Why not quality-based?

[Mr. Conflict]

Why do the authors of
DivX4/5 so strongly discourage people from using quality-based mode, saying on their
official web site that "it applies the same compression regardless
complexity of the scene so it shouldn't be used for archiving
but rather for future editing". I don't get it. Is there another disadvantage besides never knowing the final file size?
Isn't it vbr mode for keeping constant quality? When I compress a complex scenes, say a fashion show with q=93% I get over 250Kb/s
data rate as opposed to dark film scenes when it gives me as low as 40Kb/s. p.s.I compress to DivX5 from MJPEG

Mr.C.
:)

[avih]

constant quality keeps the quality constant lways regardless of the scene complexity. however, for the human eye, usually, the quality is much more important in 'quiet' scenes, where you can notice the details, than in fast moving scenes, where you can't 'see' the details. that's why constant quality gives too much quality to scenes which the eye don't really care about quality. and on top of that, the filesize is not predictable.

if u can live with these, and like the looks of constant quality, just use it.

but if you prefere predictable file size, go for other methods.

cheers
avi