Archived from groups: alt.video.laserdisc (
More info?)
"unclejr" <watsona@kenyon.edu> wrote in message
news:139de3b3.0408220524.38193131@posting.google.com...
> NFP <NFP@my.group.com> wrote...
> > I have a Pioneer CLD-2950 and disappointed in the excessive softening of
> > the comb filter for s-video output. Unfortunately I don't have a good
> > comb filter in my display device either.
> >
> > Are there any better external comb filters available and where would I
> > find them?
>
> I use the Camelot Technologies Crystal Vision VPS-1. I don't know if
> it does PAL, though. Here ia a direct comparison between it and the
> Faroudja VP-100:
>
> <http
/www.hometheaterhifi.com/volume_5_1/crystalvision.html>
>
> I bought mine on eBay for $150-200 a couple of years ago. Excellent
> little unit.
>
I tried to buy a Camelot comb a couple of years ago, but they quit
making/selling them. However, that comb was a 'really good' 2d comb
instead of a 'really good' 3d comb. (A 3d comb can actually be worse
than a 'really good' 2d comb.)
For my own purposes, I never watch LDs anymore, and only archive my
LDs onto D9, DVD and/or DV25. The combing and processing that I do
for my LDs provides better performance than one might normally achieve
with even the best 3D comb filter.
I know that this is semi-off-topic, but it is sometimes good to look towards
the future. I still love LD, but want to get the best video of my old LD
stuff
and keep perfect copies.
So, I use a TBC that even corrects very well for LD scratches, the TBC has
NR (but is too aggressive, except for the worst LDs) that can be seamlessly
enabled/disabled. Then, I use a Canopus ADVC300, which has a damned good
3D comb and damned good NR (little or no comet tailing at reasonable
noise reduction levels -- I always use the lowest levels.) By the time the
signal is processed by these two devices, and encoded into DV25, the result
is a relatively less noisy version of LD video. IF you don't NR and 3D comb
the
LD signal before doing DV25 conversion, you'll collect numerous degradations
of the signal that result from the DCT truncation (which is made very bad
because
of the LD noise and potential NTSC decoding artifacts.) If you remove the
NTSC
decoding artifacts, and remove much of the 'analog' noise, and also totally
remove
the TBC errors, the DV25 encoding does a very good job of making a near
perfect
copy of the LD.
If you look at the DV25 specs, you'll notice that superficially DV25 video
quality
fully encompasses LD. DV25 WILL NOT accurately reproduce normally noisy
LD signals, and is probably one reason why people don't feel comfortable
backing
up an LD onto DV25. Another reason is that DV25 produces VERY LARGE data
files,
and AFAIR doesn't really support 5 channels of audio. For my purposes, 2Ch
is good
enough.
After getting a slightly 'improved' copy of the LD signal, then I encode the
results
by using TMPGENC and spending LOTS of time. The noise reduction provided
by TMPGENC appears to be very good, and doesn't seem to produce excessive
comet-tailing or other evil results. Also, I perform some correction of the
transfer
function (probably resulting from group delay problems), and avoid over
processing.
The DVD result of the LD doesn't really appear to be sourced from LD, and
looks
damned good. The image has no perceptable mosquitos (because I use plenty
of
bandwidth), no significant noise, and significant correction for group delay
problems
in the analog video. The major negative result of my setup is the loss of
anything more
than 2 channels of audio.
The bottom line of my results:
1) Very minimal 'random' analog video noise.
2) Total elimination of the large area chroma instability.
(looks similar to normal pro analog video quality)
3) A scratch that would originally cause a quick loss of sync
only cause a small color glitch.
4) Sharpening of edges ONLY to the extent of partial correction
of group-delay like artifacts. Small amounts of 'ringing' that might
have originally existed in the low-quality master are partially
removed.
5) Passable image quality on an HDTV.
6) My resulting sharpness (and video frequency response) is every bit as
good as LD can attain.
If I simply want a 'perfect' copy of an LD, I just use a D9 copy, which
retains
all of the noise and large area chroma noise. However, I can then re-encode
the signal by using the above setup with no real loss. When doing a D9
copy,
I definitely use a pro-level DPS290 TBC, because the time base instability
of
the color-only-corrected LD isn't sufficiently good for my D9 equipment.
(to the extent that moving the LD player causes some time base instability
that effects the ability for the VCR to lock onto the signal.)
Without using the noise reduction, the resulting DVD looks WORSE than
an LD. Using the TMPGENC NR alone without the ADVC300 does
provide some help, but it is critical to do NR before the DV25 encoding, or
the DV25 encoding along with the video noise will cause very significant
and permanent quality loss.
John
**
Facebook
Twitter