Changeset a95b386 for src/spectral/filterbank_mel.h
- Timestamp:
- Nov 17, 2018, 4:38:29 PM (5 years ago)
- Branches:
- feature/autosink, feature/cnn, feature/cnn_org, feature/constantq, feature/crepe, feature/crepe_org, feature/pitchshift, feature/timestretch, fix/ffmpeg5, master
- Children:
- ab8e838
- Parents:
- 75f9fff (diff), 2eb52bd (diff)
Note: this is a merge changeset, the changes displayed below correspond to the merge itself.
Use the(diff)
links above to see all the changes relative to each parent. - File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
src/spectral/filterbank_mel.h
r75f9fff ra95b386 56 56 57 57 \param fb filterbank object 58 \param samplerate audio sampling rate 58 \param samplerate audio sampling rate, in Hz 59 59 60 The filter coefficients are built according toMalcolm Slaney's Auditory61 Toolbox , available online at the following address (see file mfcc.m):60 The filter coefficients are built to match exactly Malcolm Slaney's Auditory 61 Toolbox implementation (see file mfcc.m). The number of filters should be 40. 62 62 63 References 64 ---------- 65 66 Malcolm Slaney, *Auditory Toolbox Version 2, Technical Report #1998-010* 63 67 https://engineering.purdue.edu/~malcolm/interval/1998-010/ 64 68 … … 67 71 smpl_t samplerate); 68 72 73 /** Mel filterbank initialization 74 75 \param fb filterbank object 76 \param samplerate audio sampling rate 77 \param fmin start frequency, in Hz 78 \param fmax end frequency, in Hz 79 80 The filterbank will be initialized with bands linearly spaced in the mel 81 scale, from `fmin` to `fmax`. 82 83 References 84 ---------- 85 86 Malcolm Slaney, *Auditory Toolbox Version 2, Technical Report #1998-010* 87 https://engineering.purdue.edu/~malcolm/interval/1998-010/ 88 89 */ 90 uint_t aubio_filterbank_set_mel_coeffs(aubio_filterbank_t * fb, 91 smpl_t samplerate, smpl_t fmin, smpl_t fmax); 92 93 /** Mel filterbank initialization 94 95 \param fb filterbank object 96 \param samplerate audio sampling rate 97 \param fmin start frequency, in Hz 98 \param fmax end frequency, in Hz 99 100 The bank of filters will be initalized to to cover linearly spaced bands in 101 the Htk mel scale, from `fmin` to `fmax`. 102 103 References 104 ---------- 105 106 Douglas O'Shaughnessy (1987). *Speech communication: human and machine*. 107 Addison-Wesley. p. 150. ISBN 978-0-201-16520-3. 108 109 HTK Speech Recognition Toolkit: http://htk.eng.cam.ac.uk/ 110 111 */ 112 uint_t aubio_filterbank_set_mel_coeffs_htk(aubio_filterbank_t * fb, 113 smpl_t samplerate, smpl_t fmin, smpl_t fmax); 114 69 115 #ifdef __cplusplus 70 116 }
Note: See TracChangeset
for help on using the changeset viewer.