Linux Audio the way to go!?

[syntheticwave]

…thanks, Soundcheck.

I agree, the frontal reproduction is the most important. But the idea isn’t a limited by frontal reproduction.
The common practiced horizontal loudspeaker rows around the listener for wave field synthesis takes much effort for rear effects but limits the reproduction to the horizontal plain. We cannot determine the exact source position outside our visible range, but we doesnt hear like a flunder.
In principle would be possible a realy 3D Holophonic reconstruction of the genuine sound field only by loudspeaker walls all around the listener.
Our idea was, substitude the loudspeakers all around by its mirror sources, which occur behind the playback room walls. It’s a better spouse acceptance factor. , alone the speakers behind the picture screen remain. But the directed wave fronts become reflected from all directions, see that little animation:
http://www.holophony.net/sound%20fie...sformation.htm
In principle such way is used by the sound projectors. But it produced virtual loudspeakers. We want to reproduce the source itself including the correct temporal and spatial behaviour of the first recording room reflections. That fakes the size of the opera house in the living room.The reverberation comes from all directions then, like in the recording room. That seems feasible by bearable effort. And we can walking across the virtual recording room by the remote during playback hence by our idea.

H.

[syntheticwave]

Originally posted by soundcheck

Interesting "scientific" project.

Though, on a real good systems you'll actually get a really nice 3D image in front of you. It'll cover pretty much all the information which were recorded.
Cheers

...its a little besides the Linux topic, but by my opinion we are far away from a real 3D image by all phantom source based procedures. Tonal accuracy is the best that can be hoped for in a traditional audio system; true spatial accuracy will never happen. Audio products should come bearing this disclaimer:

" WARNING: IMAGE PRESENTED IS LESS THAN LIFELIKE !"

The reason for that statement is described by http://www.holophony.net/Phantom-sources.htm
but we should discuss that possibly in another thread.

H.

[soundcheck]

Originally posted by syntheticwave


...its a little besides the Linux topic, but by my opinion we are far away from a real 3D image by all phantom source based procedures. Tonal accuracy is the best that can be hoped for in a traditional audio system; true spatial accuracy will never happen. Audio products should come bearing this disclaimer:

" WARNING: IMAGE PRESENTED IS LESS THAN LIFELIKE !"

The reason for that statement is described by http://www.holophony.net/Phantom-sources.htm
but we should discuss that possibly in another thread.

H.

Sorry folks for being of topic!

-----

I don't agree on your statements/position outlined on your homepage - at least partially. Perhaps it is my limited theoretical knowledge about the subject. But a real good stereo system will deliver plenty of spatial information. You can easily identify several layers of an orchestral setup. The sound field goes even beyond the speaker limits.
This will even improve if you use a dipole speaker.

The limiting factor is IMO first of all the recording itself.
Why does people with high end-systems enjoy e.g. Denon stereo-2-Mic classic recordings from the sixities? Because these are not mixed to death, by some engineers. As you might know there are some other small labels nowadays trying to keep the recording chain as short and as simple (not in terms of equipment quality) as possible. This is probably the only way to get a close to realistic result. It'll be never perfect that's for sure. But it'll be the less of two evils.

Standard life (classical) events IMO do sound much worse, with a much higher grade of distortions in the soundfield, compared to high quality recordings played with my own setup
at home. That's why it needs real sophisticated recording sessions ( performance, location, equipment) first of all to have a good starting point for a close to realistic reproduction.

Cheers

[syntheticwave]

Originally posted by soundcheck

I don't agree on your statements/position outlined on your homepage - at least partially. Perhaps it is my limited theoretical knowledge about the subject. But a real good stereo system will deliver plenty of spatial information.

…Hi soundcheck,
only short answer because we really going off topic.
I agree good records produce plenty of spatial impression by good stereo equipment, but determined by playback room acoustics mainly. We cannot anticipate a real 3D reproduction if, for example, if no information regarding the elevation of the sound sources in the record.
Nevertheless the stereo reproduction may formidable event. But is never the true spatial reproduction of the genuine.
No disturbing sometimes, but if we would able for that would be important step without all doubt. Thus, for going back to the topic, such open minded Linux groups as I have found for proof of concept is joyfully or all audio enthusiasts because we can find possibly more sophisticated ways.

Best, H.

[soundcheck]

Hi folks.

I just tried the speaker baffle simulation tool EDGE under Wine.

EDGE baffle simulation

From what I heard, results are supposed to be pretty reliable.


Cheers

[.: ZMN :.]

To append my earlier post #873 on MPD-on-a-NSLU2

[3] To add remote control access to (or keyboard shortcuts) to MPD via mpc in a console mode: download, compile and install actkbd
First, you might need to add the tools for compilation:

sudo apt-get install make gcc

Here is an example configuration (/etc/actkbd.conf) for an Ati Remote Wonder II, to control mpc.

# play = 207
# stop = 128
# channel+ = 402
# channel- = 403
# question = 214
207:::/usr/bin/mpc play
128:::/usr/bin/mpc stop
402:::/usr/bin/mpc next
403:::/usr/bin/mpc prev
214:::/usr/bin/mpc update && /usr/bin/mpc clear && /usr/bin/mpc add /

With that added I suppose a Linksys NSLU2 NAS, combined with a powered USB hub, a USB-soundcard and a USB-remote, turns into a fairly cheap linux music server.

[soundcheck]

Hi folks.

Rob (the Sox guy in charge) just informed me that the newest SOX CVS revision offers an external FIR coefficient interface. I discussed this with him about two weeks ago. Now it's done. Great guy.

These are great news. At at least to me.

Cheers

[phofman]

Great news. I guess another useful development would be a CUDA filter, at best prepared as a standalone library called from a SoX effect.

All that (together with the piping file plugin for now) could make alsa a very powerful and configurable backend for the HIFI world.

[1audio]

Originally posted by soundcheck

As said earlier. MPD supports 24bits with latest git release. Give a manual compilation
and install a try.

Cheers [/B]

I have tried to get a true 24 bit stream from MPD without success. I have installed the latest MPD (14.2) While MPD doesn't crash and sound comes out its truncated to 16 bits. I posted a note on the MPD forum and Max from MPD asked for access to my system to troubleshoot it. I have set up access for him so hopefully he can make progress.

However with the mentions here I thought I would ask if anyone else has verified true 24 bot data from MPD. I know my hardware is working right (aplay does 24 bit) and VLC will output the full 24 bits but MPD doesn't yet.

Should anyone want to know how to check this its a little cumbersome and hardware intensive. I use a cheap dac from ebay with a 192K crystal receiver chip on it. I brought out the L/R clock and some of the other clock/data lines as well. If you sync on the l/r clock and look at the raw datastream from the SPDIF connection its easy to see which bits are changing and which are not on a scope. The visual signature for the 24 bit stream is quite different from that of a 16 bit stream. The only other way to verify this is the RR HRx files into the BADA dac and look for the HDCD light.

[phofman]

Hi Demian,

If I get the git html interface correctly, this should be from 0.14%_beta2

http://git.musicpd.org/cgit/master/m...5a58bed40#n223

When you enable debug, does mpd output the debug message about converting to 16bits?

What is your alsa output device defined in MPD?