Date: Thu, 24 Jan 2008 10:58:05 +0300 From: "Yuriy Tsibizov" <Yuriy.Tsibizov@gfk.com> To: "Mark Kane" <mark@mkproductions.org> Cc: freebsd-multimedia@freebsd.org Subject: RE: Questions about snd_emu10kx Message-ID: <78664C02FF341B4FAC63E561846E3BCC0EEC02@ex.hhp.local> In-Reply-To: <20080122154114.28be626c@mkproductions.org> References: <20080122154114.28be626c@mkproductions.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> I've been using snd_emu10k1 for a long time with my Audigy 2 Platinum > and it has been working great. Recently I moved to FreeBSD 7 and was > experiencing problems recording (I believe the same or=20 > similar issues to > kern/119759), so I had to switch to snd_emu10kx to urgently get > something recorded. With emu10kx I have noticed a few things that I > have questions about: >=20 > 1) The headphone jack on the front panel of the Audigy 2 Platinum is > not outputting sound like it did with emu10k1. I have tried with all > mixer levels up with no luck. Is this a misconfiguration, known > bug/limitation, or an unknown issue that I should file a PR for? Mark,=20 for headphones output -- as I remember it was disabled for some reasons (like this=20 DSP output pin was used for different output on some cards). I can make a patch for you to test, what happens if I enable it back. >=20 > 2) I no longer seem to be able to mute the line input from coming out > of the card while still recording from it. In emu10k1 I could set > "line" in the mixer to 0, but "rec" higher with the recording source > set to line and it would still record while not being outputted from > the card. With emu10kx, if I set "line" to 0 even with the=20 > source set to > line and "rec" at 100 no audio is recorded from the line. I have also > tried setting the recording source to "line1" and setting "line" to 0 > but then it does not record either. For some time there were separate contols for amount of signal coming into recording and into playback (see http://chibis.persons.gfk.ru/audigy/default_en.htm#routing), but with move to mixer(8) InRecVolumeN =3D InVolumeN. It is possible to set = InRecVolumeN to 100% and get behaviour similar to snd_emu10k1 one. Also, there are different line inputs on Audigy. Some of them are connected to AC97 chip and sent to DSP 'AC97 in' input. There is only one input on DSP, and AC97 chip is always in 'recording' mode (even if you want to listen to sound from you line input on speakers). I think I should add some logic to not pass output from of AC97 codec to playback (speakers) if all volume controls for it are set to 0, but this will not solve all the problem. If you are going to listen to audio CD (that's connected to AC97) in the same time you want to record from 'line in' (connected to the same chip) you will get both sources mixed togeter on both speakers and recording. As I know, snd_emu10k1 use different way for playback -- it uses AC97 codec for playback and there are no problems with unexpected sound mix. >=20 > 3) Last night I noticed that I was not able to play any audio with a > 48kHz sample rate. I'm pretty sure it was working after switching from > emu10k1 to emu10kx but last night it suddenly stopped working. In XMMS > or Audacity a buzz/stuttering sound is outputted when playing a > 48kHz file. In mplayer, playing a video with 48kHz audio causes it not > to play correctly (it will play a frame of video then stop=20 > playing). If > playing video in VLC with 48kHz audio it will play the video only with > no sound. If I try using jackd and ZynAddSubFX with a 48kHz=20 > sample rate > it also does not output any sound. There are no errors > in /var/log/messages or from the applications themselves so I'm not > sure what is causing this problem. I did update some ports but I'm not > sure if that's related to this, or what would cause all=20 > applications to > fail with anything 48kHz. I'm working on other amd64 problem (broken digital sound output on Live! card on amd64). Can you try to build your kernel without optimizations -- make buildkernel __MAKE_CONF=3D/dev/null COPTFLAGS=3D-O0 -- and try again? also, you can try to patch your /usr/share/mk with http://chibis.persons.gfk.ru/7/sharemk.patch=20 and rebuild kernel & modules if kernel built for -O0 still does not work for you. This patch will allow use of COPTFLAGS for module build (they will be built with -O2 in first case). Also, do you load driver in loader.conf or with kldload? Yuriy Tsibizov, http://chibis.persons.gfk.ru/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?78664C02FF341B4FAC63E561846E3BCC0EEC02>