HOW TO TEST SNOW
 ----------------

Snow is an experimental wavelet-based codec made by the ffmpeg developers,
and while it's still in heavy development, it's already giving very good
results.
Be very careful though, as the format of the bitstream produced might
change, don't rely on it to store videos that you value.
As such, you will need to add vstrict=-1 to you command line to let
MPlayer allow an encode.


OPTIONS OBEYED BY SNOW

 * vqscale=<0.01-255>
   Encoding quality, sane range 1-10, default: 0 (lossless).
   A given quality in snow needs a somewhat lower qscale than the same
   quality in mpeg4.
   Note that 0 may not be specified; if you want lossless encoding, you
   must leave out vqscale. (Yes, quantizer can be fractional.)

 * cmp, subcmp, mbcmp
   Set the comparison function, default: 0 (SAD).
   useful values = 0 (SAD), 1 (SSD), 11 (5/3 wavelet), 12 (9/7 wavelet).
   Experience show that SSD is the best most of the time, while SAD is
   slightly better the remainder of the time.
   You can add 256 to any of the options to enable chroma motion
   estimation for that comparison (e.g. mbcmp=257 for SSD with chroma),
   but it doesn't seem to help much for the moment.

 * pred=<0-2>
   Wavelet type. 0 (9/7 wavelet), 1 (5/3 wavelet), 2 (13/7 wavelet).
   9/7 is probably better for for lossy coding, and 5/3 for lossless.

 * last_pred=<0-3>
   Tries a few extra predicted motion vectors before doing EPZS search,
   default=0.
   This option has negligible effect on both speed and quality of snow,
   so just leave it off. (it does, however, help mpeg4.)

 * qpel
   Refines motion estimation, default: off.
   This setting always helps compressibility, but costs some cpu time
   both while encoding and decoding.

 * v4mv
   Allows smaller motion partitions, default: off.
   The current MB decision algorithm doesn't make very good use of this:
   it improves quality, but also increases bitrate. (and you could get
   more quality per bitrate by reducing quantizer instead.)


v4mv and the wavelet cmp functions are theoretically good, but in practice
won't work well until we get OBMC-aware motion estimation.

In short:
The best options in almost all cases are
vcodec=snow:vstrict=-1:vqscale=$N:pred=0:cmp=1:subcmp=1:mbcmp=1:qpel