Programmer Guide/SPU Reference/PZTF: Difference between revisions
From STX Wiki
Jump to navigationJump to search
No edit summary |
No edit summary |
||
| Line 10: | Line 10: | ||
=====Inputs:===== | =====Inputs:===== | ||
{ | {| | ||
; | ! style="text-align:leftr;"|Name | ||
! Type | |||
! Description | |||
! Default | |||
|- | |||
|SP | |||
|vec | |||
|log. amplitude spectrum | |||
|no default | |||
|- | |||
|DF | |||
|num | |||
|frequency resolution in Hz | |||
|no default | |||
|- | |||
|F0 | |||
|num | |||
|fundamental frequency in Hz | |||
|estimated if not supplied | |||
|- | |||
|F1,F2 | |||
|num | |||
|component search range for f0 estimation in Hz | |||
|75-4000Hz | |||
|- | |||
|NC | |||
|num | |||
|number of f0 components to use for harmonic mean | |||
|5 | |||
|- | |||
|IINT | |||
|num/str | |||
|envelope interpolation method (0=Linear, 1=Qaudratic, 2=SPLine) | |||
|0=Linear | |||
|- | |||
|NN | |||
|num | |||
|number of numerator coefficients (>0) | |||
|no default | |||
|- | |||
|ND | |||
|num | |||
|number of denominator coefficients (>1) | |||
|no default | |||
|- | |||
|IALG | |||
|num/str | |||
|pole-zero algorithm (0=WLLS, 1=ND) | |||
|1=ND | |||
|} | |||
Out: F0 num fundamental frequency in Hz (input or estimated, 0 for error) | Out: F0 num fundamental frequency in Hz (input or estimated, 0 for error) | ||
Revision as of 08:31, 24 February 2015
PZTF
pole-zero model estimation
Usage:
[SPU PZTF SP DF F0 FMIN FMAX NC IINT NC NN ND IALG OUT F0 ENV C NZ NP F R H E]
Inputs:
| Name | Type | Description | Default |
|---|---|---|---|
| SP | vec | log. amplitude spectrum | no default |
| DF | num | frequency resolution in Hz | no default |
| F0 | num | fundamental frequency in Hz | estimated if not supplied |
| F1,F2 | num | component search range for f0 estimation in Hz | 75-4000Hz |
| NC | num | number of f0 components to use for harmonic mean | 5 |
| IINT | num/str | envelope interpolation method (0=Linear, 1=Qaudratic, 2=SPLine) | 0=Linear |
| NN | num | number of numerator coefficients (>0) | no default |
| ND | num | number of denominator coefficients (>1) | no default |
| IALG | num/str | pole-zero algorithm (0=WLLS, 1=ND) | 1=ND |
Out: F0 num fundamental frequency in Hz (input or estimated, 0 for error) ENV vec log. spectrum envelope C vec numerator and denominator coefficients NZ num number of zeros NP num number of poles F vec frequency of zeros and poles R vec radius of zeros and poles H vec log. amplitude response E num error (avr((ENV - H)^2)
