Help with CCE 2.50 SP - 4 Pass VBR - Which M2V to use?

[digital_EUR]

hi there everyone

sorry for my question if it has been already done by someone else.
i am beginning to use CCE 2.50 doing a 4 pass VBR of a Depeche Mode concert that didn´t came out with quality using either Instant Copy, Dvd2One or Dvd Shrink. All of them failed to achieve a minimum quality in this DVD. I think is due to the hard encoding cenarios that were used in the concert, lots of extreme blue, red and green backgrounds.

It is the "One Night In Paris" in case someone else knows it.

Well, in anycase i decided to give CCE his first try.
In my first attempt i used 4 pass VBR for an average of 4000 in a interval between 500 ( i think this value is more safe than 0 but give me your opinion ) and 8500 ( this max value was the same as the original, not sure if CCE would gain with raising this value ).
The Quality Priority was the value of 4.

After 7 hours it gave me the result that after compiled in Dvd Maestro result in 3.72GB DVD.

The quality of the image was promising and knowing that i still could rise the average value i was beginning to fall in love with CCE ( LOL ).

So the next day i used the same parameters just moving the average bitrate to 4700 and after another 7 hours the result was a 4.32 DVD.

Well the size is perfect. I am going to sample the image quality after i post this just because i was wondering one thing:

CCE gave me two M2V files that from what i have read alternate in the encoding process in each pass.

But they are the same size, which should i use in the authoring process?
The one with the most recent date? The one correspondin to the first or second text box in the output definition of CCE?

Now for something basic:

( based on your experience tell me this )

- the 5th pass or higher is worth the time? the gains are noticeable?
- the quality prority parameter lower than 4 is a good idea?
- all the guides say that in a Interlaced source one should deactivate the zig zag scaning option but i didnt because i just read the guides after the encoding ( not a good idea i know ) but will the alternate scaning give me such a better result?


Software used:

DVD2AVI 1.77
VFAPI 1.05
CCE 2.50 SP


Thanks for all your patience and for any help on this subject

digital

[Skynet107]

Are you using a FAT32 drive? That would explain the 2 files split. I have never had CCE split any file on me and I have used it hundreds of times. I recommend 5 pass encoding it is the best compromise between quality and processor time. Many MANY people here will argue like hell that CCE is not the best and 5 pass encoding is a waste of time but I disagree on both counts, CCE is the best and 5 pass does make a difference on tricky material. And Yes do not use zigzag for interlaced material it will look better using the alternate option. I like to use a min. setting of 2000, sometimes I use a min. of zero but I have not seen any difference doing this and it does not save file space.

[BJ_M]

you would save yourself a lot of time if you used a bitrate calculator

https://www.videohelp.com/calc

you have a lot of expensive tools -- you should upgrade to the latest version of CCE sp ..


5th pass not worth it


alternate scaning will give a slightly better result for interlaced - though a number of other factors will make more of a diff.

[digital_EUR]

thanks for your opinions...

no, i am not using any FAT32 drive, only NTFS drives.
to be more precise, 2 drives of 80 GB each ( SATA ) being one the source and the other the output of the encoding, and 1 drive of 40 GB ( IDE ) where the DVD2AVI and VFAPI projects, the ECL and VAF files from CCE are saved.
i was not refering to a splitted file but to 2 identical files.
in the output settings of CCE in Multipass VBR one can define two M2V files.
maybe you are defining only one of the them.
i am going to give the 5th pass a try, i feel like the 4.32GB version can still get a little better, and i will take this change to use the alternate scaning to see the difference because i am still getting some blurry and blocky parts though this version is better than Instant Copy ( very Less blocky ).

the forums say that the difference between the 2.50 and the 2.66 version are not worth the upgrade but i wll think about that.

but BJ_M, what other factors do you refer about that can make more of a difference? can you give some examples?

thanks in advance

[BJ_M]

later versions work better on interlaced material (your dvd was shot on digital betacam which is interlaced top field first)

i think procoder does as good or better on such material on mastering mode (not very quick though) ..

if its blocky at that high a bit rate -- some settings may be way out of wack -- there should not be any even at 2 pass .. the higher number of passes are more usefull at lower bitrates .. also try another matrix ... make sure you use a interlaced matrix type though - big error to use progressive matrix table with interlaced source ...

use this one:

 8,13,13,17,17,21,21,28,
13,13,17,17,19,21,23,30,
13,17,19,19,21,23,28,34,
13,17,19,19,21,23,28,48,
17,19,19,19,23,28,34,48,
19,19,23,25,28,32,34,48,
19,21,23,25,28,32,34,48,
21,21,25,25,28,32,34,48

 8,11,11,15,15,17,17,24,
11,11,15,15,17,17,21,24,
13,15,15,17,21,21,26,34,
13,17,15,17,21,21,26,48,
17,21,21,23,21,30,34,48,
17,21,21,23,28,30,34,48,
19,19,25,26,28,30,48,48,
19,19,25,26,28,30,48,48

[digital_EUR]

well i wasnt thinking of giving up the hope on true encoding versus transcoding ( instant copy ) but this is getting a little confusing with so many variables envolved.

my parameters:

