Login | Register For Free | Help
Search for: (Advanced)

Mailing List Archive: MythTV: Dev

Re: [mythtv-commits] Ticket #1104: multi channel audiosupport

 

 

MythTV dev RSS feed   Index | Next | Previous | View Threaded


mark at digivation

Feb 6, 2008, 10:01 PM

Post #1 of 10 (1492 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport

> --enable-liba52bin --disable-decoder=ac3 --enable-decoder=liba52
>
> Right?

I think so. havent tried it myself for ages.

>
> If I do that the clicks and pops go away --- now I get silence with
> any of my ATSC (HDTV) shows. It does not matter if I have timestretch
> enabled or not.
>
> It does seem to work with the shows I have recorded using my PVR-500,
> though. Even the PVR-500 shows had clicks and pops without the liba52
> enabled.

now we might be getting somewhere.
I suspect the ac3_sync in internal ac3dec may not be behaving itself
sometimes.
can you do a test with audio,timestamp and libav enabled (only short) and
make the log available please.
just use internal as we dont really care about liba52 I suspect (I may be
wrong).

>
> > also what was the last patch worked?
>
> Actually, this is the first time I have tried any of the patches on
> this ticket. Can you give me a list of candidate or should I just
> start bisecting the version?

dont worry about this as they wont apply well anyway as the ones I wanted to
test are quite old (say pre december when I added freesurround).

mark

_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


jppoet at gmail

Feb 6, 2008, 10:26 PM

Post #2 of 10 (1397 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

On Feb 6, 2008 11:01 PM, Mark Spieth <mark [at] digivation> wrote:
> > --enable-liba52bin --disable-decoder=ac3 --enable-decoder=liba52
> >
> > Right?
>
> I think so. havent tried it myself for ages.
>
> >
> > If I do that the clicks and pops go away --- now I get silence with
> > any of my ATSC (HDTV) shows. It does not matter if I have timestretch
> > enabled or not.
> >
> > It does seem to work with the shows I have recorded using my PVR-500,
> > though. Even the PVR-500 shows had clicks and pops without the liba52
> > enabled.
>
> now we might be getting somewhere.
> I suspect the ac3_sync in internal ac3dec may not be behaving itself
> sometimes.
> can you do a test with audio,timestamp and libav enabled (only short) and
> make the log available please.
> just use internal as we dont really care about liba52 I suspect (I may be
> wrong).

Okay, this is weird. I tried grabbing the log for you with the liba52
version, since that is what I had installed. If I run mythfrontend
like:

mythfrontend -v audio,timestamp,libav -l logfile

I *do* get audio with a ATSC show. However, I still do *not* get
audio if I run it like this:

mythfrontend -v playback -l logfile

Why would adding the logging flags change the behavior? I tried it
each way several times, and it was consistant.

Also, as soon as I hit "TOGGLESTRETCH" mythfrontend segfaults, if
"audio,timestamp,libav" is turned on. It does not segfault with
"playback" is the only verbose option.

Do you still want the log file?


John
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


mark at digivation

Feb 13, 2008, 6:14 PM

Post #3 of 10 (1382 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

> 2) Change I was asking about was trivial swab-and-tidy.patch.
> Add unsigned char * payload = data + 8;
> Hoping to make a little more readable.

I have no problem with that and it makes sense. I just copied code form avf
where it was originally (and still is I think)

mark

_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


mark at digivation

Feb 16, 2008, 11:49 PM

Post #4 of 10 (1339 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

>I just tried svn rev 16112. With it I no longer get click & pops, and
>the sound generally seems to work.
>
>However, if I invoke timestretch and then go back to 1.0x speed,
>pass-through is not re-enabled.

as intended. is there a problem with this?

mark
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


jppoet at gmail

Feb 17, 2008, 8:09 AM

Post #5 of 10 (1331 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

On Feb 17, 2008 12:49 AM, Mark Spieth <mark [at] digivation> wrote:
> >I just tried svn rev 16112. With it I no longer get click & pops, and
> >the sound generally seems to work.
> >
> >However, if I invoke timestretch and then go back to 1.0x speed,
> >pass-through is not re-enabled.
>
> as intended. is there a problem with this?

If I start watching a HD show at 1.0x speed, pass-through is enabled,
and I get Dolby Digital 5.1 sound.

If I enable timestretch, I get Dobly Prologic II -- just like I have
ever since you added timestretch.

If I then turn timestretch off (go back to 1.0x speed), I still have
Dolby Prologic II.

Why would I not want native Dolby Digital 5.1 when available?

Thanks,

John
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


jppoet at gmail

Feb 17, 2008, 9:05 AM

Post #6 of 10 (1331 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

On Feb 17, 2008 9:09 AM, John P Poet <jppoet [at] gmail> wrote:
> On Feb 17, 2008 12:49 AM, Mark Spieth <mark [at] digivation> wrote:
> > >I just tried svn rev 16112. With it I no longer get click & pops, and
> > >the sound generally seems to work.
> > >
> > >However, if I invoke timestretch and then go back to 1.0x speed,
> > >pass-through is not re-enabled.
> >
> > as intended. is there a problem with this?
>
<snip>
>
> Why would I not want native Dolby Digital 5.1 when available?


