QMatOp for DVD-RB : QMat Optimization

[jdobbs]

The 0.32b is to be released very soon... the answer is in the quote

[Carpo]

SAPSTAR,
It appears I used v.31b, downloaded from your http://www.autoqmatenc.com/
website. Do I understand that your new version is v.32? I didn't see it there on the page, so I assumed the version there was the latest release.

there is a 0.32 release in the beta testers area, the link to which is at the bottom of autoqmatencs website, besure to get Beta Version 0.32.0.3 first as it contains new files for use on dual core systems then get Beta Version 0.32.0.9 - or you could just wait till 0.32 final is released :)

[rendez2k]

Has anybody done any comparative tests to other encoders with the new versions? What are peoples experience of the quality? I keep meaning to test myself!

[archaeo]

there is a 0.32 release in the beta testers area, the link to which is at the bottom of autoqmatencs website, besure to get Beta Version 0.32.0.3 first as it contains new files for use on dual core systems then get Beta Version 0.32.0.9 - or you could just wait till 0.32 final is released :)

thanx Carpo

I have a dual processor, so I'm looking forward to trying this version.

[Fishman0919]

Has anybody done any comparative tests to other encoders with the new versions? What are peoples experience of the quality? I keep meaning to test myself!

Have done a lot of testing in the passed with AQE and most of the time it's on-par with HC and CCE... my testing has been limited recently because of new job, lots-o-hours and building a new home... but I did just do a test of "The Bourne Identity" with CCE Basic 2.70.01.08, HC .16a and AQE 0.32.0.9... hard movie to encode... very dark with lot of detail...all three did great work but I liked AQE because dropping the extra 25% and they still looked outstanding with AQE where with HC and CCE Basic the extras wasn't as good, the the dark scenes also looked great with AQE w/QMatOp and the dark engine SAPSTAR has.

[jdobbs]

Are the encode speeds comparable?

[rendez2k]

Jdobbs, any plans to add native support for AQE, I mean as a proper selectable encoder with options rather than having to use the CCE option?

[jdobbs]

I'm not sure it makes a difference... since it works anyway.

[rendez2k]

I'm not sure it makes a difference... since it works anyway.

Guess it would be nice to include it with DVDRB (as HC etc) - can't have too many enocders! I'd also like to keep seperate settings for both CCE and AQE.

[archaeo]

I ran some quick tests to visually compare the results of an encode using SAPSTAR’s AutoQMatENC (v.32.0.9) to an encode using CCE SP 2.70.1.5 (Trial). To set up this comparison, I chose a small test clip (85 MB) from ‘Lawrence of Arabia’, a long movie (227 min) with a compression factor of 56% and an average bitrate of 2200 Mbs when placed on a DVD-5. For this test I used RB-Opt to make sure the average bitrate of the encoded clips was the same average I would receive if I encoded the whole movie onto a DVD-5 disc.

So here's the visual comparison (shown @200%) between an encoded result of AutoQMatEnc (w/self generated matrix) and that of CCE SP (w/standard MPEG matrix):


5398
(see attached)


Visually, the results are pretty distinct in the side by side comparison. The artifacts in the CCE SP encode (using MPEG matrix) are more noticeable (but do hold some detail), while the image quality of the AQE generated matrix shows less artifacts and looks smoother, but perhaps has lost a little detail as well. That's to be expected with a compression of 56% and such a low bitrate. I think the picture quality of the AQE image looks very good, better in my eyes than the standard MPEG encoded clip generated by CCE.

Overall, I found that the program worked very well, no glitches. In addition, I was able to utilize the HT feature in AQE since I run dual processors (PIII xeon 1gHz).For the record it took 9 minutes for AQE vs 4 minutes for CCE.


In case you’re interested, here’s the matrix from the original clip that I extracted thru Rockas’ RME
(turned out to be Fox Home Entertainment matrix):

8, 8, 9, 11, 13, 13, 14, 17
8, 8, 11, 12, 13, 14, 17, 18
9, 11, 13, 13, 14, 17, 17, 16
11, 11, 13, 13, 13, 17, 18, 20
11, 13, 13, 13, 16, 17, 20, 24
13, 13, 13, 16, 17, 20, 24, 29
13, 12, 13, 17, 19, 23, 28, 34
12, 13, 17, 19, 23, 28, 34, 41

8, 8, 8, 9, 9, 9, 9, 10
8, 8, 9, 9, 9, 9, 10, 10
8, 9, 9, 9, 9, 10, 10, 10
9, 9, 9, 9, 10, 10, 10, 10
9, 9, 9, 10, 10, 10, 10, 11
9, 9, 10, 10, 10, 10, 11, 11
9, 10, 10, 10, 10, 11, 11, 11
10, 10, 10, 10, 11, 11, 11, 11


Here’s the matrix generated by AutoQMatEnc:

8, 14, 17, 23, 42, 69,110,251
10, 19, 17, 22, 42,100,239,255
11, 17, 34, 56,103,209,255,255
19, 21, 47,152,216,240,255,255
36, 33,113,220,255,255,255,255
72,100,216,255,255,255,255,255
174,230,255,255,255,255,255,255
185,251,255,255,255,255,255,255

18, 23, 24, 31, 41, 57,104,149
23, 29, 27, 30, 35, 47, 68,115
24, 26, 38, 45, 50, 63,105,141
31, 30, 43, 61, 92,122,145,185
38, 35, 49, 90,135,169,214,229
52, 45, 58,115,161,220,238,251
81, 58, 93,141,207,239,253,255
127,104,129,169,224,250,255,255


As you can see these are some very high values, thus explaining the loss of detail in the AQE image.

[m1ckran]

I'm not sure it makes a difference... since it works anyway.


The difference is that if you don't include the options in DVD-RB, it looks as though you don't really endorse the product. People might think it is merely an unoffical deviation from the standard system.

I'm really interested in this, though I've not tried it. I'm waiting for you to approve its use by incorporating it into DVD-RB.

Thanks for fabulous software. :)

[Fishman0919]

Are the encode speeds comparable?


No, CCE Basic was about 100 mins, HC was about 140'ish and AQE was about 250 mins

[archaeo]

No, CCE Basic was about 100 mins, HC was about 140'ish and AQE was about 250 mins


That's in line with the times I got from my short clip tests - about twice as long as CCE.

[Skunk]

Visually, the results are pretty distinct in the side by side comparison. The artifacts in the CCE SP encode (using MPEG matrix) are more noticeable (but do hold some detail), while the image quality of the AQE generated matrix shows less artifacts and looks smoother, but perhaps has lost a little detail as well. That's to be expected with a compression of 56% and such a low bitrate. I think the picture quality of the AQE image looks very good, better in my eyes than the standard MPEG encoded clip generated by CCE.

As you can see these are some very high values, thus explaining the loss of detail in the AQE image.

OK. Between thess statements I am a bit confused. You explain that AQE has some extra loss of detail, but you think the quality is better?

I am a total matrix noob, but those ones generated by AQE are way higher than any others I've seen around. Is that the reason there is soo much loss?

[archaeo]

OK. Between thess statements I am a bit confused. You explain that AQE has some extra loss of detail, but you think the quality is better?

I am a total matrix noob, but those ones generated by AQE are way higher than any others I've seen around. Is that the reason there is soo much loss?


This is completely subjective, but yes, in this instance I'd rather see some smoothing in the image and loss of some detail than heavy artifacting around the edges.

I agree, the AQE generated matrix has some really high values, similar to what you see in Bulletproof's heavy compression matrix. It's a lot more aggressive than what I would have picked as a custom matrix. All I can say is that between these two images, I prefer the one generated by the AQE matrix. Perhaps some further tests with other custom matrices would offer more comparisons.

[Skunk]

I think the custom matrix is an awesome feature of AutoQMatEnc. It could save a ton of time on picking through matices for every scene.
I do agree (now that I can actually see the comparison attachments) that AQE looks noticeably better overall. I also see what you mean about loss of detail. Its like the CCE one is over-sharp.

Wonder what that shot would look like if you ran QMatOp on that project and THEN ran it through CCE? Do you think the CCE encoder would do a better job if it had the same 'edge'?

I'm not asking which encoder is better, just wondering how much of an edge AQE actually gets by customizing matices.

[jptheripper]

can aqe handle avisynth filters? such as resizing, sharpening, etc.. just curious

[Fishman0919]

....I haven't had any pb's with AQE and filters...

