fixed-point: changed find_spectral_pitch() to use single-precision (16-bit) FFT.
authorJean-Marc Valin <Jean-Marc.Valin@csiro.au>
Wed, 5 Mar 2008 06:20:30 +0000 (17:20 +1100)
committerJean-Marc Valin <Jean-Marc.Valin@csiro.au>
Wed, 5 Mar 2008 06:20:30 +0000 (17:20 +1100)
commitf93747c44ab8e1c2d31b458ae1ffc0f85952b67c
tree6ccaa400eeeaca92ebec47d90f26d65326f1e64d
parent4024444811949fbfe2dc13ad30868d0b30e5d5f3
fixed-point: changed find_spectral_pitch() to use single-precision (16-bit) FFT.
This involved adding kfft_single.[ch] that redefines kiss_fft a second time
with a different prefix. All this is still a bit of a mess now. The mask
had to be converted to 16-bit input, but we're still using floats to apply it.
12 files changed:
libcelt/Makefile.am
libcelt/_kiss_fft_guts.h
libcelt/celt.c
libcelt/kfft_single.c [new file with mode: 0644]
libcelt/kfft_single.h [new file with mode: 0644]
libcelt/kiss_fft.c
libcelt/kiss_fft.h
libcelt/kiss_fftr.c
libcelt/pitch.c
libcelt/pitch.h
libcelt/psy.c
libcelt/psy.h