Let src/usr.bin build with recursive parallel make..
- add .WAIT and .NOTPARALLEL in a few places
- change ${MAKE} print-objdir to ${PRINTOBJDIR}
- convert other ad-hoc forms to use ${PRINTOBJDIR}

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Enable WARNS=4 by default for usr.bin, except for:
awk bdes checknr compile_et error gss hxtool kgetcred kinit
klist ldd less lex locale login m4 man menuc mk_cmds
mklocale msgc openssl rpcgen rpcinfo sdiff spell ssh
string2key telnet tn3270 verify_krb5_conf xlint

branches: 1.4.44;
- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

-g for debug, try without when done debugging

Let src/usr.bin build with recursive parallel make..
- add .WAIT and .NOTPARALLEL in a few places
- change ${MAKE} print-objdir to ${PRINTOBJDIR}
- convert other ad-hoc forms to use ${PRINTOBJDIR}

branches: 1.1.10;
add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

separate the 3 functions midirecord uses from libaudio.a into its own
file and link it directly, instead of having an (implicit) dependancy
on usr.bin/audio/common being built before usr.bin/midirecord is linked.

Rename MKPRIVATELIB to LIBISPRIVATE, to make it clearer that this is
a variable that is used by in-tree Makefiles to control behaviour.
(MKsomevar variables are generally intended to be controlled by the end-user)

branches: 1.7.26;
Use MKPRIVATELIB=yes instead of providing an empty libinstall:: target and
setting NOLINT, NOPIC, NOPROFILE (etc)

inline the contents of auconv.c into auconv.h. this makes audiorecord(1)
slightly smaller (about 200 bytes on x86) and i would guess slightly faster.

makefile delint. use NETBSDSRCDIR as appropriate

for sun output files, if we have unsigned or little endian data, convert
it to signed big endian data. pull in a modified copy of the kernels'
auconv.[ch] to help with this (and remove some home grown code that did
some of the same things.)

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

MKfoo=no -> NOfoo

branches: 1.1.10;
add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

inline the contents of auconv.c into auconv.h. this makes audiorecord(1)
slightly smaller (about 200 bytes on x86) and i would guess slightly faster.

branches: 1.1.2;
for sun output files, if we have unsigned or little endian data, convert
it to signed big endian data. pull in a modified copy of the kernels'
auconv.[ch] to help with this (and remove some home grown code that did
some of the same things.)

Add support for recording 24bit wav files.

branches: 1.5.54; 1.5.64; 1.5.72;
Remove clause 3 and 4 from TNF licenses

branches: 1.4.20;
__inline__ -> inline

Fix fencepost errors in the conversion functions (making them a bit more like
the kernel versions).

branches: 1.2.2;
inline the contents of auconv.c into auconv.h. this makes audiorecord(1)
slightly smaller (about 200 bytes on x86) and i would guess slightly faster.

branches: 1.1.2;
for sun output files, if we have unsigned or little endian data, convert
it to signed big endian data. pull in a modified copy of the kernels'
auconv.[ch] to help with this (and remove some home grown code that did
some of the same things.)

Clean up lines containing only tabs.

branches: 1.26.8;
add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.25.8; 1.25.16; 1.25.18;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

separate the 3 functions midirecord uses from libaudio.a into its own
file and link it directly, instead of having an (implicit) dependancy
on usr.bin/audio/common being built before usr.bin/midirecord is linked.

add a decode_uint() function.

branches: 1.22.4;
move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.21.2; 1.21.8;
recognize AUDIO_ENCODING_AC3

ANSIfy. Add static. Add __dead.

remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.18.26; 1.18.28;
Add (unsigned char) cast to ctype functions

Add NetBSD RCS Ids.

- decode_int(): always use base 10, so that a time of "1:09" works.
- decode_time(): 60 seconds in a minute, really.
both from tsarna@netbsd.org.

- be verbose about the time we are going to record for, if verbose.

- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

fix 2001/03/04 change: audio_parse_wav_hdr() returns a signed value.

fix bug in audiorecord -t handling.

