diff options
author | Jamie Bullock <jamie@postlude.co.uk> | 2008-05-04 11:02:40 +0000 |
---|---|---|
committer | Jamie Bullock <jamie@postlude.co.uk> | 2008-05-04 11:02:40 +0000 |
commit | 6a4be601746f6a84fc70434a61b16bea86b6c147 (patch) | |
tree | fd8e8bfac0e6846f848a38490d583f4bfa7f6a6e /xtract/xtract_scalar.h | |
parent | d817fc9b42b0c04d6452159ca63f5a30834f5489 (diff) | |
download | LibXtract-6a4be601746f6a84fc70434a61b16bea86b6c147.tar.gz LibXtract-6a4be601746f6a84fc70434a61b16bea86b6c147.tar.bz2 LibXtract-6a4be601746f6a84fc70434a61b16bea86b6c147.zip |
Fixed bug in peak interpolation algorithm in xtract_peak_spectrum()
Diffstat (limited to 'xtract/xtract_scalar.h')
-rw-r--r-- | xtract/xtract_scalar.h | 2 |
1 files changed, 2 insertions, 0 deletions
diff --git a/xtract/xtract_scalar.h b/xtract/xtract_scalar.h index 9bccb01..fdfb42a 100644 --- a/xtract/xtract_scalar.h +++ b/xtract/xtract_scalar.h @@ -394,6 +394,8 @@ int xtract_hps(const float *data, const int N, const void *argv, float *result); * \param *result: the pitch of N values from the array pointed to by *data * * This algorithm is based on the AMDF, with peak and centre clipping. It would benefit from further improvements to improve noise robustness and overall efficiency + * + * It is based on suggestion by Robert Bristow-Johnson in a discussion on the comp.dsp mailing list, subject "Reference implementation of pitch detection" * */ int xtract_f0(const float *data, const int N, const void *argv, float *result); |