As was read all over the net, WA3 alpha has been released to let the general public be their free alpha testers. Perhaps it's time to start working on a WA3 MAD plugin so that when WA3 is finally stable, we can have a working version of MAD ready? I'm no developer but I've read that it's pretty easy to port 2.x plugins to the 3.x architecture.
The released pre-SDK doesn't look to be quite finished -- in particular, there doesn't seem to be any support yet for configuring plug-ins.
Something I read on a decoder test website post recently on the list.
"In conclusion, the MAD decoder has no major decoding bugs, and the only problem compared to other decoders is that it clips the ends off many files. It doesn't quite live up to its accuracy claims - it is better than other 16-bit decoders, but doesn't yet represent the best that can be achieved."
Room for improvements?
The end clipping issue can be addressed, as can some of the dithering issues David mentions elsewhere in that review. I plan to introduce a few different dithering options in the configuration soon so you can choose the one that sounds best.
-rob