Please educate me on what problem this "multi channel audio support"
is trying to solve.

I have a very good surround sound preamp driving a 7.2 channel speaker
setup. My preamp can decode any surround sound format other than the
new HD audio on Bluray or HD-DVD. It does a very good job at routing
the appropriate sounds to the appropriate speakers. When fed a Dolby
Digital 2.0 signal, it's ProLogic IIx decoder does an excellent job of
filling all 7.2 speakers with the appropriate audio cues. When fed a
two channel stereo signal, it has a very good "music" mode which does
an excellent job at divvying up the sound amongst all seven speakers.

How does this "multi channel audio support" improve my situation?

Thank you,

John


P.S. Mark, I really like your contributions to Myth. Please do not
take this is an attack. I just want to make sure I am using the tools
correctly.

--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


mark at digivation

Feb 17, 2008, 10:37 PM

Post #7 of 10 (1330 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

> Changeset 15893 breaks passthru for me (emu10k1). The reduction in
> buffer_time throws underruns and I get no sound. Setting this back to
>500000 fixes it.

how does mythmusic like this setting?
this impacts time to paused. I thought 1/2 second was excessive.
does 200000 work?
is it only passthru or does the same thing happen with analog out?

>
>My fragment_size also gets calculated to 6400 which is not cleanly
>divisible for 6 channels @ 16-bits. It still works like this, but I'm
>not sure what impact it actually has.

this gets calculated every time and is only relevant to the current #
channels requested.
so 6400 for a passthru which has a pseudo channel count of 2 is correct.

mark

_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


ngarratt at gmail

Feb 18, 2008, 9:20 AM

Post #8 of 10 (1307 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

> how does mythmusic like this setting?
> this impacts time to paused. I thought 1/2 second was excessive.
> does 200000 work?
> is it only passthru or does the same thing happen with analog out?

At anything less than 40000 I still get occasional loss of audio. This
only affects passthru; normal audio (which is sent to a PCM device
that routes over IEC958 to my receiver) works correctly with 10000.
Given the variety of hardware out there, 50000 is probably the safest
bet.

I don't use MythMusic, but I can built it if necessary for testing.


I've also looked at timestretch on AC3 streams and hit some problems.
When enabling timestretch it switches from my passthru device to my
default device (which only handles PCM and the card routes to S/PDIF).
Obviously this doesn't work and I just get clicking.

Setting my default device to iec958 fixes this, but I have volume
problems with PCM sent directly to this device so it's not really a
solution. Shouldn't it continue to use the original device given that
it's still going to be the same type of stream?

Lastly, I only get 2-channel AC-3 when in timestretch. You mentioned
that it should re-encode to the original number of channels?

This is with internal FFmpeg AC-3 and FFTW3; if I switch to A52 I get
garbled sound when in timestretch.
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


jppoet at gmail

Feb 18, 2008, 9:40 AM

Post #9 of 10 (1311 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

On Feb 18, 2008 10:20 AM, Neil Garratt <ngarratt [at] gmail> wrote:
>
<snip>
>
> Lastly, I only get 2-channel AC-3 when in timestretch. You mentioned
> that it should re-encode to the original number of channels?
>

There is a new setup option "Max Audio Channels" which you need to
change to 5.1. I believe it is under "General".

John
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev


nigel at ind

Mar 6, 2008, 3:52 PM

Post #10 of 10 (1214 views)
Permalink
Re: [mythtv-commits] Ticket #1104: multi channel audiosupport [In reply to]

On 14/02/2008, at 1:14 PM, Mark Spieth wrote:

>> 2) Change I was asking about was trivial swab-and-tidy.patch.
>> Add unsigned char * payload = data + 8;
>> Hoping to make a little more readable.
>
> I have no problem with that and it makes sense.

Thanks. Applied to head.


> I just copied code form avf
> where it was originally (and still is I think)

I went looking (so I could commit upstream), but couldn't find it?

--
Nigel Pearson, nigel [at] ind| "Peacemaking is our profession"
Telstra Net. Eng., Sydney, Australia |Motto of Burpelson Airforce Base
Office: 9202 3900 Fax: 9261 3912 |Commanded by Gen. Jack D. Ripper
Mobile: 0408 664435 Home: 9792 6998 | Dr Strangelove
_______________________________________________
mythtv-dev mailing list
mythtv-dev [at] mythtv
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev

MythTV dev RSS feed   Index | Next | Previous | View Threaded
 
 


Interested in having your list archived? Contact Gossamer Threads
 
  Web Applications & Managed Hosting Powered by Gossamer Threads Inc.