Video
------
intra dc precision 9
luminance level 16-235
add sequence end code
zip zag scanning order
multipass VBR 4 pass
avg 4700
( do you think this is a high bit rate?
this is 37% lower than the original! )
min 500
max 8500

all others unchecked

Gop
-------
Close all GOP´s
M=3
N/M=4
SEQ=1

Quality Settings
------------------
Image Quality Priority = 4
no anti noise filter

DVD2AVI says it is INTERLACED

BITRATE VIEWER says:

Nom. bitrate: 8400000 Bit/Sec
DCT precision: 9
Pic. structure: Frame
Field topfirst: Yes
( this is why i unchecked upper field first as said in all the guides)
DCT type: Field
Quantscale: Nonlinear
Scan type: ZigZag
( this is why i checked Zig Zag scanning )
Frame type: Interlaced
Scene change detection: NOT FOUND
Variable GOP pattern: NOT FOUND


Matrixs:

Intra:
8 16 19 22 26 27 29 34
16 16 22 24 27 29 34 37
19 22 26 27 29 34 34 38
22 22 26 27 29 34 37 40
22 26 27 29 32 35 40 48
26 27 29 32 35 40 48 58
26 27 29 34 38 46 56 69
27 29 35 38 46 56 69 83

NonIntra:
16 17 18 19 20 21 22 23
17 18 19 20 21 22 23 24
18 19 20 21 22 23 24 25
19 20 21 22 23 24 26 27
20 21 22 23 25 26 27 28
21 22 23 24 26 27 28 30
22 23 24 26 27 28 30 31
23 24 25 27 28 30 31 33


what is wrong in all of this?

thanks again

[BJ_M]

change the following and see notes:

Video
------
intra dc precision 9
luminance level 0-255 --< (already clamped)
add sequence end code
alternate scanning order <----
multipass VBR 4 pass
avg 4700
( do you think this is a high bit rate? - (((fine bitrate))))
this is 37% lower than the original! )
min 500
max 8500

all others unchecked

Gop
-------
Close all GOP´s (i would leave open if authoring program can accept - better quality)
M=3
N/M=4 <-- change to 5 for that bit rate
SEQ=1

Quality Settings
------------------
Image Quality Priority = 4
no anti noise filter

DVD2AVI says it is INTERLACED

BITRATE VIEWER says:

Nom. bitrate: 8400000 Bit/Sec
DCT precision: 9
Pic. structure: Frame
Field topfirst: Yes <<<<-- ((( CCE will ALWAYS flag top field first and the setting for this is a bit tricky .. uncheck is correct in your case but not always the case)))
( this is why i unchecked upper field first as said in all the guides)
DCT type: Field
Quantscale: Nonlinear
Scan type: ZigZag (see above)
( this is why i checked Zig Zag scanning )
Frame type: Interlaced
Scene change detection: NOT FOUND
Variable GOP pattern: NOT FOUND


Matrixs:

change to what i posted above

[digital_EUR]

Well the results with your matrize got a little away from the image style from the original, mostly in the objects borders.
Dont know if it was from the matrizes or from the alternate scanning.
So i went back to CCE standard matrizes and modified the following parameters:

Closed GOP - changed to Open GOP
( Maestro acepted it and the original was also Open GOP )

Average Bitrate climed to 4730
( it was 4700 and the final dvd got 4,35GB )

Minimum Bitrate droped from 500 to 0

Maximum Bitrate climed from 8500 to 9200

M/N got back to 4 ( the original GOP had 12 picture )

Passes - changed from 4 to 5

After almost 9 horas of analyze and encoding the result was very satisfactory. The majority of scenes that had problems with visible blocks got better, many even got equal to the original.
I can tell now by the fingers on one hand the number of scenes that still have a little bit of blocks.

I feel that with a few more passes maybe those last scenes can get perfect also.

I havent deleted the VAF file, so my question is:

Can i do another 5 pass and the encoding process will consider the previous 5 that i already made?
Maybe getting this way a 10 pass quality level divided in a 2 phases job?

Best regards and once again thanks in advance for any help.

P.S. this concert is very difficult endeed because of his original image noise that makes encoders waste bitrate with it but despite this i dont
think a noise filter would help.

[BJ_M]

Can i do another 5 pass and the encoding process will consider the previous 5 that i already made?

not if you deleted the vaf file

that many passes is a waste of time -- but if you want to do them ...

at that point you may as well adjust the bit rate by hand (which ive done) .. i don't think you are understanding what those passes are doing -- they are re-configuring the bit rate per GOP , not some kind of video processing..

[BJ_M]

the bIGGEST - by FAR - improvement you can make, is to improve the source material BEFORE encoding ...

that means remove chroma shift, noise , single pixel errors, color balance, block noise from bad prev. compression and image stablization , etc etc ... ..

[digital_EUR]

Well, as adjusting bitrate manually in advance mode would take me dedicated hours to that job, that´s not a viable solution to me.
i would instead bet in another 5 passes and let the encoder do the same thing by his own. And this is because for me leaving the pc at home working during 9 hours while i go to work it´s not time waste for me directly.
But about the improvement of the source quality how can i do that?
I am frameserving to CCE from the original using DVD2AVI and then VFAPI but i would alert to the fact that the original already has pixelation problems and i dont think i can solve that part on my own but give me your opinion anyway.

Ahh, i did not delete the VAF file so based on your answer if i would do another 5 pass in the end it will be equivalent to 10 pass, right?

Best regards