[alsa-devel] VIA VT1708 bugs
Hi,
the state of VIA VT1708 in the current kernel is far from optimal. It would be good to improve the situation, at least for 2.6.38. On the machine here, I experience at least these bugs:
1) "Jack detect" switch has no function:
* When turning "Jack detect" off, quitting alsamixer, then restarting it, "Jack detect" is still on. * When turning "Jack detect" off, then turning it on again, alsamixer crashes with the message "alsamixer: hcontrol.c:764: snd_hctl_handle_event: Assertion `elem' failed."
2) Plugging headphones in toggles "Front" switch control
* This should probably just be removed, especially with the latest opinion from Takashi of what "Front" really means. The problem is worsened by that PulseAudio seems to pick it up in some circumstances and thinks that it should mute everything.
3) Several users report headphones output being broken
* Symptoms vary a little and some might be due to bug 2), but some however report the headphones volume being "grayed out" in alsamixer. I haven't seen that in particular on the hardware here.
4) All these extra controls - "Independent HP", "Jack Detect", "Smart 5.1"
I assume they are meant to add extra functionality, and of course I don't mind that, but the problem is that since these are not very well described (or am I missing documentation somewhere?) it's difficult to know the exact intended functionality of these, and without knowing that, it's even more difficult to fix it when it's broken (or even knowing if it is broken or not!).
Any ideas and/or opinions? Should we talk to VIA about it, after all they seem to have written a lot of the code here? Or can someone here cast some light on these issues?
2011/1/4 David Henningsson david.henningsson@canonical.com
Hi,
the state of VIA VT1708 in the current kernel is far from optimal. It would be good to improve the situation, at least for 2.6.38. On the machine here, I experience at least these bugs: 3) Several users report headphones output being broken
- Symptoms vary a little and some might be due to bug 2), but some
however report the headphones volume being "grayed out" in alsamixer. I haven't seen that in particular on the hardware here.
- All these extra controls - "Independent HP", "Jack Detect", "Smart 5.1"
Independent headphone
http://www.viaarena.com/forums/showthread.php?t=41015
"Smart 5.1" similar to switch "channel mode" from 2 to 6 in intedl desktop motherboard with 3 audio jacks in rear panel
"Independent Headphome" similar to "Multistreaming" playback in those two videos
http://www.intel.com/support/motherboards/desktop/sb/cs-020642.htm
On 2011-01-04 08:43, Raymond Yau wrote:
2011/1/4 David Henningssondavid.henningsson@canonical.com
Hi,
the state of VIA VT1708 in the current kernel is far from optimal. It would be good to improve the situation, at least for 2.6.38. On the machine here, I experience at least these bugs: 3) Several users report headphones output being broken
- Symptoms vary a little and some might be due to bug 2), but some
however report the headphones volume being "grayed out" in alsamixer. I haven't seen that in particular on the hardware here.
- All these extra controls - "Independent HP", "Jack Detect", "Smart 5.1"
Independent headphone
Ok, that makes some sense. It translates to me as * With independent HP turned off - activates auto-mute between speaker-out/line-out and headphone out. Note the difference between VIA and Realtek/Sigmatel here as line-out are auto-muted as well.
* With independent HP turned on - headphone out and speaker-out/line-out should take output from different PCM subdevices.
"Smart 5.1" similar to switch "channel mode" from 2 to 6 in intedl desktop motherboard with 3 audio jacks in rear panel
But does turning "Smart 5.1" on, mean that mic-in and line-in are always turned into C/LFE and Back L/R jacks, or is it trying to do some kind of autodetection of the current thing plugged in?
Also, it is not obvious to me how "Jack detect" comes into this equation and how they affect the two parameters above...?
2011/1/4 David Henningsson david.henningsson@canonical.com
On 2011-01-04 08:43, Raymond Yau wrote:
2011/1/4 David Henningssondavid.henningsson@canonical.com
Hi,
the state of VIA VT1708 in the current kernel is far from optimal. It would be good to improve the situation, at least for 2.6.38. On the machine here, I experience at least these bugs: 3) Several users report headphones output being broken
- Symptoms vary a little and some might be due to bug 2), but some
however report the headphones volume being "grayed out" in alsamixer. I haven't seen that in particular on the hardware here.
- All these extra controls - "Independent HP", "Jack Detect", "Smart
5.1"
Independent headphone
Ok, that makes some sense. It translates to me as
- With independent HP turned off - activates auto-mute between
speaker-out/line-out and headphone out. Note the difference between VIA and Realtek/Sigmatel here as line-out are auto-muted as well.
Those video about realtek and IDT codec are all Intel DESKTOP motherboard
line out does not auto mute because of multi streaming playback
For example, you can listen to one audio source through the back panel speakers and a second audio source through front panel headphones or speakers.
- With independent HP turned on - headphone out and speaker-out/line-out
should take output from different PCM subdevices.
"Smart 5.1" similar to switch "channel mode" from 2 to 6 in intedl desktop motherboard with 3 audio jacks in rear panel
But does turning "Smart 5.1" on, mean that mic-in and line-in are always turned into C/LFE and Back L/R jacks, or is it trying to do some kind of autodetection of the current thing plugged in?
"Smart 5.1" is not specific to VIA HDA codec, it first appear in those onboard ac97 motherboard
http://www.viaarena.com/via-mainboards-smart-51.aspx?ID=76&MCatID=1
smart 5.1
http://git.alsa-project.org/?p=alsa-kernel.git;a=commit;h=1564b2878f5cf160f6...
http://www.via.com.tw/en/products/audio/codecs/vt1708b/index.jsp
To support 7.1 using 8 channels dac require retask headphone as side (I guess the motherboard only has 5 audio jacks in rear panel )
Jack detect
http://git.alsa-project.org/?p=alsa-kernel.git;a=commit;h=1f2e99febd5dd0c91f...
I think the default value of "smart5.1" switch should be :"no" instead of "Yes"
This mean that the driver should not create "smart5.1" for those 10 channels dac hda codec
May be the best way is "Gnome Sound Preference" should check whether the user can select surround 51 profile if rear mic/line is selected as input source whenever the sound card has "Smart5.1" switch or "channel mode" switch and vice versa
All those micro ATX size motherboard has only 3 audio jacks in the rear panel
on-board sound drivers use ac97 codec (e.g ATIIXP, ICH, ICH4 , VIA8233, VIA8233A , VIA8237 , NFORCE ) have the "channel mode" switch in surround40 or surround51
http://git.alsa-project.org/?p=alsa-lib.git;a=blob;f=src/conf/cards/http://git.alsa-project.org/?p=alsa-lib.git;a=blob;f=src/conf/cards/ATIIXP.con
2011/1/4 David Henningsson david.henningsson@canonical.com
Ok, that makes some sense. It translates to me as
- With independent HP turned off - activates auto-mute between
speaker-out/line-out and headphone out. Note the difference between VIA and Realtek/Sigmatel here as line-out are auto-muted as well.
The point is using ac97 front panel connector does not allow you to know whether the headphone plugged in , the presence of headphone is only know when you are using HDA front panel and HDA codec. so line-out of a desktop is usually not auto muted
http://www.intel.com/support/motherboards/desktop/sb/cs-020642.htm
participants (2)
-
David Henningsson -
Raymond Yau