QUOTE (M @ May 8 2003 - 03:25 AM)
@ Roberto: Do you know something about WavPack 4 that we don't? I remember David and Josh making a few obscure comments about the possibility of combining their efforts into a new encoder, but I hadn't heard anything recently. Any news would be apprecieated.
No, I don't know that much about WavPack4.
But, AFAIK, David and Josh aren't working together.
All I remember was Josh commenting he would look into maybe adapting WavPack 3 as an optional lossless mode for FLAC (no compability with current FLAC decoders then, of course). But he wasn't very sure about it due to patent concerns. Noone ever did a patent search on WavPack algorithms, I think.
Most of them are old technology, according to David, so it's not a big issue anyway.
And I don't remember David commenting anything about working with Josh.
What I know about WavPack4:
-It'll be much more flexible, what means more sampling rates, more bit depths, and multichannel.
-It'll be block-based, what will address the following issues (and probably others): better seekability, streamability and error robustness.
-Official tagging will be APEv2

-Compression might increase, but David isn't willing to sacrificate encoding time for that. Advanced compression modes would probably require neural networks, like the ones Monkey's Audio and LA use. But, in order to be efficient, neural networks should be coded in assembly - and assembly means a nightmare porting the code to other platforms.
WavPack is 100% ANSI C code, so it should be very easy to port.
There are other ideas, but they are still theoretical and implementing them might prove useless or unfeasible, so I'd better not speak about them now.
Regards;
Roberto.
Edit: I also don't know anything about progress/release date/etc.
@superdumbprob: it's more than immediately and less than some time.

Seriously, I'd rather not speculate. David wanted to have something ready soon, but N things might happen that would delay the project.
@layer3maniac: ???