Can't make Auto Gordian Knot use the mpeg custom to fix distorted colors

[Talayero]

Hi,

A couple of months ago I fixed an xvid movie with distorted colors problem. The movie showed bits with distorsion in my Pioneer standalone and also in my Windows XP. But after using a MPEG custom matrix in Auto Gordian Knot I fixed the problem. I don't remember now if I used the MTK or the ESS matrix.
The thing is the movie was divided in two parts and I only fixed then the first part. Now I want to fix the second part and I don't find the way to repeat the process.

I must say I'm not familiarized with Auto Gordian Knot, I always use GordianKnot or simply Virtualdubmod to encode with Xvid, but anyway it doesn't seem to be a very complicated program. The thing is now when I encode that second part with Auto Gordian Knot using the matrices for the standalone support I don't get the problem fixed, because the movie is not really encoded with an mpeg custom matrix. Mpeg4 modifier says Quant type: H.263. I tried several times changing from MTK to ESS, or changing the file to different folders, or uninstalling and reinstalling the program, but in the end I always get a file with Quant type H.263.

On the contrary, the movie I fixed months ago with the same Auto Gordian Knot had Quant type: Mpeg Custom and matrix:

Custom intra matrix:
8 16 20 23 27 30 32 34
16 16 23 24 29 31 34 38
20 23 27 28 30 35 38 40
23 24 28 29 35 39 40 44
27 29 30 35 39 42 45 50
30 31 35 39 42 46 54 59
32 34 38 40 45 54 62 72
34 38 40 44 50 59 72 84

Custom inter matrix:
17 19 21 23 25 27 29 31
19 21 23 25 27 28 29 33
21 23 25 27 29 30 32 37
23 25 27 30 31 34 40 45
25 27 29 31 38 46 54 60
27 28 30 34 46 58 72 74
29 29 32 40 54 72 90 100
31 33 37 45 60 74 100 124

Now, only once, after installing the Auto GK in other computer, I managed to encode the second part of the movie with an mpeg custom matrix (MTK compatibility) but the result didn't fix the distorsion problem.
This is the matrix according to Mpeg4 modifier:

Custom intra matrix:
8 9 10 11 12 14 16 18
9 11 12 13 15 16 18 20
10 12 14 16 17 19 21 22
11 13 16 18 20 22 24 26
12 15 17 20 23 25 28 30
14 16 19 22 25 29 34 38
16 18 21 24 28 34 46 52
18 20 22 26 30 38 52 72

Custom inter matrix:
16 16 17 17 18 19 21 23
16 17 17 18 19 20 22 24
17 17 19 20 21 23 24 27
17 18 20 22 24 25 28 32
18 19 21 24 26 29 36 40
19 20 23 25 29 40 48 52
21 22 24 28 36 48 60 64
23 24 27 32 40 52 64 106

That means probably that the matrix I used months ago to fix the issue was the ESS, I don't know. The thing is now I can't use it again. I repeated a second time in this other computer, chosing ESS compatibility, and again I only got a H.263 quant type.

My question is: How could I encode that movie with that first matrix that fixed this distorsion problem months ago?

A second question: This matrices I wrote above are really the first one for the ESS compatibility and the second one for the MTK compatibility? It's just to rule out some mistake i could have made.

Thanks in advance

[manono]

Hi-

If using the ESS option, you'll get either the H.263 or MPEG matrix, depending on the results of the compress test. Both should play fine on the computer and in all standalones.


If using either the MTK option or no option at all, you'll get one of 2 custom quant matrices. Again, the one you get depends on the results of the compress test. Both should play fine on the computer, but may not play well in all standalones.

Your first one is a custom quant matrix, one developed by Sharktooth especially for AutoGK to be the low bitrate matrix. I forget its exact name, but it's very similar to his EQM v3ULR matrix. The other one is Sharktooth's EQM v3HR higher bitrate custom quant matrix. You would have gotten either of those 2 matrices by using either the MTK option, or no option at all. Using the ESS option, you wouldn't have gotten either of those 2.

For absolute compatability, if you don't know the chipset your DVD/MPEG-4 player uses, you should encode using the ESS chipset option.

I don't know how that helps you, since it's at odds with some of the things you said. You might consider reencoding manually and choosing the MPEG or H.263 matrix, and the results (if you don't also choose Q-PEL or GMC) should play in your standalone. Or use the DivX codec in AutoGK and be pretty much guaranteed that it'll play OK.

[Talayero]

Thanks for answering.

My standalone is a Pioneer dv-470 and if I'm not wrong it has a mediatek chipset.

There are at least two weird things I noticed related to this issue with my movie.

One is that the ugly distorsion occurs both in the standalone and in the PC. You say it should play fine in the computer, but it doesn't. It only plays fine after encoding with that first matrix. I've never read that there are movies that show distorsion colors in PC, only in standalones (Philips, most of the times, I think). Originally the problematic movie was an xvid with quant type H.264 Xvid0046. I reencoded with Auto Gordian Knot using that first matrix and the problem disappeared.

The second weird thing is that I managed finally to reencode the second part of the movie, that wasn't fixed yet, using that same first matrix (Jawor's1cd) and now it didn't work. The distortion is still there. There must be some factors beyond of my knowledge.


P.S. I also reencoded it with DivX without success.

[celtic_druid]

The second one above would be cqm_autogk_sharp. The first one is Jawor's 1CD matrix, which I think was used by older versions of AutoGK maybe? Both of the above matrices should play fine on PC and MTK based SAP's. No reason why any CQM should result in playback problems for PC based playback.

For ESS, I think H.263 would always be used since AutoGK would be using the Home Theatre profiles which disable anything else.

[manono]

The second one above would be cqm_autogk_sharp. The first one is Jawor's 1CD matrix

Looking more closely at the ones I thought they were, they're only similar, and not exact. Thanks for the correction. If I'm not mistaken, and if I remember correctly, though, the one called cqm_autogk_sharp was developed by Sharktooth in response to len0x saying that whatever was being used before (EQM v3HR?) produced too much mosquito noise. Like you say, the matrices have changed some over time.

I don't know though, Talayero, why you're having trouble after reencoding. As celtic_druid and I said, anything should play OK on a computer. And H.263 (and DivX) should play fine on any standalone. Perhaps there's something else going on here, but I have no idea what.