Picture quality is good on both PVR-250 and PVR-500. Sound is good on
PVR-250. Sound is bad on PVR-500: Overdriven and scratchy. No idea
which versions the sound problem is present for: The last 0.3.4 version
I tested was 'f' which had video issues for me but the sound was ok. I
have no options set in modprobe.conf other than the tuner id's and
no_black_magic=1 for the cx25840. Currently using version 0.3.2w
Bill
Chris Kennedy wrote:
This contains some more patches from Axel Thimm, one from David DeHaven
for fixing changing inputs during audio switches, and another for fixing
switching TV/FM in tda9887.
Also I have separated the VBI work queues from encoding, because first the
decoder VBI work is not encoder related at all (holding back the others),
and vbi work can go on simultaneous with mpg work.
This seems to fix alot of issue with re-insertion, it's still not perfect,
but better. I also fixed the way CC decoder output works to have a queue
of characters and collect them and send them locking better too, so under
high CPU loads it can flex and store what needs to be displayed. It seems
our timing is not always perfect, and we probably need to even delay output
and match it with the current vsync pts time, since decoding delays output
and we bypass the decoder output with input to the saa7127 directly, so
this has alot of interesting possibilities in the future.
So everyone with VBI needs, test this, it could be a big improvement I think,
I'm still wondering if DMA mode isn't good for sliced VBI since the data
packets are so small, but may just be the buffer sizes it has problems
with (so possibly need to increase the for decoder VBI too, and then DMA can
be used for each), but that's to be experimented with later, this seems to
fix things enough for my needs right now :-). It is a very tricky thing to
output this vbi data into the saa7127 and I've just touched the tip of the
iceburg on it, anyone is welcome to come up with more advanced pts matching
schemes and methods of feeding it, lots of interesting things could be coded,
so something fun to look at for someone ;-).
#0.3.4p: http://www.ivtv.tv/releases/ivtv-0.3
Thanks,
Chris
-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
_______________________________________________
ivtv-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ivtv-devel