[Insane2986]

Anyone know why my Qmatop stopped working? I click the green checkmark and do the standard "delete WAITDVDRB file, close DVDRInterface.exe in system tray." The progress bar pops up but then closes almost immidiately :(

[Insane2986]

bump :stupid:

[Carpo]

seeing as the author stoped work on this ages ago its doubtfull you'll get a solution to your issue - unless someone else has had the same issue and found a solution

[moviefan]

When starting QMatOp the progress bar appears, but I get errors in the command line. A very confusing thing is that the resolution is totally wrong (496x56), although the avs script is correct (tested). This is what QMatOp returns in the command line:

Execing cmd.exe /C .\DVDRBInterface.exe WAITPREPARE "C:\Programme\DVD Rebuilder
\QMatOp\.\WAITDVDRB"
Execing cmd.exe /C del /Q .\frames\Intra*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Intra*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Finished
Execing cmd.exe /C del .\temp.avs
Pass 1/1 frame 3/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 129/240, 427.61fps, ETA: 00:00:00.264, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.594 (0.00fps).
Execing cmd.exe /C del /Q .\frames\Inter*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Inter*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Pass 1/1 frame 8/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 136/240, 446.67fps, ETA: 00:00:00.237, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.601 (0.00fps).
Finished
Execing cmd.exe /C del .\temp.avs
IntraMatrix : W01000000001001
8 16 16 16 17 18 21 24
16 16 16 16 17 19 22 25
16 16 17 18 20 22 25 29
16 16 18 21 24 27 31 36
17 17 20 24 30 35 41 47
18 19 22 27 35 44 54 65
21 22 25 31 41 54 70 88
24 25 29 36 47 65 88 115
InterMatrix : W01000000001001
18 18 18 18 19 21 23 27
18 18 18 18 19 21 24 29
18 18 19 20 22 24 28 32
18 18 20 24 27 30 35 40
19 19 22 27 33 39 46 53
21 21 24 30 39 50 61 73
23 24 28 35 46 61 79 98
27 29 32 40 53 73 98 127
Execing cmd.exe /C del /Q .\frames\Intra*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Intra*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Pass 1/1 frame 3/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 128/240, 420.00fps, ETA: 00:00:00.271, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.601 (0.00fps).
Finished
Execing cmd.exe /C del .\temp.avs
Execing cmd.exe /C del /Q .\frames\Inter*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Inter*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Pass 1/1 frame 3/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 127/240, 420.88fps, ETA: 00:00:00.273, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.597 (0.00fps).
Finished
Execing cmd.exe /C del .\temp.avs
IntraMatrix : W01000100002001
8 16 16 16 17 18 21 24
16 16 16 16 17 19 22 25
16 16 17 18 20 22 25 29
16 16 18 21 24 27 31 36
17 17 20 24 30 35 41 47
18 19 22 27 35 44 54 65
21 22 25 31 41 54 70 88
24 25 29 36 47 65 88 115
InterMatrix : W01000100002001
18 18 18 18 19 21 23 27
18 18 18 18 19 21 24 29
18 18 19 20 22 24 28 32
18 18 20 24 27 30 35 40
19 19 22 27 33 39 46 53
21 21 24 30 39 50 61 73
23 24 28 35 46 61 79 98
27 29 32 40 53 73 98 127
Execing cmd.exe /C del /Q .\frames\Intra*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Intra*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Pass 1/1 frame 3/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 130/240, 426.67fps, ETA: 00:00:00.262, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.601 (0.00fps).
Finished
Execing cmd.exe /C del .\temp.avs
Execing cmd.exe /C del /Q .\frames\Inter*.ppm
C:\Programme\DVD Rebuilder\QMatOp\frames\Inter*.ppm konnte nicht gef
unden werden
Execing cmd.exe /C .\avs2avi .\temp.avs .\avs2avi.log -o l -c MSVC

AVS2AVI v1.31 (c) 2002,2003 Christophe PARIS (christophe.paris@free.fr)
Contributions by: int21h
Additional code (c) 2002,2003 David Leatherdale (dave@leatherdale.net)

Source:
* Filename: .\temp.avs
* Bit depth: 32
* FOURCC: none (RGB)
* Frames: 240
* Resolution: 496x56
Compressor:
* Name: Microsoft Video 1
* FOURCC: MSVC

Pass 1/1 frame 3/240, 0.00fps, ETA: , size: 0.00MB
Pass 1/1 frame 131/240, 422.95fps, ETA: 00:00:00.262, size: 0.00MB

* Frames: 0
* Key frames: 240
* Size: 0.000MB
Pass 1/1, compression time 00:00:00.600 (0.00fps).
Finished
Execing cmd.exe /C del .\temp.avs
Execing cmd.exe /C del /Q .\intra\*.* > NUL
Execing cmd.exe /C del /Q .\inter\*.* > NUL
Execing cmd.exe /C del /Q .\pics\*.* > NUL
Execing cmd.exe /C .\Calculate.bat 17 1
Exception in thread "main" java.lang.ArithmeticException: / by zero
at QMatOp.AverageMatrix(QMatOp.java:589)
at QMatOp.ECLCalculate(QMatOp.java:347)
at QMatOp.main(QMatOp.java:68)

Can anyone help?