CCE 2.67 Settings

[GParent]

USE:

DVD2AVI 1.77.3
VFAPI 1.05
CCE 2.67.00.10 (No Guides Exist that Refer to this Version)

I have successfully joined the two sides of GOODFELLAS using some alchemy and a prayer or two. <g> My DVD of GOODFELLAS is a FLIPPER release and I now have one complete and unified set of VOBs, that works with Power DVD Player.

GOODFELLAS over 7GB in VOBS:
MPEG-2
720*480
4:3 LETTERBOXED
FRAME RATE: 29.97
NOM BITRATE: 7500000
VBV BUFFER SIZE: 112
CHROMA 4:2:0
DCT PRECISION: 8
PIC STRUCT FRAME
BOTTOM FIELD FIRST
FRAME PROGRESSIVE
ZIGZAG
QUANTSCALE NONLINEAR
_____________________________________________________________________

I am Frameserving GOODFELLAS into CCE 2.67.00.10 (DVD2AVI/VFAPI) and have setup:

TWO PASS VBR
PROGRESSIVE
ZIGZAG
DCT = 8
LUMINANCE LEVEL = 16-235
IMAGE QUALITY = 15
NOISE FILTER OFF
MIN BITRATE = 0
MAX BITRATE = 7500
AVG BITRATE = 4500
OFFSET = 1 (Same as UFF = YES)

NOTE: I made two CCE jobs. Starting with a One Pass VBR copying the ROBA Guide. I then did a two pass VBR changing the settings to the above, having retained the hints found in the first run (VAF).

I now have a 4.9GB M2V file that needs re-encoding.
_____________________________________________________________________

QUESTIONS:

(1) Can I keep retain my intermediary VAF file for the third run, or should I start from scratch?

(2) I plan to downgrade the AVG BITRATE to 4000 and check the fit. Should I keep the min at ZERO and the MAX at 7500?

(3) Under CCE 2.67, (OFFSET = 1) = (UPPER FIELD FIRST = YES).

I set OFFSET = 1, following the ROBA Guide of UFF = YES. Since the source is 99% Progressive Frames that are BFF, the CCE Guide would say to go with OFFSET = 0. Why?

Is it possible the answer has to do with Progressive frames not needing the OFFSET setting to work at all, so it doesn't matter if its set to 0 or 1. Therefore, why not set it to OFFSET = 1, which would be the best setting for the 1% of Interlaced Frames that are BFF.

Is this a stretch or does it make sense?

(4) The CCE Help says that DCT = 8 may give better quality for lower bitrates. Since its DCT = 8 in the source, I just left it that way. Did I make the right decision or should this be DCT = 10?

(5) I used 15 for Image Quality but I still feel the jury is out on this one. I have read lots of opinion but I am awaiting some general rule or tool that would give a better hint as to what it should be.

(6) The whole Luminance Level debate has me wondering if I might not understand its use. Isn't 0 to 255 the ONLY setting if your DVD will be used on a TV system? I went ahead and chose 16 to 235 as the Guides ALL recommend. But I still wonder why?

(7) When would I use 3:2 PULLDOWN on the Main Window of CCE 2.67????

NOTE: I have never once read a Guide that shows how, why, and when this setting should be used. The CCE 2.66 Manual has a whole section on 3:2 Pulldown. The mechanics are there, I just don't know how I could apply it to making GOODFELLAS look better.

_____________________________________________________________________

Thanks, I'm sorry if I asked too elementary questions but I really have read many of the messages and Guides both here and on other forums and newsgroups. As you know, there is a lot, make that, A LOT, of dispute about the same things. I follow the CCE FAQ, and the Two CCE Guides on DOOM.

[wmansir]

I'll try and answer what I can.

1. You can keep the .vaf. I think the only times you have to change the .vaf is when you change the source or turn on/off Pulldown detection.

2. You really need to use a bitrate calc. TryTHIS (http://www.dvdrhelp.com/calc.htm) online one from dvdrhelp.com, or you can download one like FitCD. Just fill in the settings for you movie time, DVD-R size, and audio bitrate and it will give you good results. You can keep your MAX and MIN as they don't affect the size of the .mpv, only the AVG bitrate does that.

3.It really doesn't matter for progressive material, so you can leave Line Offset at 1. Since your encoding as progressive, that 1% interlaced (if you notice it at all) is not going to look right regardless of whether you encode TFF or BFF, because it's at 24fps and not 30fps. But in all likelyhood you won't even see that 1%.

As is discussed in the FAQ all Line Offset does is shift the video up by 1 pixel. If your picture were full frame this would result in a virtually unnoticable 1-pixel line across the bottom of the picture, but since your video is letterboxed you don't even see that line.

4. I haven't played with DTC precision, especially not with DVD-R, so I really can't answer this. But I was thinking along the same lines as you regarding CCE's DTC vs. the source's DTC.

5. Again this is something I can't answer. For my SVCDs I prefer a relatively low IQ level, but I'm sure it's different for DVD-R.

6. Since your using Vfapi this matters for you. To be safe I would recommend using 16-235, since that's within NTSC TV specs. However if you check your settings in DVD2AVI it may have already done that.

In DVD2AVI under Video:YUV->RGB you can select PC Scale (0-255) or TV scale (16-235). If you select TV scale there you can use 0-255 in CCE, otherwise you should use 16-235.

As to why check This (http://forum.doom9.org/showthread.php?s=&threadid=28815&highlight=scale+luminance) thread. Note that most of the people saying to use 0-255 are refering to AviSynth usage. Using Vfapi, with DVD2AVI set to PC scale, is a special case that requires CCE set to 16-235.

7. Only use that setting on so-called Hybrid material. This is a stream that is a mixture of Video and Film.

Normally there are two ways to encode an NTSC stream, as Progressive Film or Interlaced Video. Video is encoded at 30fps, but Film is encoded at 24fps and then pulldown flags are added to the stream. These pulldown flags tell your DVD player to change the 24fps Film into 30fps Video on playback thru a process called Telecine (because your TV can only dispaly 30fps).

To use Pulldown detection you feed CCE 30fps. It examines the video and looks for parts that are being telecined. It then performs Inverse Telecine (IVTC) and restores the original 24fps FILM, and adds the pulldown flags to just that part of the stream. If it doesn't detect pulldown it encodes as regular Video at 30fps.

Normal DVDs are encoded like this all the time. In fact that is what DVD2AVI is telling you when it says your DVD is say 10% Video. That means 90% of your DVD has pulldown flags and 10% does not. When you use Force Film in D2V you are telling the program to ignore the pulldown flags and just output the progressive film.

[GParent]

>>> Normal DVDs are encoded like this all the time. In fact that is what DVD2AVI is telling you when it says your DVD is say 10% Video. That means 90% of your DVD has pulldown flags and 10% does not. When you use Force Film in D2V you are telling the program to ignore the pulldown flags and just output the progressive film.<<<

DVD2AVI and CCE, I read everything I can, including FAQs. As you know, information varies and even the gurus disagree on things, but its important to ask questions and I appreciate your reply. I just want to cast a vote for a future guide. Since DVD2AVI and CCE (or TMPQEnc) are so Tightly_Coupled, and so determinant of the end quality of your DVD, that it would be good to have a single source, a Bible, for how the parameters dovetail between the two. FORCE FILM in D2V and Pulldown in CCE, Luminance, et al.

Thanks Again,

Pat