Tool/software: TI-RTOS
I am looking to configure te beamforming demo to access the mic and speaker data using a frame size of 64 instead of 160. The user guide references the sys.h variables to configure. Seems that I can't change the frame size. Is this a configurable option?
From sys.h
/*======================================================================
* Static system configuration parameters
*======================================================================*/
/*======================================================================
* Static system configuration parameters
*======================================================================*/
#define SYS_MICS_MAX 7 /* Maximum number of microphones in the system */
#define SYS_VMICS_MAX 12 /* Maximum number of virtual microphones in the system */
#define SYS_ADC_FS_HZ 16000 /* ADC sampling rate in Hz */
#define SYS_FS_HZ 16000 /* Sampling rate in Hz */
#define SYS_FS_RATIO SYS_ADC_FS_HZ/SYS_FS_HZ /* Sampling rate ratio */
#define SYS_FRAME_DURATION_MS 10 /* Frame duration in ms */
#define SYS_FRAME_LENGTH (1L*SYS_FS_HZ*SYS_FRAME_DURATION_MS/1000) /* Frame length in samples */
#define SYS_FRAME_SIZE (TYP_LINSAMPLE_SIZE*SYS_FRAME_LENGTH/TYP_TWORD_SIZE) /* Frame size in bytes */
#define SYS_IN_LENGTH (2L*SYS_FRAME_LENGTH*SYS_MICS_MAX) /* Input buffer length (dual) */
#define SYS_IN_SIZE (TYP_LINSAMPLE_SIZE*SYS_IN_LENGTH/TYP_TWORD_SIZE) /* # of words in input buffer */
#define SYS_USE_DRC 1 /* Do we want use DRC */
Seems that accessing the mics is settup to be done at 16k on a 10ms frame. Is this configurable to something different. The BUFLEN is hard defined to:
#define BUFLEN 160*2 /* number of samples in the frame */
Any attempt to modify causes issues with the system.
If the frequency is lowered to 8K with a 20ms frame duration, the demo runs but the execution time that the Audio_echo_Task is required to run in remains at 10ms.
Steve
