[XviD-devel] Adaptive quantisation

[Dirk Knop]

Hi,

Michael Militzer wrote:

>new lumi masking algorithm?? I didn't even know that!
>
Hm, I thought I read about this here on this list. But I can't verify 
this since the archive was built starting with the new website design.

>.....ok, I just did some research and have to admit that I'm a bit angry
>now: uManiac seems to have overwritten my luminance masking code even while
>stating that the new code is "unoptimized". There are such nice things like
>"#ifdefs" to add experimental new code! Why does noone ask before
>overwriting code?
>
Don't be angry, they were doing this with best intentions, I'm sure 
about that.

>Nonetheless I've tested this new code and it seems total bogus:
>
>my dogma test-clip, fixed quant 4:
>
>25.442.304 Bytes, without lumi-masking
>25.403.392 Bytes, with lumi-masking
>
That's the problem why the people keep bugging me over at doom9... :)

>Therefore I'll simply switch back to my "old" code. About the user
>adjustable values: There are pretty much values that are important and it's
>maybe no good idea to let the user adjust all of them. I'd suggest to allow
>the user to adjust how "aggressive" luminance masking should work (maybe a
>nice slider) and that we'll include this option in the upcoming API 3.0.
>
I could do a "test build" with an edit box which edits this threshold. 
Let's see first if people can cope with it, I think it's too hard for 
them to get that right and if they fail, not to blame the codec. (Like 
in the new doom9-comparison, I could just give him some hints how to 
setup the codec correctly, but he failed. Since I don't have those DVDs 
I can't deliver optimized values [well, I lent the matrix now to see if 
those scenes aren't tweakable. I'm confident XviD can cope better with 
that.]. Too bad that he didn't mention that the "new" curve treatment 
improves the overall quality of an xvid-encode by a noticable amount.)

>btw: thanks koepi for letting me know at all, elsewise I would have never
>noticed that the code had been changed...
>  
>
Hm... you're welcome. I'm looking forward to redo the Matrix with this 
new old code :)

Best regards,
Koepi