A short while back I ran across the Flake FLAC encoder. Doing some simple tests and minimal research, it seems pretty solid. For some tests, I ran a standard EAC+FLAC rip of a couple CDs using A) the standard FLAC 1.1.2 encoder, and B) the latest Flake with the -12 option -- I saved 1-2 megs per track!

The audio quality seems the same (this is only be ear at the moment mind you) and my preferred playback application (WinAMP) plays them just fine as well. I have yet to test with my Squeezebox.

Is Flake ready for every day use? 1-2 megs per track is a significant improvement in compression as long as the audio quality is the same (lossless is lossless, right?!) and as long as all the standard FLAC players can play the -12 produced FLAC files.

Any more information about this would be greatly appreciated! I'd love to see more pros/cons about this new encoder, thoughts on every day use, etc.

flake had problems with support, etc. so i would wait for a more stable release.

i think the 0-8 presets are fine, iirc


later

In past discussions about this question a few users mentioned that flake -12 can result in choppy playback on certain devices. That's due to the increased maximum prediction order of 32 with flake -11 and flake -12. The flake usage table on http://flake-enc.sourceforge.net/ shows that flake -10 is the optimal setting since its maximum prediction order is way more hardware friendly.

since Flake is a lossless encoder, the data stream should still be lossless. i personally use foobar2000 which has a nice BitCompare plugin which allows you to check if the stream is actually lossless.

i have heard of a bug that Josh Coalson found which happened due to Flake's wav reading code(Winamp uses a different thing IIRC) which gave different bitstreams. other than that, i don't think that i know of anything else wrong with it.

edit: bleh. people posted before me.

And now this bug must be fixed according to this post.

more info here:
http://www.hydrogenaudio.org/forums/index....st&p=463073