Best CQM for 1600-1700 bitrate encodes

[rjc7394]

I use Shark High bitrate and Sixofnine matrixes for this bitrate area for video that is above average-very good. Just wanted to know if there any others I should look at. I have most of them. I haven't had as good luck with HVS Better and Best CQM than the ones above.

I use Jaw1CD and MPG up to 1650 kbps for video that is average-above average.

[BigDid]

I use Shark High bitrate and Sixofnine matrixes for this bitrate area for video that is above average-very good. Just wanted to know if there any others I should look at...
Hi,

For that kind of bitrate and full resolution, you could give a try to EqmV2 and the V3HR modified by Heini, more here: http://forum.doom9.org/showthread.php?p=854722#post854722

Did

[Blue_MiSfit]

All depends on the source of course, but EQM-V3HR would probably be a place to start.

[Audionut]

Take the word "best" out of your title as per rule rule 12 before a mod comes in to do their job.

[Teegedeck]

I use Shark High bitrate and Sixofnine matrixes for this bitrate area for video that is above average-very good. Just wanted to know if there any others I should look at. I have most of them. I haven't had as good luck with HVS Better and Best CQM than the ones above.

I use Jaw1CD and MPG up to 1650 kbps for video that is average-above average.It is absolutely impossible to tell which matrix to use for which bitrate because bitrates are meaningless.

http://forum.doom9.org/showthread.php?p=839185#post839185
http://forum.doom9.org/showthread.php?p=853208#post853208 !!!

And - right - there is no 'best', we have a rule about that.

[Blue_MiSfit]

Yeah, bitrate is essentially meaningless, but still. You gotta start somewhere :)

[rjc7394]

They can't be meaningless because I know if I use a low bitrate I'm going to end up with crappy output.

You're talking about variables like motion that make it very hard to choose a proper bitrate or variables that are more dependent on other XviD parameters?

[Teegedeck]

Certainly both. For the subject of XviD settings and their influence on bitrate, there certainly is an influence there: Especially VHQ mode has a strong influence on filesize. At the same quantizer you reach about the same filesize if you either use VHQ=4 with the EQMv3 HR matrix or if you use VHQ=1 with the EQMv3 LR matrix.

Secondly, the source is more decisive. On the one hand you have the controllable parameter 'resolution' - you can encode to the same filesize with EQMv3 ULR and SixOfNine if you reduce the resolution for the SixOfNine-encode. On the other hand you have the less controllable influence of video content. Bad compressibility can be caused by motion, but also by contrast, film-grain and simply - very important and often overlooked - by detail (among other factors, DVDs are also encoded using various custom quantization matrices!).

Encode some movies at the same quantizer with the same matrix and you'll get radically different results. When making size predictions with SixOfNine at quantizer=3 I get results ranging from about 1 to 3.5 GB; most movies give results roughly between 1.4 and 2.4 GB. This exemplifies that you only know whether you can use a specific CQM AFTER making a size-prediction. Add to that the often small differences in compression between CQMs (see the HR vs LR example) and you see why I don't recommend a CQM based on "1600-1700 bitrate".

Edit: Corrected obvious mix-up and typos.