diff options
Diffstat (limited to 'regress/sys/dev/audio/autest.1')
| -rw-r--r-- | regress/sys/dev/audio/autest.1 | 68 |
1 files changed, 62 insertions, 6 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. |