[MEncoder-users] Tuning hqdn3d

[RC]

On Wed, 13 May 2009 19:54:45 +0200
Nicolas George <nicolas.george at normalesup.org> wrote:

> The question is not pointless, but you fail to see my point, which is
> this: for any file size, the best overall quality will will be
> achieved when the local quality is constant.

Both CRF and 2-pass will do everything they can to acheive consistent
quality, from beginning to end.  You're making no distinction.

> And "appropriate" is precisely determined as "that will give the best
> visual quality".

That only applies to the relative distribution.  Not the absolute
quality level, which will be determined as "as high as possible with the
given number of bits".  

> On other words: if the rate control algorithm has a few spare bits to
> put either in scene A or in scene B, it needs to know which one
> already looks better, to gives them to the other.

It does, but given enough of a fudge factor, thanks to a bitrate, modest
errors will not be visible.

> - The best overall encoding should have a roughly constant subjective
>   quality.

There's no basis for this.  Consistently low quality is certainly not
better than alternating between low and high quality.

> - Constant quality encoding tries to provide it directly, target
> bitrate
>   encoding tries to achieve it with complex algorithms.

Bitrate encoding tries to provide the best quality for a given bitrate. 
It does not try to acheive a certain quality level, and throw away
available bits if it thinks it has.

> - Someone who do not care about fitting his video exactly on a CD or
> DVD wants to choose a quality level, not a target bitrate. 

In theory, yes, but CRF is simply a different equation.  It is not
"quality". 

> Finding the bitrate to achieve the desired quality is hard.

No.

> Conclusion: constant quality is the best choice for someone who do not
> care about the exact size of his videos.

Not at all.

> - Multi-pass encoding could be beneficial for global optimizations of
> the encoding, but current implementations of the codecs only use the
>   global information for rate control.

Not true.

> Conclusion: multi-pass encoding is currently useless for someone who
> do not care about the exact size of his videos.

Repeating it won't make it any more true.


-- 
Don't trust me! I'm wrong!