OSS4 won't install properly

[keantoken]

Hello all.

Whenever I use Gdebi to install OSS4, the installation hangs at the step which says "Starting Open Sound System". However I get sound from osstest.

When I try to fix dpkg from complaining, I get this:

Code: Select all

keantoken@ubuntu:~$ sudo dpkg --configure -a
Processing triggers for menu ...
Setting up oss-linux (4.2-2004) ...
Building OSS Modules for Linux-i686 2.6.39-2.dmz.1-liquorix-686

OSS build environment set up for REGPARM kernels

Building module osscore
Building module lynxone
Building module lynxtwo
Building module oss_ali5455
Building module oss_atiaudio
Building module oss_audigyls
Building module oss_audioloop
Building module oss_audiopci
Building module oss_cmi878x
Building module oss_cmpci
Building module oss_cs4281
Building module oss_cs461x
Building module oss_digi96
Building module oss_emu10k1x
Building module oss_envy24
Building module oss_envy24ht
Building module oss_fmedia
Building module oss_geode
Building module oss_hdaudio
Building module oss_hdsp
Building module oss_ich
Building module oss_imux
Building module oss_madi
Building module oss_midiloop
Building module oss_midimix
Building module oss_sblive
Building module oss_sbpci
Building module oss_sbxfi
Building module oss_solo
Building module oss_trident
Building module oss_usb
Building module oss_userdev
Building module oss_via823x
Building module oss_via97
Building module oss_ymf7xx
depmod -a
Forcing re-detection of installed soundcards
Starting Open Sound System


And it still hangs. Until I fix this error, I can't do anything with Synaptic, Aptitude or dpkg.

It appears OSS has installed fine, but for some reason the last step hangs.

Please help,
- keantoken

[keantoken]

Okay, I went back to the generic kernel and it didn't hang this time. Audacious still gives me these errors though:

Code: Select all

keantoken@ubuntu:~$ audacious

** (audacious:5669): WARNING **: SNDCTL_DSP_SETFMT ioctl failed: Operation not permitted

** (audacious:5669): WARNING **: SNDCTL_DSP_SETFMT ioctl failed: Operation not permitted

** (audacious:5669): CRITICAL **: drain: assertion `cop->drain != NULL' failed

** (audacious:5669): WARNING **: SNDCTL_DSP_SETFMT ioctl failed: Operation not permitted

** (audacious:5669): WARNING **: SNDCTL_DSP_SETFMT ioctl failed: Operation not permitted

** (audacious:5669): CRITICAL **: drain: assertion `cop->drain != NULL' failed

** (audacious:5669): WARNING **: SNDCTL_DSP_SETFMT ioctl failed: Operation not permitted

** (audacious:5669): CRITICAL **: reset_time: assertion `cop->set_written_time != NULL' failed

Google searches suggest this could be a permissions problem for the /dev/dsp or whatever thingies.

- keantoken

[keantoken]

Sound works in Audacity, so maybe it's an Audacious issue...

[keantoken]

I was using Synaptic and I discovered I hadn't uninstalled the OSS4-dev package, so I uninstalled it along with the other things I was installing. I saw this in the terminal and was quick enough to copy it:

Code: Select all

(Reading database ... 284780 files and directories currently installed.)
Removing oss4-dev ...
Removing 'diversion of /usr/include/linux/soundcard.h to /usr/include/linux/soundcard.h.oss3 by oss4-dev'
Purging configuration files for oss4-dev ...
No diversion 'diversion of /usr/include/linux/soundcard.h by oss4-dev', none removed.
Selecting previously deselected package alsaplayer-oss.
(Reading database ... 284776 files and directories currently installed.)
Unpacking alsaplayer-oss (from .../alsaplayer-oss_0.99.80-5build1_i386.deb) ...
Selecting previously deselected package alsaplayer-gtk.
Unpacking alsaplayer-gtk (from .../alsaplayer-gtk_0.99.80-5build1_i386.deb) ...
Selecting previously deselected package alsaplayer-common.
Unpacking alsaplayer-common (from .../alsaplayer-common_0.99.80-5build1_i386.deb) ...
Selecting previously deselected package liboss-salsa2.
Unpacking liboss-salsa2 (from .../liboss-salsa2_4.1-build1052b-2_i386.deb) ...
Selecting previously deselected package oss-preserve.
Unpacking oss-preserve (from .../oss-preserve_1.1-6_i386.deb) ...
Processing triggers for menu ...
Processing triggers for man-db ...
Processing triggers for bamfdaemon ...
Rebuilding /usr/share/applications/bamf.index...
Processing triggers for desktop-file-utils ...
Processing triggers for python-gmenu ...
Rebuilding /usr/share/applications/desktop.en_US.utf8.cache...
Processing triggers for ureadahead ...
ureadahead will be reprofiled on next reboot
Processing triggers for python-support ...

- keantoken

[keantoken]

Okay, I discovered in a roundabout way that soundon hangs. I got this peculiar message:

Code: Select all

keantoken@ubuntu:~$ sudo soundon
Previous start of OSS crashed the system
Please resolve the situation and remove file
"/usr/lib/oss/starting". Then start OSS by
running soundon again.

