fixed-point: Got rid of the three last float bits in the
[speexdsp.git] / libspeex / pseudofloat.h
2007-11-08 jmMoved the remaining of misc.h into arch.h
2007-10-28 jmInclude os_support.h in pseudofloat.h
2007-02-22 jmcomments
2006-11-10 jmpseudo-float divide operators should be more robust...
2006-11-10 jmDefined VSHR32 and using it instead of "if (shift>0...
2006-11-08 jmFixed potential overflow in gradient computation. Need...
2006-11-02 jmOops. Previous overflow fix had actually made things...
2006-11-02 jmfixed an overflow when the numerator is negative
2006-10-10 jmFixed a bug on 16-bit platforms (thanks Jim)
2006-08-31 jmIncrease learning rate for some mis-adaptated condition...
2006-08-08 jmadded FLOAT_SQRT
2006-05-10 jmcleaner struct assignments (some compilers complained)
2006-05-10 jmMACROS ARE EVIL. Fixed EXTEND32() and EXTRACT16() and...
2006-05-09 jmanother fix for 16-bit platforms
2006-04-29 jmpatch by Brian Retford to remove the weird struct manip...
2006-02-09 jmfixed lots of warning/errors on retarded (non-C99)...
2005-12-19 jmConverted leak spectral estimation array to int32.
2005-12-18 jmoops, forgot float version
2005-12-17 jmDIVU alias for float
2005-12-17 jmFLOAT_DIVU for computing leak estimate
2005-12-17 jmfixing the float version
2005-12-17 jmminor stuff
2005-12-17 jmgreater-than operator
2005-12-17 jmmost of leakage estimation converted
2005-12-16 jmremoved C++ style comments
2005-12-10 jmthink all the main AEC loops are in fixed-point
2005-12-08 jmconverted power normalization to pseudo-float type
2005-12-08 jmdebugged pseudo-float type, using it for correlations
2005-12-07 jmSupport for a really simple float emulating type