Don't play trailing garbage at end of RIFF WAVE (.wav) files
Patch sent by YAMAMOTO Takashi <takashi.yamamoto@bigfoot.com> in PR 12328

libaudio:
- rename audio_get_sun_encoding() to audio_sun_to_encoding()
- add audio_encoding_to_sun()
play.c:
- adapt
record.c:
- convert NetBSD encoding to sun encoding if possible, otherwise don't
output a header at all.
fixes PR#10380 from Michael Eriksson <eramore@era-t.ericsson.se>.

branches: 1.9.6;
fix off-by-one error in error message array. add a comment.

branches: 1.8.2;
Use "NULL" instead of "0" for null pointer.

branches: 1.7.2;
remove duplicate include of <stdio.h>

Only 8-bit data is unsigned in WAVE files.

.wav files are little endian.

make recording precision/encoding setting sane.

add support to play stdin.

add missing copyrights.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Clean up lines containing only tabs.

branches: 1.1.8; 1.1.18; 1.1.26;
separate the 3 functions midirecord uses from libaudio.a into its own
file and link it directly, instead of having an (implicit) dependancy
on usr.bin/audio/common being built before usr.bin/midirecord is linked.

audio_wav_parse_hdr(): avoid buffer overreads and clean up

reimplement most of this function using a new method of buffer
management to ensure that we never read beyond the provided size.

properly handle RIFF chunk lengths, instead of assuming various
offsets from most files are right.

update list of consumed documentation and fill the list of WAVE
formats from RFC 2361 (most remain not supported.)

should fix PR#57973.

tested against a large number of .wav files i have handy and with
a testsuite generator that should be incoming soon.

branches: 1.21.8;
add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.20.8; 1.20.16; 1.20.18;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

add a decode_uint() function.

branches: 1.18.4;
move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.17.2; 1.17.8;
ANSIfy. Add static. Add __dead.

- add extended WAVE header support
- attempt to play a bunch more WAV files

branches: 1.15.4; 1.15.10;
remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.14.20; 1.14.22;
variable signed/unsigned cleanup.

put the __packed__ attribute close to the struct definition
(to appease gcc-3.4), and use the compiler-independant __packed
form for consistency

apply the patch from PR#24127: make audio_wav_parse_hdr() take u_int *
parameters for encoding, precision, sample rate and channels.

- add a AUDIO_FORMAT_DEFAULT. use it. don't change functionity.

- fix the function selection in sunau conversions.

- be more verbose when given -V. (hi lennart! :)

- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

fix 2001/03/04 change: audio_parse_wav_hdr() returns a signed value.

Don't play trailing garbage at end of RIFF WAVE (.wav) files
Patch sent by YAMAMOTO Takashi <takashi.yamamoto@bigfoot.com> in PR 12328

de-__P()-ify

libaudio:
- rename audio_get_sun_encoding() to audio_sun_to_encoding()
- add audio_encoding_to_sun()
play.c:
- adapt
record.c:
- convert NetBSD encoding to sun encoding if possible, otherwise don't
output a header at all.
fixes PR#10380 from Michael Eriksson <eramore@era-t.ericsson.se>.

branches: 1.4.2;
remove unused crap.

branches: 1.3.8;
make recording precision/encoding setting sane.

add missing copyrights.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Clean up lines containing only tabs.

branches: 1.10.8;
add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.9.8; 1.9.16; 1.9.18;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.7.2; 1.7.8;
ANSIfy. Add static. Add __dead.

remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.5.16; 1.5.18;
missing initializers.

variable signed/unsigned cleanup.

Add NetBSD RCS Ids.

branches: 1.2.2;
sun linear encodings are *signed*

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

audio_wav_parse_hdr: handle zero-length data files again

the previous clean up turns zero-length data into no data and thus
an error, instead of simply doing nothing. noted by gson.

fix some sizeof() confusion.

using "const char search[4]" as a function parameter means that
"search" is actually a pointer type so "sizeof search" returns
8 on 64-bit platforms. i mis-read this and used "sizeof *search"
which is always 1, noted by rillig.

instead of trying to avoid writing "4" twice, put it in a define
and use that in various places instead. annoying.

audioplay(1): handle mis-aligned RIFF chunks.

put the code to find RIFF chunks into a new find_riff_chunk() function,
and handle mis-aligned chunk lengths. can now play files with chunks
that say they are 7 bytes long, and have 1 byte padding.

add some -V -V extra-verbose for the wav parser.

minor clean up in the RIFF/WAVE header checking.

use memcmp() not strncmp() to check for "RIFF" and "WAVE".
avoid endian convesion on part.len 3 times.

be sure to byte swap ext.len when checking/printing it.

noted by rillig@.

audio_wav_parse_hdr(): avoid buffer overreads and clean up

reimplement most of this function using a new method of buffer
management to ensure that we never read beyond the provided size.

properly handle RIFF chunk lengths, instead of assuming various
offsets from most files are right.

update list of consumed documentation and fill the list of WAVE
formats from RFC 2361 (most remain not supported.)

should fix PR#57973.

tested against a large number of .wav files i have handy and with
a testsuite generator that should be incoming soon.