Every time I run soundon, it leaves an uninterruptible instance of something called "sync" in the system monitor which won't die.

[igorzwx]

I was using Synaptic and I discovered I hadn't uninstalled the OSS4-dev package, so I uninstalled it along with the other things I was installing. I saw this in the terminal and was quick enough to copy it:

If you feel very experimental, you may try this:
viewtopic.php?f=3&t=3933#p17316
and this:
viewtopic.php?f=3&t=3543&start=30#p15711

[keantoken]

Igor, I am very experimental, in fact none of my experiments has ever concluded! Kubuntu is my "virgin" distro, but then I moved on to UbuntuStudio. Unfortunately I don't like to have to run over to the help forums every time I install a program because it doesn't work for some trivial reason.

Archlinux sounds like something I will want to try, and hopefully I won't need a college course in Linux. I do eventually want to learn Assembly programming, as it is the Way, if you can manage it. But this will be an undertaking in Faith, as I have yet to learn any other programming language.

Is it possible to fix this error?

- keantoken

[keantoken]

Okay, I figured out that Audacious wouldn't work because OSS doesn't seem to support 24-bit audio. I switched to 16 bit and it works. Underwhelming.

I tried to set my soundcard to the best rate it supports, which is 96KHz. OSS doesn't support that either. Doubly underwhelming.

Am I doing something wrong? I don't see why compile for high-performance mode if these two basic things aren't supported.

- keantoken

[igorzwx]

Okay, I figured out that Audacious wouldn't work because OSS doesn't seem to support 24-bit audio. I switched to 16 bit and it works. Underwhelming.

I tried to set my soundcard to the best rate it supports, which is 96KHz. OSS doesn't support that either. Doubly underwhelming.

Am I doing something wrong? I don't see why compile for high-performance mode if these two basic things aren't supported.

- keantoken

Yes, you are doing something fundamentally wrong.

1. OSS4 does not support those Ubuntu users who do not want to read the official manuals.
2. The manual is here: https://wiki.archlinux.org/index.php/OSS
3. OSS4 does support "high definition audio", but this depends on your soundcard.
4. To find out which formats are supported for your soundcard, you can use this command:

Code: Select all

$ ossinfo -v9

Changing the Sample Rate
Changing the output sample rate is not obvious at first. Sample rates can only be changed by the superuser and vmix must be unused by any programs when a change is requested. Before you follow any of these steps, ensure you are going through a receiver/amplifier and using quality speakers and not simply computer speakers. If you are only using computer speakers, don't bother changing anything here as you won't notice a difference.
By default the sample rate is 48000hz. There are several conditions in which you may want to change this. This all depends on your usage patterns. You want the sample rate you are using to match the media you use the most. If your computer has to change the sampling rate of the media to suit the hardware it is likely, though not guaranteed that you will have a loss in audio quality. This is most noticable in downsampling (ie. 96000hz → 48000hz). There is an article about this issue in "Stereophile" which was discussed on Apple's "CoreAudio API" mailing list if you wish to learn more about this issue.
Some example sample rates:
44100hz - Sample rate of standard Red Book audio cds.
88000hz - Sample rate of SACD high definition audio discs/downloads. It is rare that your motherboard will support this sample rate.
96000hz - Sample rate of most high definition audio downloads. If your motherboard is an AC'97 motherboard, this is likely to be your highest bitrate.
192000hz - Sample rate of BluRay, and some (very few) high definition downloads. Support for external audio reciever equipment is limited to high end audio. Not all motherboards support this. An example of a motherboard chipset that would support this includes Intel HDA audio.
Read more: https://wiki.archlinux.org/index.php/OS ... ample_Rate

You may better try to learn OSS4 with Arch LiveCD viewtopic.php?f=3&t=3933#p17316

For testing sound quality, you may better use ossplay, read this: viewtopic.php?f=3&t=3543&start=45#p15974
Audacious does not seem to be a good player for OSS4.
You may try this one: http://deadbeef.sourceforge.net/
There is also UMPlayer

Code: Select all

$ yaourt UMPlayer
1 aur/umplayer 0.95.2.1-2 (111)
    The multimedia player that fills all your needs.
2 aur/umplayer-svn 172-1 [installed] (33)
    UMPlayer is the multimedia player that fills all your needs.
==> Enter n° of packages to be installed (ex: 1 2 3 or 1-3)
==> -------------------------------------------------------
==>

[keantoken]

My card is a Creative X-Fi Extreme Audio, an underwhelming card itself, but it does support 24-bit audio and 96KHz. I did read that part of the manual, and OSS reports 48KHz is the only samplerate, and 16 bits as the only bit depth.

In the end I'm not so sure good computer sound is strongly related to samplerate and bit depth. I think it has more to do with resampling quality, driver quality and whether all the bits make it through the software intact. From what I've read many developers can't avoid simple rounding errors. I see many people liking the sound quality of OSS but I would like to see an actual bit comparison with the other sound systems. I can hear the sound is different but I am not sure if it is a good difference or not.

