[mythtv-users] Underwater audio

[Curtis Stanford]

On 29-Feb-08, at 2:18 AM, Gregorio Gervasio, Jr. wrote:

>>>>>> Michael T Dean writes:
>
> m> On 02/28/2008 07:10 PM, Curtis Stanford wrote:
>
>>>> I think I have the same problem with 32kHz audio in NUV recordings
>>>> (framegrabber). The patch in this ticket (reverts some rev 15893
>>>> changes) seems to fix the problem for me:
>>>>
>>>> http://svn.mythtv.org/trac/ticket/4764
>
>>> That totally fixed my problem. Thank you!
>
> m> Would be /really/ nice is some of you could test the value that's  
> being
> m> changed by that patch and find the smallest value that works...   
> Try
> m> starting at 120000.  See:
> m> http://www.gossamer-threads.com/lists/mythtv/dev/319315#319315
>
>
>        Based on limited testing, it's "fragment_size" rather than
> "buffer_time" that's affecting this specific "underwater" problem:
>
> 1. (rev 15893)
>   fragment_size = (audio_bits * audio_channels * audio_samplerate) /  
> (8*30);
>   buffer_time = 100000;
> -> BAD -- underwater effect
>
> 2. (pre 15893)
>   fragment_size = 6144;
>   buffer_time = 500000;
> -> OKAY
>
> 3.
>   fragment_size = (audio_bits * audio_channels * audio_samplerate) /  
> (8*30);
>   buffer_time = 120000;
> -> BAD -- underwater effect
>
> 4.
>   fragment_size = (audio_bits * audio_channels * audio_samplerate) /  
> (8*30);
>   buffer_time = 500000;
> -> BAD -- underwater effect
>
> 5.
>   fragment_size = 6144;
>   buffer_time = 100000;
> -> better -- sounds okay but audio is sometimes lost when skipping
> around (need to exit and restart)
>
> 6.
>   fragment_size = 6144;
>   buffer_time = 120000;
> -> OKAY
>
>        I will try out this last setting for a while.

I haven't had time to try lots of combinations but I can confirm that  
#6 works for me as well.

Curtis