Clean up lines containing only tabs.

Clean up lines containing only spaces.

Add support for recording 24bit wav files.

branches: 1.15.8;
add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.14.4; 1.14.6;
need getle16() for ext.sub_tag too

PR bin/52762 by Yosuke Sugahara

branches: 1.13.8;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

fix unused variable warnings

move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.10.2; 1.10.8;
ANSIfy. Add static. Add __dead.

- add extended WAVE header support
- attempt to play a bunch more WAV files

branches: 1.8.4; 1.8.10;
remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.7.32; 1.7.34;
apply the patch from PR#24127: make audio_wav_parse_hdr() take u_int *
parameters for encoding, precision, sample rate and channels.

Add NetBSD RCS Ids.

Respect the law: Use A-law and mu-law as spellings as far as easily possible.
Inspired by Igor Sobrado in PR 19680.

remove some #if 0 debug code

branches: 1.3.2;
add $NetBSD$

- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Update to correct (minor) misstatememt pointed out on netbsd-users
by Ramiro Aceves.

(Someone with better knowledge of the audio system than I should
review this and correct anything I do not correctly understand.)

branches: 1.25.4;
fix various typos in comments, documentation and messages.
mainly s/paramater/parameter/ and s/reduntant/redundant/.

audioctl.1: describe foibles

branches: 1.23.2;
Sort options (no-argument options before ones demanding arguments).

Bump date.

Remove -p option. AUDIO_SETCHAN is insecure and is obsoleted.

branches: 1.21.12;
Audio now uses fdclone, there is no longer a limitation of one audio
instance per process. Virtual channels are placed in a queue, so there is
no longer a compile time limit of VAUDIOCHANS.

A new sysctl has been added to control multiple user access.
Mixer labels on virtual channels are now prefixed by vchan.

audiobell.c and audioctl have been updated to reflect these changes.

Use of fdclone was posted to tech-kern@ and improvements made.
Multiuser access control changes and the use of a queue were suggessted by
pgoyette@

branches: 1.20.2;
The audio sub-system now supports the following features as
posted to tech-kern:

* Simultaneous playback and mixing of multiple streams
* Playback streams can be of different encoding, frequency, precision
and number of channels
* Simultaneous recording to different formats
* One audio device per process
* Sysctls to set the common format frequency, precision and channels
* Independent mixer controls for recording/playback per stream
* Utilizes little cpu time for multiple streams / good performance
* Compatible with existing programs that use OSS/NetBSD audio
* Changes to audioctl(1) to allow specifying process id for corresponding
audio device

branches: 1.19.28;
Reorder sections to canonical order.

Convert TNF licenses to new 2 clause variant

branches: 1.17.32;
.Nm does not need a dummy argument ("") before punctuation or
for correct formatting of the SYNOPSIS any longer.

New sentence, new line. Use Aq/An.

Use '-d' rather than '-f' for specification of alternative
mixer or audioctl device. The support for '-f' is kept for compatibility
for now.

Sort sections, drop some whitespace.

Xref audioplay(1) and audiorecord(1).

branches: 1.12.2;
Add missing .El line.

Oops; update the date as well.

Update comments.

Rework the examples section a little.

Use audio and mixer devices with unit numbers. Also allow the devices
without unit numbers of a compatibility option is set. Fixes PR 5991.

Chnage the name of the environment variables just introduced.

Add an environment variable for the default audio device.

fix .Nm usage

Fix tipo inherited from old version of TNF copyright template.

Change to use /dev/audioctl.

Add HISTORY section.

Add audio utility programs.
The audio device can be inspected and changed by audioctl,
and the mixer device by mixerctl. The command syntax was
inspired by sysctl.

remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.3.34; 1.3.36;
de-__P()-ify

add missing copyrights.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Remove -p option. AUDIO_SETCHAN is insecure and is obsoleted.

branches: 1.43.12;
Renane ioctl AUDIO_SETPROC to AUDIO_SETCHAN.

