Matrix Basics. What are others doing?

[reefskinner]

I have been using Rebuilder for a while now, but haven't messed with the Matrices much. I am just wondering how others are going at it. Are people setting it up to use multiple matrices in the setup options and just leaving them as a sort of default? i.e. Low = x matrix, very low = y, extras = z etc.
this seems like a good way to go, but was wondering if others that have been using matrices for a while have an opinion on this.
Also, what are some of the better matrices to use? Just looking for some sort of starting point, kinda like a default matrix.

Thanks a bunch......

[Voodoochild]

Unusually in low compression movies let's say above 80% I use the same Matrix as the movie use and do a normal 2 pass vbr.

If Compression drops below 80% I do CCEAQM=1 (in the ini file) which means the encoder chooses the best Matrix for the segment, according to the matrix you choose. Meaning if you choose the CCE default matrix, it will be cut by half or quarter every time the encoder changed it. In this case I do 3 pass.
Again in this case I usually go for avamat6 matrix which is good from 3500-9000 br.

8 16 19 22 26 27 29 34
16 16 22 24 27 29 34 35
19 22 26 27 29 34 35 38
22 22 26 27 29 34 35 40
22 26 27 29 32 35 40 48
26 27 29 32 35 40 48 50
26 27 29 35 40 48 50 60
27 29 35 40 48 50 60 62

16 20 24 28 32 36 40 44
20 24 28 32 36 40 44 48
24 28 32 36 40 44 48 52
28 32 36 40 44 48 52 56
32 36 40 44 48 52 56 58
36 40 44 48 52 56 58 60
40 44 48 52 56 58 60 62
44 48 52 56 58 60 62 62


Today I tested low compression movie (93%) with 1 pass vbr, using the default matrix of cce since 1 pass vbr can't change matrix. Have to say got perfect results with half the time. Though I still need to experiment with this.

Hoped I helped a bit,

Elad

[techreactor]

its confusing and with so many matrices available, its difficult to decide/choose.

Use Autoqmat enc with Qmat enabled, it chooses or rather creates most suitable matrices based on source.

[kumi]

The differences between one matrix and another are so subtle that I don't pay much attention to that part of setting up an encode. Often I go by the numbers, using the 1-8 matrices from Rebuilder Matrix Editor.

[mutenroid]

Again in this case I usually go for avamat6 matrix which is good from 3500-9000 br.


I'm a little confused about choosing matrices. I have two aditional matrices (avamat6 & avamat7) in the defult matrix path of DVD-RB.

Before rebuild a movie i execute "prepare phase", this give to me an overall bitrate, if this value it's between 1800-3500 then i choose avamat7, else avamat6. But it's very difficult to decide when this value it's around 3500, like a range 3300..3700 :confused:

Are there other type of decisions that allow me to choose correctly between one matrix or other?

:helpful:

[rahzel]

get RME. it comes with several matrices and has a brief description of each. from what i understand, the lower the numbers, the more detail it retains. a matrix targeted for low compression will not be so good for a movie that requires high compression.

with CCE, i use the same matrix that was used for the movie if its 90%+. for 70%+ and if the bitrate is over 4000, ill use the default. anything below 3800-4000 bitrate, ill use avamat6 (good for higher compression/lower bitrate).

i'm starting to like procoder now where the default matrix does a good job for low bitrate movies. i haven't tried a high bitrate movie yet though.

you can extract matrices from DVD's with RME too. this is how i find out what matrix was used for the DVD. RME can also import matrices as well.

[G_M_C]

I just choose from the matrices that are available in the Matrix Editor. 4000 and up, i go for the "High-xxxx" matrices.
For bitrates lower than 4000, i choose accordingly ("Medium-xxxx").

In short: I always try to choose the highest possible setting, according to the available bitrate, allowing for some overhead.

Before choosing i do the prepare phase, and than look at the avg/min/max bitrates, and look at a few of the ECL's. I also use a very-very slight noise filtering in the filter-editor to help compressabillity. And now the downside: After choosing the matrix that i'm gonna use ... i have to do the prepare-phase all over again.

For filtering i usually put FFT3DGPU with a sigma(2,3,4) between 0.2 and 1.0. The filtering helps compressabillity, and often improves quality/clarity/edge-definement without the need for sharpening (and through sharpening the increase of noise, and the lowering of the compressabillity).

[deity67]

Do I need to change matrixs or do you just stick with the defaults no matter witch movie i doing.

[Digga]

Do I need to change matrixs or do you just stick with the defaults no matter witch movie i doing.well, the idea about a matrix is that you can pick the one that is best suited for a specific movie, so yes, changing the matrix (or considering which one to use) per movie is due.

just imagine you'd use a low bitrate matrix for all your movies. some would come out suboptimal.
the same principle also holds true for the default matrix to a certain degree. but this matrix has been chosen (I'd imagine) because it is the 'best' matrix for the widest range of movies.

the above is only a well educated guess, as I myself don't use different matrixes (damn, what's the plural of matrix?!).

of course if you don't have a clue what matrix or filter to chose you might as well mess up things if choosing the the 'wrong' one.

[G_M_C]

Do I need to change matrixs or do you just stick with the defaults no matter witch movie i doing.

If you read my posting above, you can get an idea on how things wor and what you have to do to get it to work (assumed you have DVD-RB pro).

I just choose from the matrices that are available in the Matrix Editor. 4000 and up, i go for the "High-xxxx" matrices.
For bitrates lower than 4000, i choose accordingly ("Medium-xxxx").

In short: I always try to choose the highest possible setting, according to the available bitrate, allowing for some overhead.

Before choosing i do the prepare phase, and than look at the avg/min/max bitrates, and look at a few of the ECL's. I also use a very-very slight noise filtering in the filter-editor to help compressabillity. And now the downside: After choosing the matrix that i'm gonna use ... i have to do the prepare-phase all over again.

For filtering i usually put FFT3DGPU with a sigma(2,3,4) between 0.2 and 1.0. The filtering helps compressabillity, and often improves quality/clarity/edge-definement without the need for sharpening (and through sharpening the increase of noise, and the lowering of the compressabillity).

[rahzel]

in general, matrix' don't make a huge difference imo. unless you're compressing quite a bit and the bitrate is low, the difference between using a matrix targeted for low bitrate/high compression vs and matrix targeted for high bitrate/low compression is noticeable.

some people may have really good eyes and can spot differences, but i personally can't, unless i take still images and zoom them.

[Sworm]

Hi all,
Avamat6 is good for low bitrate high compression, and the avamat7?

[techreactor]

avamat7 is meant for bitrates lower than 2000 whereas avamat6 does a good job of bitrates above 2000.

Use avamat7, see similar post here (http://forum.doom9.org/showthread.php?t=108313)