[mythtv-users] Weird audio bleed-over with analog recordings

[Rod Smith]

On Monday 26 March 2007 17:09, Bob Sully wrote:
> Rod Smith wrote:
> >
> > Have you tried muting the "Duplicate Front" mixer channel? I ran into
> > this and
> > added a brief mention of it to the wiki:
> >
> > http://www.mythtv.org/wiki/index.php/Sound_Troubleshooting#Echo_on_audio_
> >input
>
> Unfortunately, there's no such channel in Alsamixer (and for some reason,
> KMix doesn't show the input channels on Audigy cards).  Any other mixers
> that you know of which might?

The mixer programs just make available the mixer channels provided by your 
hardware. If there's no Duplicate Front channel in your mixers, that's 
because there's nothing with that label for your hardware. There may be 
something with the same function but a different name, though, so you might 
want to try muting everything except what you're positive should be left 
unmuted until you find the right one (if you do -- I make no promises you 
will).

Also, the alsamixer program separates audio channels into two groups: Playback 
and Capture. On my systems, it comes up showing the Playback group only. Be 
sure to check both groups by pressing the F3 and F4 keys to toggle between 
them.

As a side comment, this is part of what's frustrating about configuring sound 
on MythTV: There's little standardization on the names for mixer functions 
across different sound hardware. This makes it difficult to have sensible 
conversations about what mixer settings work and what ones don't. (This isn't 
a MythTV issue per se; it's an issue with the Linux drivers.)

> In Alsamixer, there is a PCM capture, which I
> can lower but not mute - that one seems to be the culprit, but when I
> bring up mythfrontend again, the PCM Capture level is back up to where it
> was previously (yes, I do save the setting on exit)...I don't know why it
> won't stay down.

You said that the problem occurs when watching live TV, but I don't recall if 
you said it occurred when recording. If not, then my guess is that you should 
*NOT* be muting the PCM Capture channel but you SHOULD be muting something 
else. I'd expect the PCM Capture channel to be the one used to record audio 
inputs, and as MythTV always records, you'd want to record this. With the PCM 
Capture channel set low, try pausing your live TV. If the audio continues to 
play while the recording is paused, thus getting out of sync when you start 
playing again, then you've got the wrong channel. If the audio pauses along 
with the video and stays in sync when you start up again, then you've got the 
right channel.

If I'm right (that you've identified the wrong channel), then I suggest you 
start going through the audio channels again, leaving PCM Capture set to 
whatever value is good and muting or lowering other channels until you find 
the one that's causing the echo.

If I'm wrong and you've located the correct channel, then you might look at 
the MythTV recording options. Some of these relate to audio volume, and it's 
likely that one of these is adjusting the PCM Capture channel, probably 
thinking that it's the channel from which MythTV is recording audio.

One other possibility occurs to me: Some sound cards can be configured to 
provide audio input devices for Linux. I don't believe my pcHDTV 3000 is one 
of them, but I don't know about the pcHDTV 5500. If you've got a Linux device 
file for your pcHDTV 5500's audio channel, then it's conceivable that MythTV 
is recording audio from that and that, independent of MythTV, your mixer is 
configured to pass through audio from the sound card's inputs to the 
speakers. If this is the case, then unplugging the cable from the pcHDTV 5500 
to your sound card should fix the echo while leaving recordings working. 
There is one problem with this hypothesis, though: If it were true, I'd 
expect you to hear audio from that card at all times, or at least whenever 
you're recording from the card. Perhaps there's some reason this isn't 
happening, though, and given that you just need to unplug that one cable to 
test the hypothesis, it's worth checking.

-- 
Rod Smith
http://www.rodsbooks.com