#include <lpc.h>
Data Fields | |
| int | blocksize |
| int | max_order |
| enum FFLPCType | lpc_type |
| double * | windowed_buffer |
| double * | windowed_samples |
| void(* | lpc_apply_welch_window )(const int32_t *data, int len, double *w_data) |
| Apply a Welch window to an array of input samples. | |
| void(* | lpc_compute_autocorr )(const double *data, int len, int lag, double *autoc) |
| Perform autocorrelation on input samples with delay of 0 to lag. | |
Definition at line 50 of file lpc.h.
| void(* LPCContext::lpc_apply_welch_window)(const int32_t *data, int len, double *w_data) |
Apply a Welch window to an array of input samples.
The output samples have the same scale as the input, but are in double sample format.
| data | input samples | |
| len | number of input samples | |
| w_data | output samples |
Referenced by ff_lpc_calc_coefs(), ff_lpc_init(), and ff_lpc_init_x86().
| void(* LPCContext::lpc_compute_autocorr)(const double *data, int len, int lag, double *autoc) |
Perform autocorrelation on input samples with delay of 0 to lag.
| data | input samples. constraints: no alignment needed, but must have have at least lag*sizeof(double) valid bytes preceding it, and size must be at least (len+1)*sizeof(double) if data is 16-byte aligned or (len+2)*sizeof(double) if data is unaligned. | |
| len | number of input samples to process | |
| lag | maximum delay to calculate | |
| autoc | output autocorrelation coefficients. constraints: array size must be at least lag+1. |
Referenced by ff_lpc_calc_coefs(), ff_lpc_init(), and ff_lpc_init_x86().
| double* LPCContext::windowed_buffer |
| double* LPCContext::windowed_samples |
1.5.8