[mythtv-users] Video Quality
ziplux at comcast.net
Thu Mar 11 19:42:36 EST 2004
You're able to capture at 640x480 on an Athlong 1GHz? I have an Athlon XP
1800+ and it can barely do MPEG4 at 480x480 with no filters. Are there any
performance tweaks you did, or maybe something I'm doing wrong?
From: mythtv-users-bounces at mythtv.org
[mailto:mythtv-users-bounces at mythtv.org]On Behalf Of Alan Snyder
Sent: Thursday, March 11, 2004 12:29 PM
To: 'Discussion about mythtv'
Subject: RE: [mythtv-users] Video Quality
> I am in the process of ironing out a few kinks with my mythtv
> but I was wondering about what type of quality people are
> experiencing with their recordings off of digital cable/cable
> TV. I'm using an Osprey 200, an entry level video card.
> How much of a video quality upgrade does one get by moving
> from a basic card like that to one of the PVR cards like the
> Hauppauge PVR 350. For example does the on board Mpeg 2 make
> that much of a difference? After all can't one just compress
> the signal after capture (sacrificing time and HD space for
> $$$) or is there more to it than that.
> Perhaps people could post samples of video captured. I don't
> want to get anyone in trouble - maybe they could just post
> something off of public TV etc... I just want to get a sense
> of the quality. For example Tivo's high quality setting looks
> to my eye exactly like the broadcast signal coming in
> - there are no artifacts on a TV. Is this possible with a
> Linux PVR rig? Also is it only possible with the higher end
> capture cards?
> Personally I probably won't want to use mythtv to actually
> watch my recordings. I'm perfectly content saving them to
> DVD-R and watching them on my DVD player. I'm just wondering
> if that will look as good as the original broadcast or will
> it look strange. How successful is mythtv as a digital VCR?
First principle: No matter what hardware/software you use,
capture (frame grab) can be perfect for all practical purposes
but the compress-uncompress process involves some loss (assuming
no use of lossless codecs like huffyuv).
The key elements are avoiding unecessary loss due to poor
cable signals, poor tuner quality, poor video processing/
digitizing, etc., then looking for compression methods that
keep signal loss and artifacts as un-noticable as possible
on your display device.
With a dirt-cheap bt878 capture card, generic
GEForce-2 video card, and mid-90's 27" TV, with myth doing
MPEG-4 at 640x480, 2200kbps, the difference between live TV
through myth and live TV off the cable is noticable but
not bothersome. The CNN "crawl" is a little jumpy. There's
some loss of detail in the highlights (I think that might
be a bt878 chipset issue). Up close (closer than I watch),
some blockiness is present (replacing the less noticable noise
in an analog picture). Football can seem harder to follow,
probably because of the limitations of the compression settings
to handle the frame-to-frame changes in content. But overall,
the difference is similar to the difference between a good cable
signal and an "OK" cable signal. I do notice that for certain
channels where the cable signal is weaker, the noise in the
original results in more artifacts in the processed video.
This is most likely because the compression scheme needs to
deal with all that extra "information."
It would be interesting to hear how hardware MPEG-2 compares
with MPEG-4 in a system without my CPU limitations. I can't go
beyond the settings listed above with my 1 GHz Athlon.
More information about the mythtv-users