diff options
author | Dale Rahn <drahn@cvs.openbsd.org> | 2005-09-27 02:53:44 +0000 |
---|---|---|
committer | Dale Rahn <drahn@cvs.openbsd.org> | 2005-09-27 02:53:44 +0000 |
commit | b175f0b94f1858b9a5a79334507588477d0488b8 (patch) | |
tree | 265fb1e68eb8cefbc0bbbed7a065b01169851c45 /regress/sys | |
parent | c3472f9013768c705615e3e89b073506ba3af2e6 (diff) |
report time duration required to play sample then % error of sample rate.
from PR 4304. Also added support for different sample rate requests.
ok jason@
Diffstat (limited to 'regress/sys')
-rw-r--r-- | regress/sys/dev/audio/autest.1 | 68 | ||||
-rw-r--r-- | regress/sys/dev/audio/autest.c | 176 |
2 files changed, 195 insertions, 49 deletions
diff --git a/regress/sys/dev/audio/autest.1 b/regress/sys/dev/audio/autest.1 index 920d49a4b26..f25c0eb648e 100644 --- a/regress/sys/dev/audio/autest.1 +++ b/regress/sys/dev/audio/autest.1 @@ -1,4 +1,4 @@ -.\" $OpenBSD: autest.1,v 1.6 2003/06/02 19:15:38 jason Exp $ +.\" $OpenBSD: autest.1,v 1.7 2005/09/27 02:53:43 drahn Exp $ .\" .\" Copyright (c) 2002 Jason L. Wright (jason@thought.net) .\" All rights reserved. @@ -33,6 +33,7 @@ .Sh SYNOPSIS .Nm autest .Op Fl f Ar device +.Op Fl r Ar rate .Sh DESCRIPTION The .Nm @@ -47,17 +48,24 @@ if not specified, and iterates through all of the encodings supported by the device playing a 440Hz tone in the proper format. The tone should sound almost identical in each of the formats. +The +.Fl r +rate +can be used to specify the audio rate to test, It will request the +audio subsystem to play that Hz, however the audio device may return +a different speed. This can be useful to test different speeds, eg 8000, +44100, 48000. .Pp .Nm can produce tones in any of the following formats and will skip other formats if supported by the device: .Bl -tag -width XXXXXXXXXX .It Cm mu-law -8bit mu-law companded +8 bit mu-law companded .It Cm A-law -8bit A-law companded -.It Cm adpcm -4 bit adaptive differential pulse code modulation +8 bit A-law companded +.\" .It Cm adpcm +.\" 4 bit adaptive differential pulse code modulation .It Cm ulinear 8 bit unsigned linear .It Cm ulinear_le @@ -71,7 +79,55 @@ formats if supported by the device: .It Cm slinear_be 16 bit signed linear big endian (twos complement) .El +.Sh OUTPUT +Interpreting the output of +.Nm +is a little tricky. +The output below is from an +.Xr auich 4 . +.Bd -literal +ulinear:8...mono(s 44100 c 45167 e 2.4%)...stereo(s 44100 c 45162 e 2.4%) +mulaw:8*...mono(s 44100 c 45166 e 2.4%)...stereo(s 44100 c 45157 e 2.3%) +alaw:8*...mono[Invalid argument]...stereo[Invalid argument] +slinear:8*...mono(s 44100 c 45171 e 2.4%)...stereo(s 44100 c 45170 e 2.4%) +slinear_le:16...mono(s 44100 c 45171 e 2.4%)...stereo(s 44100 c 45170 e 2.4%) +ulinear_le:16*...mono(s 44100 c 45167 e 2.4%)...stereo(s 44100 c 45168 e 2.4%) +slinear_be:16*...mono(s 44100 c 45169 e 2.4%)...stereo(s 44100 c 45167 e 2.4%) +ulinear_be:16*...mono(s 44100 c 45167 e 2.4%)...stereo(s 44100 c 45160 e 2.3%) +.Ed +.Pp +.Nm +loops through each mode claimed to be supported by the chip (emulated or not). +For each emulated mode, an asterisk is appended to the mode name. +If the tone sounds different from all the other tones for the device, +it is likely that either the emulation is wrong, or the mode is not +set on the device correctly. +.Pp +Also, for the mono and stereo versions of each mode, +.Nm +prints the claimed sample rate, +.So s Bo rate Bc Sc , +computed sample rate, +.So c Bo rate Bc Sc , +and the percent error between them, +.So e Bo percent Bc Sc . +If the percent error is high (greater than 10 percent or so), +either the sample rate is not being correctly returned by +the device, or it is not being set correctly on the device. +.Pp +Interestingly, when +.Nm +requests +.Sq alaw +encoding, the device driver returns +.So Invalid argument Sc . +This indicates that the device includes +.Sq alaw +in its mode enumeration, but does not support it for playback. +This is very likely a bug in the driver. .Sh SEE ALSO .Xr audio 4 .Sh BUGS -The ADPCM encoding sounds very noisy on CS4231 (it's probably incorrect). +There is partial support for ADPCM, adaptive differential pulse code +modulation, but it is not enabled by default +since it does not appear to be correct. diff --git a/regress/sys/dev/audio/autest.c b/regress/sys/dev/audio/autest.c index 608386f68f7..171b99dbcb6 100644 --- a/regress/sys/dev/audio/autest.c +++ b/regress/sys/dev/audio/autest.c @@ -1,4 +1,4 @@ -/* $OpenBSD: autest.c,v 1.10 2003/08/06 16:15:44 jason Exp $ */ +/* $OpenBSD: autest.c,v 1.11 2005/09/27 02:53:43 drahn Exp $ */ /* * Copyright (c) 2002 Jason L. Wright (jason@thought.net) @@ -44,18 +44,31 @@ #include "adpcm.h" #include "law.h" +struct ausrate { + struct timeval tv_begin; + struct timeval tv_end; + u_int r_rate; /* requested rate */ + u_int s_rate; /* rate from audio layer */ + u_int c_rate; /* computed rate */ + int bps; /* bytes per sample */ + int bytes; /* number of bytes played */ + float err; +}; + int main(int, char **); -void check_encoding(int, audio_encoding_t *); -void check_encoding_mono(int, audio_encoding_t *); -void check_encoding_stereo(int, audio_encoding_t *); -void enc_ulaw_8(int, audio_encoding_t *, int); -void enc_alaw_8(int, audio_encoding_t *, int); -void enc_ulinear_8(int, audio_encoding_t *, int); -void enc_ulinear_16(int, audio_encoding_t *, int, int); -void enc_slinear_8(int, audio_encoding_t *, int); -void enc_slinear_16(int, audio_encoding_t *, int, int); -void enc_adpcm_8(int, audio_encoding_t *, int); +void check_encoding(int, audio_encoding_t *, int); +void check_encoding_mono(int, audio_encoding_t *, int); +void check_encoding_stereo(int, audio_encoding_t *, int); +void enc_ulaw_8(int, audio_encoding_t *, int, int); +void enc_alaw_8(int, audio_encoding_t *, int, int); +void enc_ulinear_8(int, audio_encoding_t *, int, int); +void enc_ulinear_16(int, audio_encoding_t *, int, int, int); +void enc_slinear_8(int, audio_encoding_t *, int, int); +void enc_slinear_16(int, audio_encoding_t *, int, int, int); +void enc_adpcm_8(int, audio_encoding_t *, int, int); void audio_wait(int); +void check_srate(struct ausrate *); +void mark_time(struct timeval *); #define PLAYFREQ 440.0 #define PLAYSECS 2 @@ -68,12 +81,16 @@ main(int argc, char **argv) audio_info_t ainfo; char *fname = NULL; int fd, i, c; + int rate = 8000; - while ((c = getopt(argc, argv, "f:")) != -1) { + while ((c = getopt(argc, argv, "f:r:")) != -1) { switch (c) { case 'f': fname = optarg; break; + case 'r': + rate = atoi(optarg); + break; case '?': default: fprintf(stderr, "%s [-f device]\n", argv[0]); @@ -98,7 +115,7 @@ main(int argc, char **argv) enc.index = i; if (ioctl(fd, AUDIO_GETENC, &enc) == -1) break; - check_encoding(fd, &enc); + check_encoding(fd, &enc, rate); } close(fd); @@ -106,20 +123,46 @@ main(int argc, char **argv) } void -check_encoding(int fd, audio_encoding_t *enc) +check_srate(struct ausrate *rt) +{ + struct timeval t; + float tm, b, r, err; + + timersub(&rt->tv_end, &rt->tv_begin, &t); + tm = (float)t.tv_sec + ((float)t.tv_usec / 1000000.0); + b = (float)rt->bytes / (float)rt->bps; + r = b / tm; + + err = fabs((float)rt->s_rate - r); + err /= r * 0.01; + rt->err = err; + rt->c_rate = rintf(r); + printf("(s %u c %u e %3.1f%%)", + rt->s_rate, rt->c_rate, rt->err); +} + +void +check_encoding(int fd, audio_encoding_t *enc, int rate) { printf("%s:%d%s", enc->name, enc->precision, (enc->flags & AUDIO_ENCODINGFLAG_EMULATED) ? "*" : ""); fflush(stdout); - check_encoding_mono(fd, enc); - check_encoding_stereo(fd, enc); + check_encoding_mono(fd, enc, rate); + check_encoding_stereo(fd, enc, rate); printf("\n"); } void -check_encoding_mono(int fd, audio_encoding_t *enc) +mark_time(struct timeval *tv) +{ + if (gettimeofday(tv, NULL) == -1) + err(1, "gettimeofday"); +} + +void +check_encoding_mono(int fd, audio_encoding_t *enc, int rate) { int skipped = 0; @@ -129,23 +172,23 @@ check_encoding_mono(int fd, audio_encoding_t *enc) if (enc->precision == 8) { switch (enc->encoding) { case AUDIO_ENCODING_ULAW: - enc_ulaw_8(fd, enc, 1); + enc_ulaw_8(fd, enc, 1, rate); break; case AUDIO_ENCODING_ALAW: - enc_alaw_8(fd, enc, 1); + enc_alaw_8(fd, enc, 1, rate); break; case AUDIO_ENCODING_ULINEAR: case AUDIO_ENCODING_ULINEAR_LE: case AUDIO_ENCODING_ULINEAR_BE: - enc_ulinear_8(fd, enc, 1); + enc_ulinear_8(fd, enc, 1, rate); break; case AUDIO_ENCODING_SLINEAR: case AUDIO_ENCODING_SLINEAR_LE: case AUDIO_ENCODING_SLINEAR_BE: - enc_slinear_8(fd, enc, 1); + enc_slinear_8(fd, enc, 1, rate); break; case AUDIO_ENCODING_ADPCM: - enc_adpcm_8(fd, enc, 1); + enc_adpcm_8(fd, enc, 1, rate); break; default: skipped = 1; @@ -155,16 +198,16 @@ check_encoding_mono(int fd, audio_encoding_t *enc) if (enc->precision == 16) { switch (enc->encoding) { case AUDIO_ENCODING_ULINEAR_LE: - enc_ulinear_16(fd, enc, 1, LITTLE_ENDIAN); + enc_ulinear_16(fd, enc, 1, LITTLE_ENDIAN, rate); break; case AUDIO_ENCODING_ULINEAR_BE: - enc_ulinear_16(fd, enc, 1, BIG_ENDIAN); + enc_ulinear_16(fd, enc, 1, BIG_ENDIAN, rate); break; case AUDIO_ENCODING_SLINEAR_LE: - enc_slinear_16(fd, enc, 1, LITTLE_ENDIAN); + enc_slinear_16(fd, enc, 1, LITTLE_ENDIAN, rate); break; case AUDIO_ENCODING_SLINEAR_BE: - enc_slinear_16(fd, enc, 1, BIG_ENDIAN); + enc_slinear_16(fd, enc, 1, BIG_ENDIAN, rate); break; default: skipped = 1; @@ -176,7 +219,7 @@ check_encoding_mono(int fd, audio_encoding_t *enc) } void -check_encoding_stereo(int fd, audio_encoding_t *enc) +check_encoding_stereo(int fd, audio_encoding_t *enc, int rate) { int skipped = 0; @@ -186,23 +229,23 @@ check_encoding_stereo(int fd, audio_encoding_t *enc) if (enc->precision == 8) { switch (enc->encoding) { case AUDIO_ENCODING_ULAW: - enc_ulaw_8(fd, enc, 2); + enc_ulaw_8(fd, enc, 2, rate); break; case AUDIO_ENCODING_ALAW: - enc_alaw_8(fd, enc, 2); + enc_alaw_8(fd, enc, 2, rate); break; case AUDIO_ENCODING_ULINEAR: case AUDIO_ENCODING_ULINEAR_LE: case AUDIO_ENCODING_ULINEAR_BE: - enc_ulinear_8(fd, enc, 2); + enc_ulinear_8(fd, enc, 2, rate); break; case AUDIO_ENCODING_SLINEAR: case AUDIO_ENCODING_SLINEAR_LE: case AUDIO_ENCODING_SLINEAR_BE: - enc_slinear_8(fd, enc, 