Thanks for recommending some sound players. Since most of my music is CD quality, the soundcard settings aren't a big issue. I really must find a way to better coexist with Linux, without being a parasite to support forums.

Thanks for your help, let me go find something else to bug you with,
- keantoken

[igorzwx]

My card is a Creative X-Fi Extreme Audio, an underwhelming card itself, but it does support 24-bit audio and 96KHz. I did read that part of the manual, and OSS reports 48KHz is the only samplerate, and 16 bits as the only bit depth.

In the end I'm not so sure good computer sound is strongly related to samplerate and bit depth. I think it has more to do with resampling quality, driver quality and whether all the bits make it through the software intact. From what I've read many developers can't avoid simple rounding errors. I see many people liking the sound quality of OSS but I would like to see an actual bit comparison with the other sound systems. I can hear the sound is different but I am not sure if it is a good difference or not.

Thanks for recommending some sound players. Since most of my music is CD quality, the soundcard settings aren't a big issue. I really must find a way to better coexist with Linux, without being a parasite to support forums.

Thanks for your help, let me go find something else to bug you with,
- keantoken

The difference between OSS4 and ALSA is that you can disable resampling with OSS4 and play your HD wave directly to the proper sound device.
In this case, the sound quality depends only on your soundcard and OSS4 drivers.
Moreover, it is very easy to disable resampling with OSS4.

For playing CD formats, e.g. FLAC/APE + CUE, I am using DeadBeef player with a special OSS4 plugin made by Prof. Sergey Petrov.
It has the so-called "exact scientific resampler" inside, and, therefore, the sound quality is perfect.
This plugin is free (you can get it from the author).
Alternatively, you can use the standard OSS4 plugin for Deadbeef (they told that they need OSS4 testers).
The "Production quality resampler" of OSS4 is relatively good for upsampling, but it can produce distortions during downsampling.
You can also try Petrov's pcm_conv for resampling and format conversion.
A very old version of pcm_conv is here viewtopic.php?f=3&t=3543&start=45#p15974
The newest version can be obtained from the author (it is free).

USB audio devices are problematic with OSS4.
Perhaps, you may also want to read this: http://www.tweakheadz.com/soundcards_fo ... studio.htm

Most soundcards on the market have acceptable quality DACs these days, compared to a few years ago when you had to spend premium dollars to get good DACs. It makes sense if you think of your CD player when thinking of the quality of DACs. Early cd players had a harsh sound to them and were often considered "brittle". As a result of that criticism, audio engineers found ways to make the DACs sound better, with oversampling, error correction, interpolation and other ways to "smooth" the output. http://www.tweakheadz.com/soundcards_fo ... studio.htm

[keantoken]

Thanks.

Google is not very forthcoming on how to contact Petrov. As far as I know you're the only one who has ever talked about using Petrov's OSS4 plugin! Do I need a passphrase or something?

I was previously using Pulse with Speex resampling level 10, and that seemed to work better than Sinc. I downloaded DeadBeef, but the resampler plugin still has the default Sinc interpolation. Once I get this all worked out I will reinstall Pulse to use OSS and compare.

- keantoken

[igorzwx]

Thanks.

Google is not very forthcoming on how to contact Petrov. As far as I know you're the only one who has ever talked about using Petrov's OSS4 plugin! Do I need a passphrase or something?

I was previously using Pulse with Speex resampling level 10, and that seemed to work better than Sinc. I downloaded DeadBeef, but the resampler plugin still has the default Sinc interpolation. Once I get this all worked out I will reinstall Pulse to use OSS and compare.

- keantoken

I do not think that Sergey has time to explain you how to solve very simple problems with Linux. He seems to be very busy now. On the other hand, if you feel comfortable with PulseAudio, it is not likely that you can hear the difference in sound quality.

The standard solution for playing CD format, FLAC/APE and mp3:
1. disable resampling in DeadBeef,
2. enable "Production quality" resampler in vmix,
3. set the proper sample rate in vmix https://wiki.archlinux.org/index.php/OS ... ample_Rate

This standard solution should be sufficient for you, because it is very unlikely that you are able to hear the difference between "Production quality" and Petrov's resampler. Although, of course, you can try it too, it is here: viewtopic.php?f=3&t=3543&start=45#p15974
There is also a plugin for ALSA with a special configuration tool http://forum.ubuntu.ru/index.php?topic=112480.0

[keantoken]

Igor, I can't say I feel comfortable with with any sound, although designing my own speaker amplifier has helped with that. Analog equipment can also be blamed for "sterile sound". Even if I can't hear the difference, it would still be nice to have accurate sound in order to do sensitive tests on my designs.

Wait, should I feel stupid? Can you send me the plugin if I PM you my email address?

- keantoken

[igorzwx]

Can you send me the plugin if I PM you my email address?

I am not going to distribute this particular software.
It is free, but it is not "open source". The "open-source" enthusiasts proved to be very intolerant. I do not want to create "ideological problems". We had already "theological discussions" of the sort.

Do you hate "terminal"? If not, you may try this: viewtopic.php?f=3&t=3543&start=45#p15974