Add an ioctl to return channel number (AUDIO_GETCHAN). This can be used
on audio/sound/audioctl devices.

Return EIO in read/write/ioctl/poll/stat if fp has been closed or is
invalid.

Update audio.4, audioio.h and audioctl(1) to reflect these changes.

Audio now uses fdclone, there is no longer a limitation of one audio
instance per process. Virtual channels are placed in a queue, so there is
no longer a compile time limit of VAUDIOCHANS.

A new sysctl has been added to control multiple user access.
Mixer labels on virtual channels are now prefixed by vchan.

audiobell.c and audioctl have been updated to reflect these changes.

Use of fdclone was posted to tech-kern@ and improvements made.
Multiuser access control changes and the use of a queue were suggessted by
pgoyette@

branches: 1.41.2;
The audio sub-system now supports the following features as
posted to tech-kern:

* Simultaneous playback and mixing of multiple streams
* Playback streams can be of different encoding, frequency, precision
and number of channels
* Simultaneous recording to different formats
* One audio device per process
* Sysctls to set the common format frequency, precision and channels
* Independent mixer controls for recording/playback per stream
* Utilizes little cpu time for multiple streams / good performance
* Compatible with existing programs that use OSS/NetBSD audio
* Changes to audioctl(1) to allow specifying process id for corresponding
audio device

branches: 1.40.8;
move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.39.2; 1.39.8;
ANSIfy. Add static. Add __dead.

fix -Wsign-compare and other WARNS=4 issues

branches: 1.37.8;
Remove clause 3 and 4 from TNF licenses

branches: 1.36.16;
use c99 initializer.

netbsd.org->NetBSD.org

Add NetBSD RCS Ids.

compatibility, not compatiblity.

Test for default device in a portable way. Thanks to Michael Ericsson
for spotting the problem.

Use _PATH_SOUND etc, all fall back to _PATH_SOUND0 (etc.) if that fails.

Delete the old code which made audioctl use stderr for output
if the stdout was same as the audio device. It's no longer relevant,
and the check was broken anyway (code used st_dev rather than st_rdev
when comparing the character devices).
A by-product of this is that 'audioctl -a > /dev/null' now works.

Okay'ed by Lennart.

Use '-d' rather than '-f' for specification of alternative
mixer or audioctl device. The support for '-f' is kept for compatibility
for now.

- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

Fix white space and formatting nits.

convert to use getprogname()

fix nested extern

de-__P()-ify

branches: 1.23.10;
audio paths are in <paths.h>

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Make the copyright header conform to the NetBSD template.

Fix thinko in last commit.

Use audio and mixer devices with unit numbers. Also allow the devices
without unit numbers of a compatibility option is set. Fixes PR 5991.

audio_prinfo.seek is now a UINT

Make it possible to do:
audioctl -w play=44100,2,16,slinear

Not the greatest implementation.

__AUDIT__ cleanup.

Fix typo. From Feico Dillema <dillema@acm.org>

Chnage the name of the environment variables just introduced.

Add an environment variable for the default audio device.

Update for SunOS compatibility.

Print properties differently.
Change indentation.

Fix tipo inherited from old version of TNF copyright template.

Get rid of warnings.

Don't try opening the device O_RDWR.

Change to use /dev/audioctl.

Modernize for new audio features.

Sync with audioio.h changes.

Sync with the recent changes of the audio variables.
s/AUDIO_ENCODING_LINEAR_LE/AUDIO_ENCODING_SLINEAR_LE/
s/AUDIO_ENCODING_LINEAR_BE/AUDIO_ENCODING_SLINEAR_BE/

Fix pasto.

Fix some buglets and clean up a little.

Add audio utility programs.
The audio device can be inspected and changed by audioctl,
and the mixer device by mixerctl. The command syntax was
inspired by sysctl.

Enable WARNS=4 by default for usr.bin, except for:
awk bdes checknr compile_et error gss hxtool kgetcred kinit
klist ldd less lex locale login m4 man menuc mk_cmds
mklocale msgc openssl rpcgen rpcinfo sdiff spell ssh
string2key telnet tn3270 verify_krb5_conf xlint

branches: 1.3.14;
Convert to raise_default_signal(3).

branches: 1.2.26;
Rototill play_fd() to fix some original problems and some added later.
Also, fix up for and enable WARNS=2. This fixes problems playing from
pipes or standard input, and playing from funny filesystems or other
cases that can't be mmap(2)ed.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

audioplay: add -n flag that doesn't write audio data.

this will be used in an upcoming testsuite for the wav parser.

update my email address.

audioplay.1: fix sentence punctuation

audioplay(1): fix typo

From Jim Spath in PR/57228

branches: 1.30.6;
audioplay.1: generalize a reference to web browsers

Remove list of audio drivers, moved to audio(4).

Fix a typo. Fix some xrefs, comment out references to non-existing man pages.

add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.26.18; 1.26.26; 1.26.28;
Merge riastradh-drm2 to HEAD.

Use Mt for email addresses.

branches: 1.24.6; 1.24.12; 1.24.16;
Use better reference after sorting in -f description.

Sort some more.

add a new "-B buffersize" flag to both audioplay and audiorecord.
suggested by jmcneill@.

sort audioplay manual some.

remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.20.32; 1.20.34;
audiofiles -> audio files

New sentence, new line. Use Aq/An.

Update man pages to reflect the use of /dev/sound instead of /dev/audio.

Generate <>& symbolically. I'm avoiding .../dist/... directories for now.

Retire -C, do all the ioctls on same audio device we do i/o

Sort SEE ALSO.

- add signed linear support to audiorecord's WAV support.
- swap big endian samples for WAV.
- update audioplay(1) and audiorecord(1) audio driver xrefs.

XXX fix support for unsigned linear samples -- it seems they need
to be converted to signed for WAV???

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

Sort sections, sort SEE ALSO, close quotes in two places.

Whitespace cleanup.

Don't give .Os an argument, not even 'NetBSD' (default includes version).

X68k built-in voice synthesizer supports 4bit ADPCM.

remove extra period in SEE ALL section

branches: 1.7.8;
add a note about "audioctl encodings", to find available supported encodings.

Name possible encodings for "-e" option and add an example provided by
Ben Wong in PR bin/8391. Thanks must go to Ross Harvey for helping me
with the "nroff" stuff.

If not specified on the command line, attempt to fall back to the environment
variables AUDIODEVICE (for -d) and AUDIOCTLDEVICE (for -C, already used in
audioctl(1)) before falling back to _PATH_* constants.

Fix various buglets.

branches: 1.3.2;
document new play -f/-s/-P-/c/-e, s/name/named/ from mason in pr#7292. update dates.

add -f (force) flag to audioplay, plus -s/-P/-c/-e switches for force setting sample rate, precision, channels and encoding. fix some man pages missing .El bits (via veego)

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

audioplay: add -n flag that doesn't write audio data.

this will be used in an upcoming testsuite for the wav parser.

Revert accidentially committed part. The 'latency' report isn't ready
to be useful.

Add support for recording 24bit wav files.

branches: 1.61.2;
fix verbose log for 64 bit float

Spell the largest 32bit signed value as INT32_MAX. Explicitly cast to
float, because it is not precisely representable.

copyright maint.

add support for playing IEEE float32 and float64 RIFF WAVE
samples on platforms that have these types natively, and
can handle signed linear 32 bit samples. explicitly
disabled on vax, run-or-compile-time sizeof() check
disabled for everyone else

now i can play a float32 .wav file i found.
float64 not tested.

copyright maint, update HISTORY, update audio drivers list.

branches: 1.57.2;
Use err(3)/errx(3) properly to avoid "write failed: Undefined error: 0".

The test for regular file was wrong.

From yarl on freenode, thanks!

branches: 1.55.8; 1.55.14; 1.55.16;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

ANSIfy. Add static. Add __dead.

