Intel HD audio problem

[dsjonny]

I have integrated Intel audio on my board (Intel DH61AG):

• device/flags: 1
• device/id: 0x1c20
• device/interface: 0x0
• device/subtype: 0x3
• device/type: 0x4
• device/vendor: 0x8086
• device name: 6 Series/C200 Series Chipset Family High Definition Audio

The problem is: there is very bad output. There is sound, but something is hears bad. I have tried to change all thing on the media settings panel, but not helped.

And there is another problem: no HDMI audio output.

I have attached an mp3 file save from the output.

[dsjonny]

One thing: I have found about 2-3 ticket about there is no audio output on Intel HD audio, but I think this problem is not the similar, because I have audio output, but it sound very bad.

[dsjonny]

I have attached a syslog file. In the I found that it detects the HDMI as output as well.

And I remember that there was a same problem (the bad sound) on BeOS too, and I could solve it when I changed the multi audio media addon to another. Unfortunatelly I do not remember what files I have changed.

[dsjonny]

I remember now: on BeOS the problem was the same, and I replaced the original BeOS multi_audio.media_addon to a haiku cvs version multi_audio.media_addon. That solved my problem. That was a VIA chipset, but the problem was the same. Unfortunatelly that addon is not working with Haiku now.

[dsjonny]

It is very strange... If I play a video in MediaPlayer, than if the MediaPlayer is on the top, than I hear just a bit noise, but if it is in full screen mode, than the sound is perfect. If the player is not on the top or in full screen, than when I switch to another window, than the sound will be terrible again.

[axeld]

I can also confirm the same oddities here - playing back video considerably improves the sound quality. Changing the workspace causes audio to be bad again. Full screen video makes the sound almost perfect (it only crackles rarely). Amazing. I'm attaching my syslog here as well. Interestingly, I find the same two suspicious lines, too:

hda: DMA position for stream (id:2) seems to be broken. Switching to using LPIB.
hda: DMA position for stream (id:1) seems to be broken. Switching to using LPIB.

[axeld]

Axels syslog hda excerpt

[jrepan]

Constantly moving around a window(for example Terminal) also improves quality. Could it be something related to inter-process communication?

[axeld]

(In #9095) This sounds a lot like #8949 -- closing as a duplicate.

[axeld]

Fixed in hrev44757.

At BeGeistert, Marcus made me aware that snooping is only enabled for a specific Intel device -- when I played with that earlier, I did not notice this at all.

[dsjonny]

Replying to axeld:

Fixed in hrev44757.

At BeGeistert, Marcus made me aware that snooping is only enabled for a specific Intel device -- when I played with that earlier, I did not notice this at all.

Good news! :) Can you add this to the nightly build too? Than I can check it.

[luroh]

Replying to axeld:

Fixed in hrev44757.

Yes, this got my audio [8086:1c20] working as well (you probably noticed the horrible distorted noise emanating from my laptop at BG). Thanks!

[dsjonny]

Perfect! Thank you!

[Pete]

Replying to axeld:

Fixed in hrev44757.

I'll add my many thanks to Axel here. I've been bashing away at this same problem for months on my ZaReason laptop (where everything Haiku worked beautifully except the HDA sound). Somehow this ticket slipped by me til now, but the symptoms are identical. With this fix, audio is suddenly beautiful too! For reference, the hardware is reported as "7 Series/C210" with codec ALC269VC (from Linux).