Hello,

I recently bought a PCI tuner card (Medion 7134 with tuner Philips FM1216ME MK3)
to try saving to a PC contents of some old VCR cassets using a miraculously still
functioning VCR. The card and the tuner got autodetected correctly (and anyway
I know their numbers from cards and tuners lists) and I started using VLC
to tune onto the VCR signal and save to a file. At first I attempted saving
as MP4 but got 100% CPU on all 4 cores (AMD X4 2+ GHz)and viewing the resulting
file revealed quite a few missing frames and freezes so I switched to saving
as MP2 ("MPEG-2+MPGA"). This time the saved video was OK (and only 25% CPU) but
once in a minute or two I noticed a gap in the sound - a gap of about 0.5-1
second. I experimented with different audio encoders avaliable in the VLC
"Convert/Save" function but even when selecting "WAV" (as I understand, just
dump the sound with no compressing) still I've got the "gaps" in the sound.
The periods between gaps are not periodic - their variance is very high,
sometimes only few seconds between one "gap" and the next but for most of the
time - it's about once in a minute or two.

How can I debug this?
Where to start looking at?
Contents of what files/outputs of what commands
can be usefull for debugging this problem?

TIA!

Not sure if it makes a difference: Did you try some real capture-application? I just don't trust vlc's stream-save-feature. Tried movie-conversion with vlc this way a few times. Never got satisfiying results. Switching to some real encoding-app did help. Maybe it's the same with capturing TV. Don't know which program to use instead though.

Hi "whocares02", thanks for the reply.

Before "VLC" I attempted to use "mencoder/mplayer" but without success:
I don't know how to tune "mencoder" to a particulat custom frequency (591750KHz, from "tvtime" scanner output).

Anyway, I have some more info regarding the VLC behavior: I wrote down the exact
times (hour:min:sec) when the sound gaps appear (I always start from exact beginning
of the cassete in the VCR). Then I repeated it, expecting the times to change at random
with some mean and variance. But what I've got were EXACTLY (to the second) times
I wrote down previously. I repeated the process few times and each time, the "gaps" appeared
at exact times. Rarely, a gap didn't appeat in its "slot" but when it did - it was the same
"slot" always, accurate to the second.

So now it seems like an input problem: in particular places on the cassete (exact times)
something causes the tuner to lose the audio (at least, that's my current theory).

I suspected some "spike" of the signal causing the tuner to lose momentary the audio.

That's all I gathered so far,
TIA for any ideas,

Some more info:

I attempted to capture with "mencoder" again, using the following configuration file lines:

Although the created avi had the video running about twice the speed
of audio but I managed to listen to the exactly same "gaps" I heared
using "vlc", so it doesn't looks like "vlc" problem.

I'm now beginning to suspect either the TV tuner card or the ALSA.
The fact that the "gaps" occur at specific (to the second) times
causes me to suspect the tuner card more than ALSA. Let's assume
I'll manage to get the audio signal from the VCR (or the TV) and connect
it to the "Line IN" of the m-board. What should I change in the VLC
in order to get the audio (together with the TV video) from the "Line IN"
instead of from the TV tuner audio?

TIA!