[X4U] Most advanced Video Codec WMV?

[B. Kuestner]

One should note that there are different H.264 encoders out there and  
they do not all do their job equally well.

Basically, like QuickTime is just a container, so H.264 is just a  
codec, i. e. the format description allowing some techniques to  
compress video. But the encoder is the hard part: Because it must  
analyze the video material and then find the best method which of  
these allowed techniques to use in which way.

For instance, if you see an object moving through the screen, will  
the encoder realize what is happening and apply the best combination  
of H.264-techniques for the job?

So it is obvious that encoding quality (measured in small size vs.  
good quality) depends a lot on how much time the encoder spends on  
analyzing the raw material. Some encoders have settings to tell them  
how much effort they should spend on the analysis.

By the way, you also have that for MP3: You can tell iTunes to encode  
MP3 and invest more or less effort in the task. So setting it to 128  
kb/s VBR will give you different results if you tell iTunes to be  
quick about it or analyze the audio material much longer to get the  
best result. But even then the LAME encoder still does better at the  
same bitrate than iTunes' built-in encoder as far as MP3 is  
concerned, even if you put iTunes to the highest encoding effort/ 
quality. Some people also say that songs from the iTMS sound better  
than if you encode your own CDs at the same settings (AAC 128 kb/s).

Back to video: The leader of the pack when it comes to H.264 seems to  
be Nero these days. Their encoder seems to deliver much better  
quality/size at the same settings than say the QT encoder which is  
about average as I read.

But they are all just at the beginning of the learning curve for H. 
264. MPEG-2 encoders at the beginning were slow and the results  
relatively weak. Now at the end of the MPEG-2 lifecycle the  
developers have really learnt to make the most of the MPEG-2  
techniques, they employ efficient strategies in their encoders to  
analyze video material and pick the best combination of MPEG-2  
techniques. For H.264 that is much more difficult, because they have  
a vastly larger toolset at hand, and they have to decide under which  
condition which compression feature is more preferable over the other.

Björn