex" | 04:13 |
| joerg | should spread input channels 0,1 to out 0,1,2,3,4,5,6,7 aiui | 04:15 |
| gnarface | you mean with ttable lines uncommented? | 04:17 |
| gnarface | trying... | 04:17 |
| gnarface | nah, doesn't help | 04:18 |
| gnarface | hmm, maybe it's academic, because the center channel doesn't seem louder to my ears anymore | 04:24 |
| gnarface | so maybe they fixed that part | 04:24 |
| gnarface | anyway, thanks for the help joerg | 04:24 |
| joerg | type route, check how they do ttable for example in http://www.sabi.co.uk/Cfg/ALSA/asound.conf (^f_ind: ttable) | 04:25 |
| gnarface | hmm, rear left seems real quiet, something may still be wrong with it | 04:27 |
| joerg | amixer (or alsamixer) might show some hw volume controls for each DAC | 04:28 |
| * joerg wonders if this is really on topic, but assuming every other distro just uses polypaudio... maybe we're on topic with alsa ;-) | 04:29 | |
| gnarface | yea sorry for the noise, but nobody was answering over there | 04:30 |
| onefang | Sounds like what you need is more noise. B-) | 04:31 |
| rrq | gnarface: sox can make an 8 channel wav from a 1 channle wav.. "sox $in $out channels 8" | 04:35 |
| joerg | gnarface: https://alsa.opensrc.org/DmixPlugin seems useful source of info | 04:37 |
| rrq | gnarface: and even use `play "|sox $in -t sox - channels 8"` to send it on-the-fly to the cards | 04:38 |
| gnarface | ah, thanks rrq | 04:39 |
| rrq | (doesn't solve the ttable thingy though) | 04:39 |
| Xenguy | Gawd save the queen | 04:40 |
| Xenguy | (We mean it, man) | 04:40 |
| joerg | nevermind the bollocks | 04:45 |
| Xenguy | Don't worry about that stuff behind the curtains | 04:46 |
| Xenguy | Or | 04:46 |
| Xenguy | Keepin' it real | 04:46 |
| joerg | I *think* the format is ttable.<inchan>.<outchan> <gain> | 04:52 |
| joerg | for playback. No idea if and how this translates for concurrent recording | 04:54 |
| joerg | the point being: I don't see how you could play 8 channels of source via a ttable.[01].[0-7] that only defines an "upstream" stereo channels 0 and 1 | 04:58 |
| rrq | mmm I understood the "type asym" of "duplex" lets it be 8 channels on playback and 2 channels on capture | 05:05 |
| rrq | ah, now I was confused again.. yes the upstream for "default" is 2 channels.. | 05:12 |
| rrq | not sure how tell "play" to use "output" directly.. aybe "AUDIODEV=output play ..." | 05:18 |
| gnarface | oh hmm, yes i may have been using those ttable outputs to map stereo to 8 channels as well, for some reason that was working right before, like it was smart enough to ignore it when there were more than 2 channels? now that makes it seem curious that it would have ever worked, so maybe that's a sign i did this wrong after all | 05:46 |
| gnarface | maybe they just changed how it worked, what i vaguely remember was that it'd ignore the upstream channel setting if there was already a real channel present? | 05:47 |
| gnarface | hmmm | 05:47 |
| gnarface | my first test of 5.1 sound through vlc failed with utterly garbled sound (like feeding a binary directly through the audio interface raw type of sound) but after removing the config and verifying vlc works without the config, vlc still worked right after replacing the config, so that's weird | 05:50 |
| gnarface | but not bad as long as it stays working i guess | 05:51 |
| gnarface | yea they must have changed how ttable treats used channels or something | 05:56 |
| Xenguy | Anyone else receive a sizable apt-get update this morning? (I'm still on Beowulf here) | 16:04 |
| Xenguy | The following packages will be upgraded: | 16:05 |
| Xenguy | eboard gir1.2-rsvg-2.0 gnome-orca grub-common grub-pc grub-pc-bin grub2-common htmldoc htmldoc-common krb5-locales libfreetype6 libfribidi0 libgssapi-krb5-2 libk5crypto3 libkrb5-3 libkrb5support0 libnet-ssleay-perl libqt5concurrent5 libqt5core5a libqt5dbus5 libqt5gui5 libqt5network5 libqt5printsupport5 libqt5sql5 libqt5sql5-sqlite libqt5test5 libqt5widgets5 libqt5xml5 librsvg2-2 librsvg2-common libwireshark-data libwireshark11 libwiretap8 libwscodecs2 | 16:05 |
| Xenguy | libwsutil9 mutt orca procmail publicsuffix python3-twisted python3-twisted-bin qt5-gtk-platformtheme tshark unrar wireshark-common | 16:05 |
| fsmithred | Xenguy, I see 74 packages on beowulf to be upgraded. My last upgrade was mid-July. | 16:15 |
| fsmithred | compared to 30 on daedalus where I upgraded three days ago. | 16:16 |
| Xenguy | fsmithred, I think I'm just the suspicious type, I always feel the need to ask, when I see large updates roll in ; -) | 16:20 |
| * Xenguy bites the bullet... | 16:21 | |
| nemo | don't have anything on beowulf anymore myself | 16:37 |
| nemo | was pretty eager to get off it actually, had a bunch of missing goodness in terms of drivers that I needed | 16:37 |
| brocashelm | ironically, being on ceres, i'm thinking of staying on beowulf for the sake of a less shitty xfce/gtk experience | 20:46 |
| brocashelm | backports and appimages can help a bit | 20:47 |
| fsmithred | can you forward-port xfce-4.12? | 20:47 |
| brocashelm | i have been able to downgrade to xfce 4.12 by carefully checking dependencies, but i think it would be better to use more compatible libs | 20:49 |
| brocashelm | it's not so much xfce as it is gtk | 20:50 |
| brocashelm | i purposely use gtk2 versions of programs whenever available. i keep a gtk2 version of leafpad | 20:51 |
| fsmithred | brocashelm, have you tried any gtk4 yet? | 21:01 |
| fsmithred | I some in ceres | 21:01 |
| fsmithred | losing whole words now | 21:03 |
| fsmithred | s//see/ | 21:03 |
| brocashelm | yeah, i have some gtk4 packages installed because of network-manager | 21:27 |
| brocashelm | libgtk-4-1 and libgtk-4-common to be exact | 21:27 |
| golinux | Ugh | 21:28 |
| golinux | It's worse than monkey pox | 21:29 |
| golinux | A ubiquitous infection . . . | 21:29 |
| brocashelm | i think lxde is the only fully gtk2 DE left | 21:36 |
| golinux | Maybe it should be the default going forward . . . | 21:59 |
| brocashelm | afaik, it's a mature product and mostly maintained by arch devs. some programs like lxhotkeys don't work anymore, sadly | 22:01 |
| brocashelm | lxhotkeys has been giving me an error message saying "openbox not supported right now, sorry" | 22:01 |
| fsmithred | is some other wm supported? | 22:01 |
| fsmithred | what is lxqt using now? | 22:02 |
| brocashelm | i don't think they have their own wms, but openbox seems to be suggested by default for both lxde and lxqt | 22:03 |
| fsmithred | I seem to recall that lxqt gave a choice of either openbox or another. possibly xfwm | 22:04 |
| brocashelm | the lxqt package itself suggests obconf-qt | 22:05 |
| brocashelm | but maybe it's just due to the toolkit | 22:05 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!