Ideal sound card for multichannel scan and modulations.

[carmangary]

I'm not sure what happens if you feed a 7.1 wav into a soundcard with 5.1 analog outputs and 7.1 capable SPDIF. Does the soundcard only send the lower 6 channels to the audio outputs and the total to the SPDIF output? Or does the soundcard attempt to mix all channels into the analog outs? Is it possible to send channels 1-6 to analog and 7 & 8 to SPDIF? Perhaps the soundcard driver allows you to set that up. I don't know.

But from a external programming standpoint all I have available is one or more sound devices and a wav data. I can create a single wav file with 8 channels and send it to a single device or 4 wave files with 2 channels and send them to 4 devices. And, although I can tell all 4 devices to start right after another, that doesn't mean that they actually will... which is the synchronization problem. Are you saying that the pro audio stuff will appear as multiple devices in windows? You've kind of lost me on all of that since I am only familiar with the cheap sound cards.

[The_Doctor]

From http://www.cmedia.com.tw/forums/view...start=15&t=863

The way that Vista access sound now is very different. The sound infastructure has been re-written meaning that Vista wont access 5.1 & 7.1 via normal hardware methods as it did in XP - in fact its no longer supported in games using EAX and DirectSound. Creative have gotten around this by using ALchemy, which is in the form of whats called a 'wrapper' to access Vistas 'OpenAl' method of accessing sound - but even they are not supporting this fully across their range of soundcards. Patience! At least a lot of you are getting sound - even though it is just in stereo and not glorious 5.1 or 7.1. I sympathise with those of you who cannot get a bean, but eventually you will.

Looks like it might be best to work with stereo pairs for now. At least that way people can get something solid during the transistion. This is why pro audio stays with stereo pairs, they know it's going to work and stay working.


Steve-o, I call my brain-to-audio interface 'whistling'. Sadly, it hasn't got enough bass.

[carmangary]

Not sure what that's all about because I didn't read it all, yet. But I am currently running Vista and have no problem writing sound out. My CM106 based USB soundcard doesn't work but my internal soundcard works fine. I have sent sound to it using DirectSound (2 channel) and WaveOut (6 channel).

Edit:
I do remember reading that Vista won't support stuff like EAX. So, any hardware processing on the soundcard isn't supported in Vista. But we don't want any of that for laser control anyway.

[The_Doctor]

I once tried to understand 5:1 stuff better, and couldn't find guidance that helped me. My own explorations didn't find much either, except that Cool Edit Pro (now Adobe Audition) could load the multichannel WAV format and let me save as separate stereo files. Whether that means it can access 5:1 outputs with that file, or only disassemble it for stereo pairs, I don't know.

I do know that the professional standard is for stereo pairs though. ALL high quality audio programs allow that, and depend on good multiclient audio drivers and hardware to render their output as analog signals. This is a deeply established standard with security that ILDA will work for years to attain. Surely it's worth writing to meet that if you're meeting the ILDA standards.

The 5:1 output on my mainboard allows multiclient audio. I can send a stereo music signal to it from Foobar while Media Player Classic is playing a movie. While this implies a mixer somewhere addressing the same stereo output pair, you could solve all problems if you can multiplex them into a format for the 5:1/7:1 system's driver. I think whatever you do, you have to start with stereo pairs.

[carmangary]

A 5.1 wav file is not much different than a 2 channel wav file. It's basically just 4 more groups of data stuck into it. It's trivial to rip apart wavs into their components or stack them to create multiwave tracks.

When you play two things at once, Windows is usually taking care of the mixing. You could play a stereo track and a quad track with silence on channels 1&2 and you would end up with the stereo track on speakers 1&2 and the quad track on 3&4. You can't create a 2 channel wav that only targets 3 & 4, though. It has to include all channels (1 to n).

You can also use the Windows mixer to pan and add 3d and stuff. So, you might be able to take to stereo tracks and play one on 1&2 and mix the other onto 3&4. I haven't looked into that. It wouldn't make sense to do that for laser control since it is easier to just create a multitrack wave and add all the data and pass it on.

For multi devices you'd have to create multi waves and send each to the correct device and somehow sync them. It wouldn't be all that difficult to write a driver that allows you to specify which channel/card is for x, y, r, g, b and let it send to the correct wav and device. It just adds another level of complexity.

If the pro audio stuff supports only stereo pairs then I would think it would fall into the latter category above. But I wouldn't exclusively do it that way because it is problematic.

Well, I got my mica and plastic screws today so its time to switch gears for awhile and get back to hardware.

[The_Doctor]

It's a convention. First, based on stereo, second, based on multiplexing pairs into groups of stereo ports or into multichannel formats. Proof? Ever seen a wave format with an ODD number of channels?

It doesn't matter that a wave format with multiple channels can't address 3 and 4 without 1 and 2, you still have to decide what goes where at some level. I think, but probably I'm getting this wrong: Left front, Right front; Left rear, Right rear; Centre, Bass; Left surround, Right surround. As we need all 8, whatever we call them, that makes it the same, they're multiplexed either into a multistream format, or into separate feeds to stereo pair outputs. It's only at that point of branching that any difference matters.

When sending to stereo pairs, you don't have to sync them. That's the hardware/driver system's job. All you have to do is send each channel a byte or two or three (depending on bit depth) in sequence, which is exactly what you do when laying data into a wave file, whatever format it's in. You don't have to specify which output pair each will route to, but I strongly recommend allowing it for the user to set, as those kinds of cards are usually going to be dual purpose, often plumbed into existing systems. It could default to the same convention you use for laying the data into the multi-track wave format though.


Time for me to switch gears too, I am now entering Food Coma. Good luck with that nasty, but pleasantly shiny mica.

[steve-o]

IMHO-plasic screws dont provide enough torque to provide hard surface-to-surface contact to for good heat transfer. Maybe use alot (but not toomuch) goose-grease for better contact. I too am presently having a component / chassis contact dilemma...
hopefully gonna get it worked out :-/

[carmangary]

Laser is running cool so it's good enough for me.

[steve-o]

Cool........................................

[Angelillolaser]

I built and use my own sound card some years ago using pcm2900 form texas, you can make waves 0.1hz to 22.2khz at the same amplitude, pcm2900 is the best DA converter that I used, use usb conections..