How to choose a quantization matrix in Hcencoder?

[katieburchett]

I was wondering what the difference is between the different matrices in Hcencoder and how I can figure out which one to use? There is a lot of motion in my footage but there are a few scenes with little motion also and obviously I want to get the maximum quality I can. (I've been using 'manono' so far but I have heard good things about 'fox1'?) Thanks

[manono]

They don't make a whole lot of difference, but I, too, try to match the quantization matrix with the material.

It's kind of a trade-off. You want to use the best quality matrix you can without sending the resulting average quant so high that you begin to see artifacts such as smearing or blocks.

In general, the lower the numbers in the matrix (particularly in the upper left) the 'better quality' it is. But it doesn't follow that you can or should use the very best matrix (Fox1). I don't have HCEnc installed on this computer so I can't be more precise as I'm working from memory.

You make an educated guess which to use. The one with these numbers is my usual 'go-to' choice, the one I use much of the time:

8 8 8 9 11 13 14 17
8 8 9 11 13 13 14 17
8 8 11 12 13 14 17 94
9 11 13 13 14 17 17 94
11 11 13 13 14 17 94 94
13 13 14 16 17 20 94 94
13 13 14 17 94 94 94 94
13 14 17 94 94 94 94 94

12 12 13 14 15 16 22 26
12 13 14 15 16 22 26 32
13 14 15 16 22 26 32 41
14 15 16 22 26 32 41 53
15 16 22 26 32 41 53 94
16 22 26 32 41 53 70 94
22 26 32 41 53 70 94 94
26 32 41 53 94 94 94 94

I forget what it's called in HCEnc but think it's one of the three 'manono' matrices. When done the encoding run the MPV/M2V through Tecoltd Bitrate Viewer and check the average Q-level. If it's up around 6-7 or higher, you might consider using a lower quality matrix and doing the encode over again. If down around 3 or so, you might want to switch to a higher quality matrix. I make my DVDs of films for DVD5s and the one above is the one I wind up using about half the time or more. If you're encoding interlaced material the matrix should usually be a lower quality one. If you're reencoding films for DVD9s, then most of the time you can use the Fox1 matrix. There are a lot of factors involved and it's impossible to give any hard and fast rules, particularly in advance of doing the actual encoding. But after the encoding the Q-level (quality scale) will provide the information you need to help decide if you should reencode or not. It's not worth it, in my opinion, to make a compression test for this, not when the encoding is so fast when using a decent modern computer.

[katieburchett]

When done the encoding run the MPV/M2V through Tecoltd Bitrate Viewer and check the average Q-level. If it's up around 6-7 or higher, you might consider using a lower quality matrix and doing the encode over again. If down around 3 or so, you might want to switch to a higher quality matrix. I make my DVDs of films for DVD5s and the one above is the one I wind up using about half the time or more. If you're encoding interlaced material the matrix should usually be a lower quality one. If you're reencoding films for DVD9s, then most of the time you can use the Fox1 matrix. There are a lot of factors involved and it's impossible to give any hard and fast rules, particularly in advance of doing the actual encoding. But after the encoding the Q-level (quality scale) will provide the information you need to help decide if you should reencode or not. It's not worth it, in my opinion, to make a compression test for this, not when the encoding is so fast when using a decent modern computer.

Thanks for all that information! Tecoltd Bitrate Viewer doesn't run on my PC for some reason, 'it says this app cannot run on your PC', it won't open in compatibility mode either. I am running windows 8.1 64bit. Is there another way to determine my average Q-Level?

[manono]

Is there another way to determine my average Q-Level?

None that I know of. CCE gives a Q-Level as it's encoding, but you're using HCEnc so that won't do you much good. If you or someone else comes up with one, know that the Tecoltd Bitrate Viewer has its own scale which won't be the same as that used for a different program. For example, you can run single passes for a defined quality level using HCEnc, but the numbers aren't the same as Tecoltd's ones.

You could try first running single-pass quality-based encodes using HCEnc and come up with your own guidelines, followed by two-pass encodes for whatever size you want, if necessary adjusting the matrix. The idea here is to find an HCEnc quality level and matrix with which you're satisfied for the size you want for your particular movie, and then to do the full two-passes to zero in on the final size more precisely. It might take several tries until you get the hang of it but after a very short period you'll be doing one single-pass test, make the adjustments, and then do the final two-pass encode.

The HCEnc Q-scale numbers are less than the corresponding Tecoltd ones, so a Tecoltd 7 might be a 5 for HCEnc. I don't know for sure, though, having never tested these myself.