Sony YUV and Canopus Procoder: levels problem

john-beale wrote on 3/22/2006, 1:16 PM
This seems to be a problem with Canopus Procoder, not Vegas, but maybe someone here has advice. Procoder is expanding my ITU-R BT.601 levels (16-235) to 0-255 RGB levels and causing clipping in the process, when converting Sony YUV source avi files. It leaves levels unchanged when converting DV source avi files.

I also have CCE Basic which has a BT.601 level setting option, but my tests on my material with my bitrates show Procoder has the best image MPEG2 image quality. Using Sony YUV (4:2:2) instead of DV (4:1:1) as the intermediate file clearly gives better resolution on red objects on the final DVD, but the levels problem is killing me. I could use a virtual intermediate (frameserver) but the conversion time almost doubles (2-pass MPEG2 encoding). I'd rather not pre-compress levels in Vegas to (32-215) since that's throwing away information and causing some posterization; I shouldn't have to do that. What is the best way to fix this?


Specific example of the problem:
Viewing my HDV-intermediate project in the Vegas 6.0d timeline, the RGB Parade monitor view shows my R,G,B values range from 20..235.
I rendered this to Sony YUV .avi (720x480 widescreen), and re-imported that back to Vegas as a check. The R,G,B values range from 20..235 as expected.

I then converted the avi to MPEG2 using Canopus ProCoder Express, and as a final check I imported this MPEG file into to Vegas.

Now my R,G,B values range from 6..255 and the image is too contrasty in Vegas and in all my MPEG2 viewers as well.




Comments

john-beale wrote on 3/22/2006, 2:55 PM
update: I'm now using the DebugMode frameserver to export from Vegas (as RGB24) and feeding that into Procoder. This is slow with 2-pass encoding, but it works OK. With this workflow the video levels stay correct through to the output MPEG.
plasmavideo wrote on 3/23/2006, 9:14 AM
Just out of curiosity, what are you gaining by using Procoder over the Mainconcept encoder for mpeg?
john-beale wrote on 3/25/2006, 3:04 PM
I haven't tested Mainconcept lately. When I used it before it had inferior quality as compared with Procoder on my high-motion low bitrate 2-pass encodes.
Lots of people argue about which encoder is better but it depends on a lot of details.

With my cameras, techniques, subject material, and bitrates, my tests on my display monitors show Procoder output looks better to me (fewer noticible artifacts).