Is CCE horrible with interlaced NTSC DVD's (TV episodes)

[raddygast]

For the first time ever, I tried to use DVD-RB and CCE 2.70 on a DVD that was television episodes. The original source looks pretty crap on my screen to begin with -- it's a show from the early nineties and definitely not in HD, maybe even a bad transfer.

It's "My So-Called Life". Anyway, there are four episodes per DL disc, which is like 3.5 hours roughly. I tried using DVD-RB on it, and I optimized the QMats by using QMatOp (tolerance=2, framesplit=12, QualityPrec,VbrBias,GOPAuto all set to 1). I used 4-pass CCE!

Anyway, the result not only looks like crap (which is maybe not so surprising), but also the bitrate seems to be higher than the original! I watched a few parts of the video at the chapter breaks and the original was peaking at say 2.79 Mbps and the re-encoded output is now peaking at say 3.23 Mbps at the same approximate spot. What gives there? I can see with my eyes that the quality is far worse than the original, and yet the bitrate is higher?

I must be missing some obvious technical fact. Can someone point it out? And can I have some advice -- is it better to use DVD Shrink or something to copy TV stuff, or can RB do a decent job? Should I switch to HC for TV discs?

Any advice really appreciated. I really don't want to burn the disc the way it is -- it looks horrible.

[hutch1711]

Quote "You can't make a silk purse out of a sow's ear"
The meaning being there is very little you can do to improve the quality of the source material. If it's crap to begin with, the output will generally also be crap. Sofar as the bitrates are concerned, look at it like being an envelope. You have your source data with all the bits that make up the movie in a small envelope. You take thoses source bits and process them as output into a larger envelope. You haven't increased the number of source bits, all you have provided is a bigger space to distribute those bits, there is no increase in quality.

[raddygast]

I am obviously not expecting to improve the quality. I am just confused as to why the bitrate goes up when the quality goes down. Or for that matter, why the bitrate would go up at all.

I'm also very confused that the bitrate would go up and the entire DVD would fit on a DVD-5, whereas the original source had to fit on a DVD-9 (but seemed to have a lower bitrate).

I think this may all stem from being interlaced. There's something I'm not understanding about that part of it. Interlaced stuff generally looks horrible on computer monitors, correct?

[hutch1711]

Sorry it's late and I must have suffered a brain cramp. I'm sure I read somewhere that DVD Rebuilder encodes cells according to their original bitrate. If yours is different I guess we'll both have to wait and see what the real experts have to say. And yes interlaced material does look terrible on a computer monitor. Cheers :)

[jdobbs]

You are talking about peak bitrates. It could very easily peak higher if the encoder decided it should. That's really not unusual. For example -- think about what would happen if a CBR source were fed to the encoder and converted to VBR, the peaks would look nothing alike... because the original had no peaks.

[Fishman0919]

and I optimized the QMats by using QMatOp (tolerance=2, framesplit=12, QualityPrec,VbrBias,GOPAuto all set to 1).

Make sure you have Animation=0 with QmatOp or you will get bad results with a reg movie.
Also make sure in dvd-rb you don't have Half D1 or Steal Space from Extra on both not a good idea with television episode discs.

[raddygast]

Yeah, animation is off of course, and I never use Half D1 or steal space.

[onesoul]

For interlaced content try hc or quenc instead.

[borgraf]

I'm certainly no expert on interlacing, matrices etc. but I recently tried QMatOp on a 3 hour source and the result was pretty bad. I used QLB matrix instead and it came out fine, you could also try with KVCD matrix. Or encode with ProCoder if you have it, I always get good result with high compression.

bon voyage :)