PDA

View Full Version : What CCE-settings for an interlaced PAL-DVD


Deltis
11th January 2002, 18:12
Hi!

First of all - sorry for the cross-posting - I got no reaction in the DVD2SVCD-section and don't know how to move a message...

I recently made "Good Morning Vietnam" (PAL) that has been identified by DVD2AVI as interlaced source. To explain in details: The DVD was made in interlaced-mode but the source was a cinema-movie(progressive), so there are no true interlaced-artifacts visible. Using DivX you don't have to perform a deinterlacing with that kind of DVDs


By now I read two guides on DVD2SVCD/SVCD-conversion written by doom9 and robshot. They seem to advice different settings:

Robshot advises: "To make the video parameter exactly the same as the original, click on the Video button. Compare the menu below with the Bit rate Viewer session. As you can see, the parameters are: DVD compliant, Zigzag scanning order. If frame Type (in Bitrate viewer) is interlaced then uncheck the Progressive frames in CCE because while the picture structure is FRAME (it's from FILM transfer), the frame type is interlaced (PAL video, remember?). Otherwise, set it to Progressive. Upper field first = Field top first. I changed the Quant scale to Linear to get better quality. DCT precision is set to auto."

doom9 writes: "The default for the anti-noise filter seems good to me. Also leave the other options checked. (Progressive=ON, ZigZag=ON, Linear Quantizer Scale=ON)"

CCE manual states: "However, the non-linear quantization scale is recommended unless there is a special reason."


- Does it really matter that much considering quality if I encode the above DVD in progressive-mode?

- What errors (in the worst case) can be seen on TV during playback of a SVCD made this way?

- What quantizer-scale is best for SVCD?

Thanks!
Thomas

Doom9
23rd January 2002, 20:01
afaik the frame format is just a flag in the mpeg2 stream and has no real influence, though I could be wrong and I have no means of testing it on a standalone.. I'm afraid you'll have to make your own tests here.

as for the quant scale... who do you trust more? me and rob or the cce manual? ;) once again.. I have no means of properly verifying so you must do so on your own.

Deltis
24th January 2002, 02:11
Hi doom9!

I recently finished converting an interlaced DVD ("Tomb Raider") to an interlaced SVCD. Alternate scanning was enabled. Until now I can't see any difference on my PC-screen nor on my TV to progressive-encoded SVCDs of an interlaced DVD-source (mentioned frame-type). Quality is good as ever :)

I just asked as I have some "combing edges" in scenes with fast moving red areas in my progressive-encoded SVCD of "Charlie's Angels". But for the rest it doesn't have the typical interlace-errors that appear in case of a missing deinterlace-stage. Therefore I presumed it might have had something to do with having choosen progressive encoding instead of interlaced encoding.

If you've got further information on how to handle interlaced DVDs which weren't initially recorded interlaced (unlike captured TV-movies), please post it in one of your guides ;)

Until then I'll stick to progressive encoding as I can't see any differences (beside the case with "Charlie's Angels"). I hope the future will become a progressive future - and I'll be ready for it :D

cu
Thomas

TactX
24th January 2002, 12:39
Quoted from the Berkeley Multimedia Research Center MPEG FAQ (http://bmrc.berkeley.edu/frame/research/mpeg/mpeg2faq.html)

Non-linear macroblock quantization was introduced in MPEG-2 to increase the precision of quantization at high bit rates (hence, low quantiser_scale values), while increasing the dynamic range for low bit rate use where larger step size is needed. The quantization_scale_code is switchable between the linear (MPEG-1 style) or non-linear scale on a picture coding (frame or field) basis. This new MPEG-2 non-linear scale corresponds to a dynamic range of 0.5 to 54 with respect to the old linear (MPEG-1 style) range of 1 to 31.

In other words: for MPEG-2 use non-linear.