Sort usage.

allow for buffer sizes < 32k

add a new "-B buffersize" flag to both audioplay and audiorecord.
suggested by jmcneill@.

sort audioplay manual some.

fix -Wsign-compare and other WARNS=4 issues

branches: 1.49.6;
remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.48.4; 1.48.6;
convert __attribute__s to applicable cdefs.h macros

Convert to raise_default_signal(3).

branches: 1.46.10;
Coverity CID 3187: Fix memory leak

branches: 1.45.2;
play.c: allow empty files to exit gracefully rather than with an error
of "header seems really large". reported by wiz.

record.c: don't use total_size in the header if is -1 (it's initial value)

appease gcc -Wuninitialized

branches: 1.43.2;
There is no -m option -- remove it from the usage.
From Kouichirou Hiratsuka in PR 26192.

branches: 1.42.4;
Add NetBSD RCS Ids.

fix an error message

fix PR#18909, but not quite exactly the same as the patch provided. also,
check if the filesize causes a size_t overflow and if so, use the fallback
read/write loop method.

Move close so that last ioctl is before close call.
Fix duplicate word in comment.
Based on patch from Christian Biere in PR 18902, with a style nit fixed.

print the file name in verbose mode as well. avoid printing (null) for encoding.

sweep of errx/warnx, remove unnecessary trailing \n

define: datasize == 0 means use the whole file. fixes part of PR#12809.

Test for default device in a portable way. Thanks to Michael Ericsson
for spotting the problem.

Retire -C, do all the ioctls on same audio device we do i/o

Actually be verbose if the -V flag is given.

Use _PATH_SOUND etc, all fall back to _PATH_SOUND0 (etc.) if that fails.

KNF.

flush the audio buffer on signals; from Ben Wong <hackerb9@u.washington.edu> in PR#9134

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

fix regression, err() -> errx()

Rototill play_fd() to fix some original problems and some added later.
Also, fix up for and enable WARNS=2. This fixes problems playing from
pipes or standard input, and playing from funny filesystems or other
cases that can't be mmap(2)ed.

X68k built-in voice synthesizer supports 4bit ADPCM.

Fix white space and formatting nits.

lp64/printf

Don't play trailing garbage at end of RIFF WAVE (.wav) files
Patch sent by YAMAMOTO Takashi <takashi.yamamoto@bigfoot.com> in PR 12328

convert to use getprogname()

fix nested extern

- move the mmap playing code into new play() function
- be more (properly) quiet with `-q'

de-__P()-ify

libaudio:
- rename audio_get_sun_encoding() to audio_sun_to_encoding()
- add audio_encoding_to_sun()
play.c:
- adapt
record.c:
- convert NetBSD encoding to sun encoding if possible, otherwise don't
output a header at all.
fixes PR#10380 from Michael Eriksson <eramore@era-t.ericsson.se>.

branches: 1.17.4;
honour -q (quiet) flag a bit more.

When we have to print usage instructions, don't indicate success in the exit
status.

* Don't treat failure to open a file as fatal error; print a warning to
standard error, arrange for a non-zero exit status and continue with the
next file. Addresses PR bin/8748.
* GCC vs. ANSI: a cast doesn't yield an lvalue.
* Use MAP_FAILED.

fix bug introduced in 1.12:
only return a header length of 0 if we are using -f (force), not
in the case when we know how long the header is. this stops from
playing the audio header.

honour -f, even if we appear to have a sun audio file. un-byte-screw the encoding for printing, too.

branches: 1.12.2;
Fix broken "-f" and "-P" options. Patches supplied by Ben Wong in
PR bin/8390.

If not specified on the command line, attempt to fall back to the environment
variables AUDIODEVICE (for -d) and AUDIOCTLDEVICE (for -C, already used in
audioctl(1)) before falling back to _PATH_* constants.

fix defect where specifying channels would also specify the same string
as the control device (missing 'break')

try using normal read/write is mmap() fails. tell the VM system we are going to be doing sequential access via madvise().

Fix various buglets.

branches: 1.7.2;
Insert an AUDIO_DRAIN at the end of each file.

add -f (force) flag to audioplay, plus -s/-P/-c/-e switches for force setting sample rate, precision, channels and encoding. fix some man pages missing .El bits (via veego)

expand usage message to reality.

add support to play stdin.

add missing copyrights.

audio paths are in <paths.h>

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Convert to raise_default_signal(3).

branches: 1.1.36;
add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

update my email address.

branches: 1.44.6;
Add an example for recording CD quality audio

audioamd is in the SPARC port

branches: 1.42.18; 1.42.26; 1.42.28;
Merge riastradh-drm2 to HEAD.

Use Mt for email addresses.

branches: 1.40.6; 1.40.12; 1.40.16;
Sort some more.

add a new "-B buffersize" flag to both audioplay and audiorecord.
suggested by jmcneill@.

sort audioplay manual some.

note that -t can take [.dddddd]. sort -q.
from dhgutteridge@sympatico.ca in PR#36928.

remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.36.20; 1.36.22;
Recording to standard output is specified by a file name of -,
not by omitting the file name.

Add -q to SYNOPSIS. From Kouichirou Hiratsuka in PR 26193.

Add arguments to options in long description.

minor consistency nit

.Nm does not need a dummy argument ("") before punctuation or
for correct formatting of the SYNOPSIS any longer.

Be consistent about putting commas between items in an enumeration.

fix the description for "-p" to match reality, new text provided in PR#18949.

New sentence, new line. Sort options. Bump Dd for .wav-file handling.
Some nits.

if the output file is called *.wav, output a WAV file.

+ rephrase a bunch of stuff
+ fix totally bogus -i text

Update man pages to reflect the use of /dev/sound instead of /dev/audio.

Be more verbose in description of -e.

Generate <>& symbolically. I'm avoiding .../dist/... directories for now.

Retire -C, do all the ioctls on same audio device we do i/o

update .Dd

Fix typos.

note about WAV sample conversions.

add a note about RIFF/WAVE support and one about signed/unsigned big/little conversion.

Sort SEE ALSO.

- add signed linear support to audiorecord's WAV support.
- swap big endian samples for WAV.
- update audioplay(1) and audiorecord(1) audio driver xrefs.

XXX fix support for unsigned linear samples -- it seems they need
to be converted to signed for WAV???

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

Sort sections, sort SEE ALSO, drop trailing '.' in SEE ALSO, use standard
headers.

Don't give .Os an argument, not even 'NetBSD' (default includes version).

Explain '-c' option.

X68k built-in voice synthesizer supports 4bit ADPCM.

cleanup. including proper fix for previous

add missing chapter number from ym(4) crossreference

Reported by Tetsuya Isaki in bin/12513

remove extra period

branches: 1.8.4;
When no -e flag is used to specify the encoding, default to "ulaw". This
fixes PR bin/9329. Thanks to Wolfgang Rupprecht (wolfgang@wsrcc.com)
for filing the PR and identifying the bug. This change is slightly different
than the one provided in the PR.

Also, document the default encoding in the manpage.

This default is consistent with the default for audiorecord under solaris.

add description of the -t option

If not specified on the command line, attempt to fall back to the environment
variables AUDIODEVICE (for -d) and AUDIOCTLDEVICE (for -C, already used in
audioctl(1)) before falling back to _PATH_* constants.

reword description to not be audioplay, as noted by pr#7379.

branches: 1.4.2;
document new play -f/-s/-P-/c/-e, s/name/named/ from mason in pr#7292. update dates.

add -f (force) flag to audioplay, plus -s/-P/-c/-e switches for force setting sample rate, precision, channels and encoding. fix some man pages missing .El bits (via veego)

add missing copyrights.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..

Fix various typos, mainly in comments.

match message & conversion function name properly.

probably should make this into an array with two members.

Clean up lines containing only spaces.

Add support for recording 24bit wav files.

branches: 1.56.2;
Let user gracefully abort recording with SIGINT.

audiorecord(1): Handle read(2) return value gracefully.

branches: 1.54.8; 1.54.18;
rename struct write_info as struct track_info, and the variable name 'wi'
as 'ti', as this info is useful for reading as well.

move most of the header writing code from record.c into wav.c and sun.c.

introduce a struct write_info for communication between front ends and
backends.

libaudio API is still pretty ugly, and needs to stay local here for now,
but it is a little more usable for other tools now.

branches: 1.52.2; 1.52.8;
PR/45384: Henning Petersen: Swapped arguments in lseek

ANSIfy. Add static. Add __dead.

Sort usage.

allow for buffer sizes < 32k

add a new "-B buffersize" flag to both audioplay and audiorecord.
suggested by jmcneill@.

sort audioplay manual some.

fix -Wsign-compare and other WARNS=4 issues

branches: 1.46.6;
remove clause #3 from my license where there are no other
copyright holders involved.

branches: 1.45.4; 1.45.6;
convert __attribute__s to applicable cdefs.h macros

Convert to raise_default_signal(3).

branches: 1.43.12;
variable signed/unsigned cleanup.

