H.264 Doubts

[Dave1024]

Let me ask few doubts relate to H.264 concepts. Please help me to find out a good reason for the following.

1. What is the advantage of using DCT in H.264 ?

2. What is the meaning of AC and DC coefficients ?

3. For High Profile what are the extra feature than main profile?

4. what is the use of calculating colocated blocks for B frames


:thanks:
Dave

[LoRd_MuldeR]

1. What is the advantage of using DCT in H.264 ?

Advantage compared to what?

2. What is the meaning of AC and DC coefficients ?

See:
http://en.wikipedia.org/wiki/JPEG#Discrete_cosine_transform

3. For High Profile what are the extra feature than main profile?

See:
http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC#Profiles

[Dave1024]

I want to know what is the reason for using DCT instead of FFT?

Thanks for mentioning the reference spots

Dave

[Mug Funky]

DCT is fast, integer-izable (yay! new word!), and well established (meaning there's tons of hardware for it and loads of pre-existing very fast code and concepts).

DFT would deliver a similar result at a higher complexity.

FFT i'm not really sure if it would have any advantages over DCT at all (though FFT3dfilter is a very good denoiser)

[pandy]

DCT is special case for DFT/FFT where FFT is special case of the DFT.
DCT when compared to other DxT gives "better energy focusing/concentration" - search for comparison DCT vs DST
H.264 doesn't use DCT per se but special "like DCT" transformation (integer implementation, fully reversible/lossless)

http://en.wikipedia.org/wiki/Discrete_cosine_transform
To be honest all Your question are well answered in many papers and articles - easy to find with Google help - try!