Patch by Robert Edele,  yartrebo <<at>> earthlink <<dot>> net

Originally committed as revision 5184 to svn://svn.ffmpeg.org/ffmpeg/trunk

This one sets up a snow.h and makes snow use the dsputil function pointer
framework to access the three functions that will be implemented in asm
in the other parts of the patchset.
Patch by Robert Edele < yartrebo AH earthlink POIS net>
Original thread:
Subject: [Ffmpeg-devel] [PATCH] Snow mmx+sse2 asm optimizations
Date: Sun, 05 Feb 2006 12:47:14 -0500

Originally committed as revision 5172 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 5164 to svn://svn.ffmpeg.org/ffmpeg/trunk

+15% speed, +.1 dB

Originally committed as revision 5158 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 5157 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 5145 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 5144 to svn://svn.ffmpeg.org/ffmpeg/trunk

patch by Dieter <freebsd at sopwith solgatos com>

Originally committed as revision 4913 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4891 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4876 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4875 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4868 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4864 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4842 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4841 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4839 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4824 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4821 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4819 to svn://svn.ffmpeg.org/ffmpeg/trunk

fix 4mv playback distortions

Originally committed as revision 4813 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4794 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4793 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4789 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4783 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4782 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4781 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4764 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4758 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4749 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4718 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4717 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4609 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4535 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4205 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4161 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4159 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4153 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4132 to svn://svn.ffmpeg.org/ffmpeg/trunk

about 10% lower bitrate for -qscale 32 (forman & some music video)
worst case bitrate increase <0.1% (lossless or low qscale)
and now the bad news, even though this just adds a single subtraction and an if() into the medium sized unpack_coeffs() loop and the if() will only be false once per unpac_coeff() call, gcc produces 50% slower code, i didnt look at the generated asm yet, not sure if i want to ...

Originally committed as revision 4131 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4130 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4129 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4128 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4127 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4126 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4119 to svn://svn.ffmpeg.org/ffmpeg/trunk

this is needed as the quantization stepsize for each subband is also in this precission and insignificant changes to the wavelet like scaling its coefficients slightly differently would lead to wildly variing PSNR and bitrate
note, a encoder could also simply choose to leave the least significant bits of the quantization parameters zero which would give the exact previous behaviour except a y very tiny number of bits in  the header

Originally committed as revision 4115 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4114 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4108 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4107 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 4104 to svn://svn.ffmpeg.org/ffmpeg/trunk