2); + enc_slinear_8(fd, enc, 2, rate); break; case AUDIO_ENCODING_ADPCM: - enc_adpcm_8(fd, enc, 2); + enc_adpcm_8(fd, enc, 2, rate); break; default: skipped = 1; @@ -212,16 +255,16 @@ check_encoding_stereo(int fd, audio_encoding_t *enc) if (enc->precision == 16) { switch (enc->encoding) { case AUDIO_ENCODING_ULINEAR_LE: - enc_ulinear_16(fd, enc, 2, LITTLE_ENDIAN); + enc_ulinear_16(fd, enc, 2, LITTLE_ENDIAN, rate); break; case AUDIO_ENCODING_ULINEAR_BE: - enc_ulinear_16(fd, enc, 2, BIG_ENDIAN); + enc_ulinear_16(fd, enc, 2, BIG_ENDIAN, rate); break; case AUDIO_ENCODING_SLINEAR_LE: - enc_slinear_16(fd, enc, 2, LITTLE_ENDIAN); + enc_slinear_16(fd, enc, 2, LITTLE_ENDIAN, rate); break; case AUDIO_ENCODING_SLINEAR_BE: - enc_slinear_16(fd, enc, 2, BIG_ENDIAN); + enc_slinear_16(fd, enc, 2, BIG_ENDIAN, rate); break; default: skipped = 1; @@ -233,9 +276,10 @@ check_encoding_stereo(int fd, audio_encoding_t *enc) } void -enc_ulinear_8(int fd, audio_encoding_t *enc, int chans) +enc_ulinear_8(int fd, audio_encoding_t *enc, int chans, int rate) { audio_info_t inf; + struct ausrate rt; u_int8_t *samples = NULL, *p; int i, j; @@ -243,6 +287,7 @@ enc_ulinear_8(int fd, audio_encoding_t *enc, int chans) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -253,6 +298,10 @@ enc_ulinear_8(int fd, audio_encoding_t *enc, int chans) printf("[getinfo: %s]", strerror(errno)); goto out; } + rt.r_rate = inf.play.sample_rate; + rt.s_rate = inf.play.sample_rate; + rt.bps = 1 * chans; + rt.bytes = inf.play.sample_rate * chans * PLAYSECS; samples = (u_int8_t *)malloc(inf.play.sample_rate * chans); if (samples == NULL) { @@ -275,9 +324,12 @@ enc_ulinear_8(int fd, audio_encoding_t *enc, int chans) } } + mark_time(&rt.tv_begin); for (i = 0; i < PLAYSECS; i++) write(fd, samples, inf.play.sample_rate * chans); audio_wait(fd); + mark_time(&rt.tv_end); + check_srate(&rt); out: if (samples != NULL) @@ -285,9 +337,10 @@ out: } void -enc_slinear_8(int fd, audio_encoding_t *enc, int chans) +enc_slinear_8(int fd, audio_encoding_t *enc, int chans, int rate) { audio_info_t inf; + struct ausrate rt; int8_t *samples = NULL, *p; int i, j; @@ -295,6 +348,7 @@ enc_slinear_8(int fd, audio_encoding_t *enc, int chans) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -305,6 +359,10 @@ enc_slinear_8(int fd, audio_encoding_t *enc, int chans) printf("[getinfo: %s]", strerror(errno)); goto out; } + rt.r_rate = inf.play.sample_rate; + rt.s_rate = inf.play.sample_rate; + rt.bps = 1 * chans; + rt.bytes = inf.play.sample_rate * chans * PLAYSECS; samples = (int8_t *)malloc(inf.play.sample_rate * chans); if (samples == NULL) { @@ -327,9 +385,12 @@ enc_slinear_8(int fd, audio_encoding_t *enc, int chans) } } + mark_time(&rt.tv_begin); for (i = 0; i < PLAYSECS; i++) write(fd, samples, inf.play.sample_rate * chans); audio_wait(fd); + mark_time(&rt.tv_end); + check_srate(&rt); out: if (samples != NULL) @@ -337,9 +398,10 @@ out: } void -enc_slinear_16(int fd, audio_encoding_t *enc, int chans, int order) +enc_slinear_16(int fd, audio_encoding_t *enc, int chans, int order, int rate) { audio_info_t inf; + struct ausrate rt; u_int8_t *samples = NULL, *p; int i, j; @@ -347,6 +409,7 @@ enc_slinear_16(int fd, audio_encoding_t *enc, int chans, int order) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -357,6 +420,10 @@ enc_slinear_16(int fd, audio_encoding_t *enc, int chans, int order) printf("[getinfo: %s]", strerror(errno)); goto out; } + rt.r_rate = inf.play.sample_rate; + rt.s_rate = inf.play.sample_rate; + rt.bps = 2 * chans; + rt.bytes = 2 * inf.play.sample_rate * chans * PLAYSECS; samples = (int8_t *)malloc(inf.play.sample_rate * chans * 2); if (samples == NULL) { @@ -388,9 +455,12 @@ enc_slinear_16(int fd, audio_encoding_t *enc, int chans, int order) } } + mark_time(&rt.tv_begin); for (i = 0; i < PLAYSECS; i++) write(fd, samples, inf.play.sample_rate * chans * 2); audio_wait(fd); + mark_time(&rt.tv_end); + check_srate(&rt); out: if (samples != NULL) @@ -398,9 +468,10 @@ out: } void -enc_ulinear_16(int fd, audio_encoding_t *enc, int chans, int order) +enc_ulinear_16(int fd, audio_encoding_t *enc, int chans, int order, int rate) { audio_info_t inf; + struct ausrate rt; u_int8_t *samples = NULL, *p; int i, j; @@ -408,6 +479,7 @@ enc_ulinear_16(int fd, audio_encoding_t *enc, int chans, int order) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -424,6 +496,10 @@ enc_ulinear_16(int fd, audio_encoding_t *enc, int chans, int order) warn("malloc"); goto out; } + rt.r_rate = inf.play.sample_rate; + rt.s_rate = inf.play.sample_rate; + rt.bps = 2 * chans; + rt.bytes = 2 * inf.play.sample_rate * chans * PLAYSECS; for (i = 0, p = samples; i < inf.play.sample_rate; i++) { float d; @@ -449,9 +525,12 @@ enc_ulinear_16(int fd, audio_encoding_t *enc, int chans, int order) } } + mark_time(&rt.tv_begin); for (i = 0; i < PLAYSECS; i++) write(fd, samples, inf.play.sample_rate * chans * 2); audio_wait(fd); + mark_time(&rt.tv_end); + check_srate(&rt); out: if (samples != NULL) @@ -459,7 +538,7 @@ out: } void -enc_adpcm_8(int fd, audio_encoding_t *enc, int chans) +enc_adpcm_8(int fd, audio_encoding_t *enc, int chans, int rate) { audio_info_t inf; struct adpcm_state adsts; @@ -471,6 +550,7 @@ enc_adpcm_8(int fd, audio_encoding_t *enc, int chans) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -533,17 +613,19 @@ out: } void -enc_ulaw_8(int fd, audio_encoding_t *enc, int chans) +enc_ulaw_8(int fd, audio_encoding_t *enc, int chans, int rate) { audio_info_t inf; int16_t *samples = NULL; int i, j; u_int8_t *outbuf = NULL, *p; + struct ausrate rt; AUDIO_INITINFO(&inf); inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); @@ -554,6 +636,10 @@ enc_ulaw_8(int fd, audio_encoding_t *enc, int chans) printf("[getinfo: %s]", strerror(errno)); goto out; } + rt.r_rate = inf.play.sample_rate; + rt.s_rate = inf.play.sample_rate; + rt.bps = 1 * chans; + rt.bytes = inf.play.sample_rate * chans * PLAYSECS; samples = (int16_t *)calloc(inf.play.sample_rate, sizeof(*samples)); if (samples == NULL) { @@ -582,10 +668,13 @@ enc_ulaw_8(int fd, audio_encoding_t *enc, int chans) } } + mark_time(&rt.tv_begin); for (i = 0; i < PLAYSECS; i++) { write(fd, outbuf, inf.play.sample_rate * chans); } audio_wait(fd); + mark_time(&rt.tv_end); + check_srate(&rt); out: if (samples != NULL) @@ -595,7 +684,7 @@ out: } void -enc_alaw_8(int fd, audio_encoding_t *enc, int chans) +enc_alaw_8(int fd, audio_encoding_t *enc, int chans, int rate) { audio_info_t inf; int16_t *samples = NULL; @@ -606,6 +695,7 @@ enc_alaw_8(int fd, audio_encoding_t *enc, int chans) inf.play.precision = enc->precision; inf.play.encoding = enc->encoding; inf.play.channels = chans; + inf.play.sample_rate = rate;; if (ioctl(fd, AUDIO_SETINFO, &inf) == -1) { printf("[%s]", strerror(errno)); |