Re: AW: glitches, another story *g* - more bad news ;-(

[Guido Fiala <gfiala@s.netic.de>]

Am Thu, 12 Oct 2000 schrieb Guido Fiala:
>Am Thu, 12 Oct 2000 schrieb Schuett Thomas:
>>Hi Guido,
>>
>>>Another thing: even with the "small" 256k buffers i could not produce any
>>>record-buffer-fulls recording with "cat /dev/video > test.pav".
>>
>>I have glitches in the record (havent checked logfile, but it is in the record,
>>on playback always the same positions etc.) when recording with 
>>cat /dev/video > file. I am just going to try buffer -i /dev/video > file
>>to see if it makes a difference.
>>
>>But after all the solutions that I have seen in the ml, and that worked for
>>some and not for others, I am really close to sure, that there are
>>
>>           several reasons for the glitches.
>>
>>We should have a matrix of possible solutions and who has tried
>>what and if it worked completely/partial/not.
>>
>I made further tests with different outstream-formats - it seems that there are
>at least 2 occurences of recorded glitches: the one with
>buffer-full/missed-video-data producing blocking-failures and another type
>- even if there are no errors shown in log -
>producing audio/video-latency with flashing image disturbances every a few
>seconds.
>
Cooling seems not to help! (lmsensors report now 5 degree less temperature)
Still to much glitches (with no buffer-full-syslogs)

But something to think about:
Watching TV the same time does not show any failures, these are just in the
recorded files - i suspect that only these part of the card/chip is
affected/buggy.
My feeling was, it started with the later drivers (>0.5). Will make some tests
with "cat /dev/video >test.pav" to check it.

But it would help, if someone can verify if these failures happen under W95 too,
who can do it?