Quote:
Originally Posted by Prodater64
4 - Im sure that rds is getting 1% undersize in samples, not whole movie. But I can be wrong. When he do whole encoding, he will find a surprise, I think.
|
Nope, I'm getting undersize ~0.05% on predictions.
On real encodes I'm getting around 1% or even less.
But I don't know if I have already emphasized that I am running for a 2 movies in one DVD-5.
That means you work with ~4.000.000KB available for video whereas I work with ~1.900.000KB.
There can be the reason why you get bigger undersize figures than I do
.
Just as an example, I got one prediction with full encode of "The Matrix I":
Ideal Video Size=1.861.038,4KB
Encoded Video Size=1.840.011KB
This is just slightly over 1.1%
.
BUT my latest encode with "Back to the Future I" was again near the 1.8% which IMHO is a bad prediction.
So this means that the prediction behaviour is far from perfect.
Quote:
Originally Posted by Prodater64
5 - @rds: Please, if you find out the solution, let me know how? It would be great that I can include that maths in my tool. Thanks.
|
I sure will if I happen to find a way.
I just would need HC to give me a break
.
I gotta tell you: I have been more optimistic than I am at this moment.
Maybe there really isn't any way of achieving what we're seeking
.
Quote:
Originally Posted by Prodater64
Edited: Im thinking in doing a last 5% prediction and see where is going it. With this value, to do a correction of maxbitrate. Any idea.
|
I've been there already.
One of the 1st things I thought: maybe this doesn't work with 1%.
So I tried 2,3 and 5%.
Results were veeery similar.
So I went back to 1% which is faster and gives me the same figures.
Anyone else besides the 2 of us working on OPV prediction with HC?
Anyone??
Do I hear Karl?
Do I hear Andrej?
Phil?
Cheers