+14 -1

CHANGELOG

reviewed

··· 1 1 + 1.3.0 2012-07-18 2 2 + removed non-standard malloc.h from kiss_fft.h 3 3 + 4 4 + moved -lm to end of link line 5 5 + 6 6 + checked various return values 7 7 + 8 8 + converted python Numeric code to NumPy 9 9 + 10 10 + fixed test of int32_t on 64 bit OS 11 11 + 12 12 + added padding in a couple of places to allow SIMD alignment of structs 13 13 + 1 14 1.2.9 2010-05-27 2 15 threadsafe ( including OpenMP ) 3 16 ··· 24 37 25 38 1.2.4 (Oct 27, 2005) The "oops, inverse fixed point real fft was borked" release. 26 39 Fixed scaling bug for inverse fixed point real fft -- also fixed test code that should've been failing. 27 27 - Thanks to Jean-Marc Valin for bug report. 40 40 + Thanks to Jean-Marc Valin for bug report. 28 41 29 42 Use sys/types.h for more portable types than short,int,long => int16_t,int32_t,int64_t 30 43 If your system does not have these, you may need to define them -- but at least it breaks in a

+1 -1

Makefile

reviewed

··· 1 1 - KFVER=129 1 1 + KFVER=130 2 2 3 3 doc: 4 4 @echo "Start by reading the README file. If you want to build and test lots of stuff, do a 'make testall'"

+2 -2

README

reviewed

··· 40 40 * fast convolution FIR filtering (not available for fixed point) 41 41 * spectrum image creation 42 42 43 43 - The core fft and most tools/ code can be compiled to use float, double 44 44 - or Q15 short samples. The default is float. 43 43 + The core fft and most tools/ code can be compiled to use float, double, 44 44 + Q15 short or Q31 samples. The default is float. 45 45 46 46 47 47 BACKGROUND: