Need facts>OSS: podcast w. music recording output AND input?

OSS related discussion in general (i.e. not limited to Linux/Solaris/BSD)

Moderators: cesium, dev, kodachi, hannu

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby cesium » Sun Dec 27, 2009 11:03 pm

Sorry for late response, was busy lately.. I understand nothing is recorded at all? Can you paste 'ossmix' output?
cesium
Developer
 
Posts: 902
Joined: Sun Aug 12, 2007 12:51 am

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby yvon » Tue Dec 29, 2009 10:50 pm

TOP NOTE: I am very much appreciative of how useful the forum and the best help possible is.
Great way to end a study year and start a new one!)))

RE: OSS emulating software mixing.


I took a day to think though I'm still not skilled/understanding how I can 'try' to emulate the mix ability of a
hardware mixing card through software mixing alone in OSS4.2.
Clearly a lack of how things flow and moreso what actual lines to type/alter.

As I move an external USB boot drive to different machines I will detail the machine I'm on as a good choice.
At the bottom here is the oss -v though hoping to grasp an understanding as would like to do
software mixing in hda-intel and a couple of other machines.



For this machine the configs of interest may be:
osscore.conf
oss_ich.conf
oss_audioloop.conf
oss_userdev.conf
soundon.user

The following are many of the things I think about that may work:

[stop vmix from auto-attaching]
In osscore do this then restart oss so that vmixctl won't give error: vmix_no_autoattach=1

[attempt to understand/figure-out best usage of vmixctl with various machines]
Add an appropriate line to soundon.user after using a line like this and confirming it does what is needed:
Code: Select all
vmixctl attach /dev/oss/oss_hdaudio0/pcm0 /dev/oss/oss_hdaudio0/pcmin0 


(the machine I am now on is ich0 not hdaudio - yet there appears no uniquely named input. Maybe need to run vmixctl?
see output at bottom, thank you)

NOTE: I do not yet know how 'practically' vmixctl works with audioloop and audiodev though I will now read this entire post over again from the beginning.

I would like to continue on this subject after finding out how to invoke vmixctl, specifically to understand.
(which I would then write into soundon.user)
and also what is needed to try as far as audioloop and userdev.

The first concern is as to what on this machine I would do to get the right vmixctl line working.
Currently vmix is set to not auto-attach. This I believe is necessary to be able to link things together differently.

I have also read about changing default output in conjuntion with audioloops etc. I think it was.

- - - - - - - - - - - - - ossinfo -v WITHOUT any vmix / vmixctl auto-attached - - - - - - - - - - - - - - - - -

Number of audio devices: 2
Number of audio engines: 7
Number of MIDI devices: 0
Number of mixer devices: 2


Device objects
0: osscore0 OSS core services
1: oss_via823x0 VIA VT8235
2: oss_usb0 USB audio core services
3: oss_imux0 OSS IMUX driver

MIDI devices (/dev/midi*)

Mixer devices
0: VIA823x AC97 Mixer (VT1612A) (Mixer 0 of device object 1)
Device file /dev/oss/oss_via823x0/mix0, Legacy device /dev/mixer3
Priority: 10
Caps:
1: IMUX Control panel (Mixer 0 of device object 3)
Device file /dev/oss/oss_imux0/mix0, Legacy device /dev/mixer1
Priority: 0
Caps:

Audio devices
VIA VT8235 /dev/oss/oss_via823x0/pcm0 (device index 0)
Legacy device /dev/dsp7
Caps: DUPLEX TRIGGER MMAP
Modes: IN/OUT
Engine 1: Available for use
Engine 2: Available for use
IMux0 audio record /dev/oss/oss_imux0/pcmin0 (device index 1)
Legacy device /dev/dsp1
Caps: TRIGGER VIRTUAL
Modes: INPUT
In engine 1: Available for use
In engine 2: Available for use
In engine 3: Available for use
In engine 4: Available for use
In engine 5: Available for use

Nodes
/dev/dsp -> /dev/oss/oss_via823x0/pcm0
/dev/dsp_in -> /dev/oss/oss_ich0/pcm0
/dev/dsp_out -> /dev/oss/oss_ich0/pcm0
/dev/dsp_ac3 -> /dev/oss/oss_ich0/pcm0
/dev/dsp_mmap -> /dev/oss/oss_ich0/pcm0
/dev/dsp_multich -> /dev/oss/oss_ich0/pcm0
yvon
Member
 
Posts: 18
Joined: Sat Dec 19, 2009 2:09 am

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby cesium » Thu Dec 31, 2009 1:47 am

Well, there's a vmixctl manpage which explain a bit. It's mainly used when default output is changed, and the new output needs software mixing. Stopping autoattach isn't necessary for vmixctl to run, but is sometimes useful (to unclutter mixer if default is unused or to clear the input node to allow a vmix with a different output node set to attach to it).

Secondly, some devices have duplex nodes which do both input and output (oss_ich is one of those), so your case is actually fine. Best way to see this is to run "ossinfo -v3". You'll see under each node a "Modes" line. If it has "INPUT" or "IN/OUT" written, than this means node does input or (respectively) input and output.

Lastly, in most cases the default output is fine, and you don't have to mess with vmixctl to get the software mixing.
cesium
Developer
 
Posts: 902
Joined: Sun Aug 12, 2007 12:51 am

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby igorzwx » Wed Jan 20, 2010 6:49 pm

yvon wrote:Hi!
USING ONLY OSS (latest mercurial install) - NO ALSA

I need to be able to have a skype-oss or voip (oss) conversation and while playing a music track INTO the conversation record the whole thing as as podcast.


It is very easy to do, if you have ancient computers with ancient sound cards, such as:

00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 01)

00:11.5 Multimedia audio controller: VIA Technologies, Inc. VT8233/A/8235/8237 AC97 Audio Controller (rev 30

The sound quality is sufficient for the purpose, provided the levels are set correctly.

The method:

1. I select "vol" as recording source in the mixer
2. then I start Audacity and press "record" button (/dev/dsp)
3. then I launch Skype and call to a friend
4. then I run a player and play music

We listen to music and discuss it.
And Audacity records all these.

These are my mixer settings:

$ ossmix
Selected mixer 0/ICH AC97 Mixer (AD1981B)
Known controls are:
vol [<leftvol>:<rightvol>] (currently 56:56)
vol.rec ON|OFF (currently ON)
pcm [<leftvol>:<rightvol>] (currently 57:57)
speaker <monovol> (currently 0)
line [<leftvol>:<rightvol>] (currently 0:0)
line.rec ON|OFF (currently OFF)
mic <monovol> (currently 38)
mic.rec ON|OFF (currently OFF)
cd [<leftvol>:<rightvol>] (currently 0:0)
cd.rec ON|OFF (currently OFF)
igain [<leftvol>:<rightvol>] (currently 31:31)
aux1 [<leftvol>:<rightvol>] (currently 0:0)
aux1.rec ON|OFF (currently OFF)
phone [<leftvol>:<rightvol>] (currently 0:0)
phone.rec ON|OFF (currently OFF)
mono <monovol> (currently 0)
mono.rec ON|OFF (currently OFF)
video [<leftvol>:<rightvol>] (currently 0:0)
video.rec ON|OFF (currently OFF)
spdout.enable ON|OFF (currently OFF)
spdout.adc/dac ON|OFF (currently OFF)
spdout.pro <Consumer|Professional> (currently Consumer)
spdout.audio <AUDIO|DATA> (currently AUDIO)
spdout.copy ON|OFF (currently OFF)
spdout.pre-emph ON|OFF (currently OFF)
spdout.rate <48000|44100|32000> (currently 48000)
spdout.vbit ON|OFF (currently OFF)
vmix0-enable ON|OFF (currently ON)
vmix0-rate <decimal value> (currently 48000) (Read-only)
vmix0-channels <Stereo|Multich> (currently Stereo)
vmix0-src <Fast|High|High+|Production|OFF> (currently Production)
vmix0-outvol <monovol> (currently 23.4 dB)
vmix0-invol <monovol> (currently 25.0 dB)
vmix0.pcm1 [<leftvol>:<rightvol>] (currently 25.0:25.0 dB)
vmix0.pcm2 [<leftvol>:<rightvol>] (currently 25.0:25.0 dB)
vmix0.pcm3 [<leftvol>:<rightvol>] (currently 25.0:25.0 dB)
vmix0.pcm4 [<leftvol>:<rightvol>] (currently 25.0:25.0 dB)

For both cards, the settings are the same.

This method works well with ancient onboard sound cards produced in 2001-2003.
However, this method may not with modern HDA soundcards.
Perhaps, it is a kind of "global technological regress" in digital sound industry.

--------------------------------------------------------------------------------------------------------------
EDIT:
You can can do exactly the same with Windows XP and Intel HDA soundcards, such as:

00:1b.0 Audio device: Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller (rev 01)

You have to select the proper things in the mixer.
Then you can call by Skype, play music through the Skype, discuss it,
and record all these as a kind of podcast.

Hardware loopback recording is possible with Intel HDA soundcards with Windows XP. But it might be impossible with Linux, because the proper drivers are missing. You can do this with Linux too, if you have a very old computer with non-HDA onboard soundcard.
igorzwx
Supporter
 
Posts: 987
Joined: Sun Jun 28, 2009 9:31 pm

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby yvon » Fri Jan 22, 2010 9:25 am

Extremely complimentary and useful post along with cesium's brilliant help. MUCH appreciated, with great thanks.

Yes, funny that older internal chips are cooler. I recently had a couple of interesting sessions/attempts with HDA intel where the distortion was huge on just one side. See my other post with 'distortion' in the title.

I'm hoping to get some clear head and time soon to try your method. really really great to know about.

I have 3 term windows open and do ossrecord in one, (to record podcast) though I do that after I've started ossplay in another (play music track into skype conversation) and then to hear the recorded podcast back I later do another ossplay or whatever player I wish to use.
To play mp3 i use mpg123 (maybe piped into mpg123 via mplayer)
I am forcing myself and enjoying no GUI for some months now though it can be slow going for my skill level...hehehe

For OSS latest version and maybe to fix distortion with HDA-intel audio:
I'm going to have to look around to see if there's been any changes in the mercurial updates. i'm not sure if one can update by just reinstalling/rebuilding package through mercurial. I had mercurial version a couple of months back then did one of my usual reinstalls and just went with latest release version.

I'll be writing back for sure!!!


might be impossible with Linux, because the proper drivers are missing. You can do this with Linux too, if you have a very old computer with non-HDA onboard soundcard.

P.S. very interesting. Hopefully someday correct drivers for hda to do this on linux.
yvon
Member
 
Posts: 18
Joined: Sat Dec 19, 2009 2:09 am

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby igorzwx » Fri Jan 22, 2010 3:05 pm

yvon wrote:I recently had a couple of interesting sessions/attempts with HDA intel where the distortion was huge on just one side. See my other post with 'distortion' in the title.

For OSS latest version and maybe to fix distortion with HDA-intel audio:
I'm going to have to look around to see if there's been any changes in the mercurial updates. i'm not sure if one can update by just reinstalling/rebuilding package through mercurial. I had mercurial version a couple of months back then did one of my usual reinstalls and just went with latest release version.


Hi Yvon!

What do you mean by "distortions"?

You see, Cesium tends to advise everyone to use PulseAudio with OSS4. There should be a reason for this.

PulseAudio does produce strong distortions, but, on the other hand, distortions are a kind of "special sound effect". There are, for example, sophisticated Audacity plugins, with which you can produce "non-linear harmonic distortions". There is a hypothesis that certain harmonic distortions may have a kind of psychedelic effect. This may explain the phenomenon of "PulseAudio addiction".

Your Intel HDA soundcard is likely to have a very bad resampler inside, and if it is the case, sound distortions might be audible with Skype, in particular. This problem might be cured with "Production quality resampler" which you can enable in the mixer, provided you know how to recompile and hack OSS4 (in some very special way)
viewtopic.php?f=3&t=3543

There are special test files, but they should be used with a great precaution (!!!).
Before using such files, it is advisable to read the entire thread
viewtopic.php?f=3&t=3543

Another source of distortions might be Linux audio players.
With OSS4, the native player "ossplay" might be better than others
http://4front-tech.com/hannublog/?page_id=34

It makes sense to clarify the problem of sound distortions with Intel HDA soundcards.
If the cause is simply a bad resampler, it might be easy to fix.

Best,
Igor
igorzwx
Supporter
 
Posts: 987
Joined: Sun Jun 28, 2009 9:31 pm

Re: Need facts>OSS: podcast w. music recording output AND input?

Postby yvon » Fri Jan 22, 2010 7:22 pm

Thank you Igor! Very helpful bits in your post.
Appears the heavy distortion on the right side is a bug/issue/problem that's known about.

viewtopic.php?f=16&t=3553

I tried pulse / jack etc. months ago and am firmly decided not to use pulse or any other layers. Just OSS. Happily OSS alone is great.

I am looking forward to trying many more things based on the above posts as soon as I can. Ideally I will have my skills get good enough to be able to 'wire' together the podcasting setup with mostly whatever machine I am on. Great to be learning. The HDA drivers will hopefully be changing for the better at some point.
yvon
Member
 
Posts: 18
Joined: Sat Dec 19, 2009 2:09 am

Previous

Return to General Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron