M
Das Format der Sound-Daten, die du an /dev/dsp schickst, stellt man normalerweise über die ioctl()-Aufrufe
SNDCTL_DSP_SETFMT
SNDCTL_DSP_STEREO
SNDCTL_DSP_SPEED
ein. Beispiel aus meinem Buch "C und Linux":
/*
rec_play.c - Audio-Signal aufnehmen u. wiedergeben
*/
# include <stdio.h>
# include <unistd.h>
# include <fcntl.h>
# include <sys/ioctl.h>
# include <linux/soundcard.h>
# define NUM_SAMPLES 100000
unsigned char buffer[NUM_SAMPLES];
int main()
{
int fd, i, format=AFMT_U8;
long length;
char input[16];
if ((fd = open("/dev/dsp", O_RDWR)) == -1)
{
perror("rec_play: Can't open device");
return(1);
}
if (ioctl(fd, SNDCTL_DSP_SETFMT, &format) == -1)
perror("rec_play: Can't set format");
i = 0;
if (ioctl(fd, SNDCTL_DSP_STEREO, &i) == -1)
perror("rec_play: Can't set to mono");
i = 22050;
if (ioctl(fd, SNDCTL_DSP_SPEED, &i) == -1)
perror("rec_play: Can't set sampling rate");
printf("Press <RETURN> to start recording. ");
fgets(input, 16, stdin);
if ((length = read(fd, buffer, NUM_SAMPLES)) == -1)
{
perror("rec_play: Can't record audio data");
return(1);
}
printf("done (%ld bytes).\n"
"Press <RETURN> to start playing. ", length);
fgets(input, 16, stdin);
if (write(fd, buffer, length) == -1)
perror("rec_play: Can't play audio data");
close(fd);
return(0);
}
Martin