Do not treat all file names of length 1 as meaning standard output.

play.c: allow empty files to exit gracefully rather than with an error
of "header seems really large". reported by wiz.

record.c: don't use total_size in the header if is -1 (it's initial value)

move some more processing before opening the audio device. don't
default the encoding to ulaw; default it to what ever the device
has. the latter fixes PR#30612.

check that an output file is given before doing any other work.
avoids touching the audio device and then printing an error.

branches: 1.38.2;
Convert 8-bit signed linear data to unsigned for .wav files.

Fix fencepost errors in the conversion functions (making them a bit more like
the kernel versions).

branches: 1.36.2;
Don't set an int to NULL.

Sort usage, and move a word.

fix a now-wrong comment.

Add NetBSD RCS Ids.

- decode_int(): always use base 10, so that a time of "1:09" works.
- decode_time(): 60 seconds in a minute, really.
both from tsarna@netbsd.org.

- be verbose about the time we are going to record for, if verbose.

if the output file is called *.wav, output a WAV file.

sweep of errx/warnx, remove unnecessary trailing \n

fix an error message.

_PATH_SOUND0 should be opened O_RDONLY (obvious pasto).

- add a AUDIO_FORMAT_DEFAULT. use it. don't change functionity.

- fix the function selection in sunau conversions.

- be more verbose when given -V. (hi lennart! :)

apply the same fix as the previous to WAV output creation.

convert the host (non _LE or _BE) encodings we can, as well. fixes part
of PR#12809...

Test for default device in a portable way. Thanks to Michael Ericsson
for spotting the problem.

Retire -C, do all the ioctls on same audio device we do i/o

Use _PATH_SOUND etc, all fall back to _PATH_SOUND0 (etc.) if that fails.

- WARNS=3 (from ross@netbsd.org)
- fix a bug in WAV conversions

WAV seems to be signed only; convert unsigned samples.

for sun output files, if we have unsigned or little endian data, convert
it to signed big endian data. pull in a modified copy of the kernels'
auconv.[ch] to help with this (and remove some home grown code that did
some of the same things.)

- add signed linear support to audiorecord's WAV support.
- swap big endian samples for WAV.
- update audioplay(1) and audiorecord(1) audio driver xrefs.

XXX fix support for unsigned linear samples -- it seems they need
to be converted to signed for WAV???

- add rudimentary "wav" recording support: don't think it copes with
big-endian systems properly yet -- need to add RIFX support or swap
samples while writing. more on this coming soon.

- new "-F <format>" option for audiorecord. supported values of <format>
are "sun", "wav" and "none", plus a few aliases.

- split sun & wav support out into their own files.

- don't assume that WAV header elements are properly aligned for the
type of data they hold.

XXX: still need to clean up libaudio interfaces a lot.

only warn about not writing the sun header once. PR#12809

X68k built-in voice synthesizer supports 4bit ADPCM.

convert to use getprogname()

fix nested extern

de-__P()-ify

libaudio:
- rename audio_get_sun_encoding() to audio_sun_to_encoding()
- add audio_encoding_to_sun()
play.c:
- adapt
record.c:
- convert NetBSD encoding to sun encoding if possible, otherwise don't
output a header at all.
fixes PR#10380 from Michael Eriksson <eramore@era-t.ericsson.se>.

branches: 1.10.4;
When no -e flag is used to specify the encoding, default to "ulaw". This
fixes PR bin/9329. Thanks to Wolfgang Rupprecht (wolfgang@wsrcc.com)
for filing the PR and identifying the bug. This change is slightly different
than the one provided in the PR.

Also, document the default encoding in the manpage.

This default is consistent with the default for audiorecord under solaris.

When we have to print usage instructions, don't indicate success in the exit
status.

run forever if no -t is specified, from dan mcmahill

branches: 1.7.2;
fix -t option

If not specified on the command line, attempt to fall back to the environment
variables AUDIODEVICE (for -d) and AUDIOCTLDEVICE (for -C, already used in
audioctl(1)) before falling back to _PATH_* constants.

Fix -C behavior.

branches: 1.4.2;
expand usage message to reality.

make recording precision/encoding setting sane.

add missing copyrights.

add audiorecord and audioplay tools. audioplay does .au and .wav files. audiorecord makes .au files only. uses parts of audioctl..