Ideal sound card for multichannel scan and modulations.

[steve-o]

I mentioned ASIO because latency was brought up. Personally I don't use it on my DAW. (digital audio workstation[multitrack recorder]) and I'm no expert on ASIO; I have read that it is preferable though in recording applications due to low latency, but like you said unless youre using multiple hardware setup prob no diff-

[The_Doctor]

Nor do I. I use the MME (WaveOut) selection for output port type, using the WavePipe in Cakewalk to get me the low latency I need.

Gary, the sync problem exists, but is one of the reasons to use pro sooundcards when using separate stereo pairs. If you code for sound cards, please allow the option. Let the end user choose their card to solve this, cos that's where it's solved. Layla cards have sample accurate sync between outputs. M-Audio is probably the same, it's a strict requirement of all pro audio stuff. The OS only has to not actually DROP a sample byte. So long as they all arrive, the hardware takes care of sync and timing accuracy. The driver takes care of buffering for the latency reduction, and the program only needs to work with the driver. If an end user needs to sync two cards, they'll choose two cards that can be synced. That feature would be something the maker would boast about.

Leave all you can to them, as most of it has been solved. That's the main strength of the sound card idea.

I'm glad you're working on this, I think it will be valuable to all Windows users. James Lehman (writer of LaserBoy) described how Linux write to all IO as if to a file, whereas Windows has its own peculiar methods. What's really needed is a kind of translator. On that thought, it could be VERY cool to be able to use a small code adapter to take from any software that doesn't need specialised hardware processing. This code could patch various known data channels from various known laser show software controllers, and you just tell it what you're using, and it lets you tell it what pairs of channels to connect to what available outputs. Could become as universal and indispensible to laser show people (and many others) as Virtual Audio Cable and Hubi's Loopback adapters are to audio and music people. The only problem with it is it might undermine the market for specialised DAC's for laser software, but as it could result in an explosion of software sales they might not mind that much.


EDIT:
That USB decoding thing has bugged me too. I know it's a scheme that uses the timing of a bidirectional change in level about a zero volt point, but I haven't found an IC that can decode it and present a simple output I can figure out how to use. I'll keep looking. Maybe the simplest answer is to use one of the Cirrus Logic C-Media IC's. Most 7:1 stuff seems to use them anyway. They're tiny, hard to work on without good SMT tools, but there are several types, and they're cheap. If one can be coded to take S/PDIF in and send it out of analog outputs, that would be great. The main problem for me is I don't even know if their S/PDIF is just two channels of a 7:1 stream, or all 8 channels. It could be either, in theory. If they chose to send just two, it might be easier to find an IC to do it. An IC by some other firm, anyway. It's easy to find an audio pre-amp that takes S/PDIF in anyway, if it's on two channels.

[steve-o]

using the WavePipe in Cakewalk

I got Adobe. The latency isn't noticable. Are U a musician Dr?

[The_Doctor]

I don't think so cos I don't do it much now. I play a bit of bass guitar, and good(ish) but one-handed keyboard. I used to program synthesizers, mainly Yamaha FM. Made an accurate emulation of analog subtractive synthesis for FM that managed to convince the owner of three Minimoogs of its effectiveness, and helped to persaude him to buy the SY99 I sold him at that time. It allows simple and single parameter changes to change between sawtooth/square and invert them, as on decent analog oscillators. This method includes a psuedo random pitch control that closely emulates the fine drifts in pitch with analog oscillators. It can be set so finely that it is barely noticeable, or made crude if wanted, too. It also formed the basis of a true white noise generator that manages to avoid any repetition, and might be an interesting study in clocks and logic. I still don't really know how it manages to do that when it's built in what is basically a computer. There's also an equal-tempered tuning scale that puts the two filters together to allow accurate tuning so thay can be correctly pitched when close to or at oscillation. As their native resolution is based on binary, 16 values per octave, it took some effort to find a pattern that worked well to make it span equally and accurately through 12 tones per octave. My frustration with trying to make music led to this as a useful tool for anyone with an SY99, SY77, TG77 who wants to get very realistic analog waves and filtering as an extension of existing abilities. Could be adapted to others, but it need three feedback loops per algorithm to be efficient.

The thing that could probably excite me most is a decent pitch to MIDI controller. Doepfer made a ribbon to MIDI controller than I adapted to solve part of the problem, fluent transistions across each note with pitch bends to allow a true free pitch sweep of arbitrary length and speed with extreme accuracy. They're still working on a pitch recognition circuit though, it's not yet available. The minimum I need is a good lin/log converter of at least 12 bit accuracy, (analog equivlent to that). I want this because I can whistle very expressively. I can play a vocal line more convincingly that way than with any other means, so I want to use that as a raw input to MIDI control, it would save time and make much better music than anything I've ever managed before. The only way a single person can make good multi-part music is to use methods so different that it makes them think as differently as if they were two different people. Bass guitar and keyboard helps, but it's not the answer I want. I need free form solo line control, and I never found anything that could translate that to MIDI. Not well, anyway. Only Doepfer's stuff seems to be headed the right way at good economy.

I stopped with music mostly, because it became frustrating trying to get what I wanted with impossible means, and started messing with diode lasers......

[carmangary]

Actually, I was just looking at the cirrus chips. I found a CS8414 (superseded by CS8416) that looks promising. But, I haven't dug into it enough to understand how the outputs work. It would still need to be piped to some type of DAC chips. I found a webpage that shows how to interface to it for L&R out but I am not sure what happens when you want 8 channels out. The chips don't have too many pins on them so they would be manageable on an experimental basis.

[The_Doctor]

Got to be a single chip to take S/PDIF and spit out analog audio though. I'll work on finding it. Might be some custom part without public documentation though. The C-Media range (Cirrus) are the nearest candidates, I think, but these things move fast, there might be others now.

Edit:
http://sound.westhost.com/project85.htm
Might be helpful, that.

[steve-o]

I stopped with music mostly, because it became frustrating trying to get what I wanted with impossible means, and started messing with diode lasers......

I know what you mean. By the time I lay the drum track down...
What was that riff

[carmangary]

yea, that link is the webpage I was refering to that showed how to interface to two channels. Maybe it is as simple as hooking up 4 of those dac chips to the digital audio receiver chip. I know that there is channel data coming from the receiver chip so there ought to be a way to use that to select one of the dac. Maybe. Is there an electronic engineer in the house?

[The_Doctor]

Do you need to use more than two S/PDIF channels though? While I see that many 5:1 systems are based on 7:1 capable IC's, can't they use their existing analog outs for 6 channels, and the S/PDIF for the remaining two, or must that S/PDIF always be all eight?

I think it IS important to allow for the separate pairs. By trying to focus solely on the cheap 5:1 (7:1) stuff you're limiting the chance of high quality output unless you get into increasing expense and difficulty with dedicated hardware. The whole point of using a decent multichannel sound unit (professional audio type) is all that's been done, the only thing to do is the DC bypass. I don't think you'll EVER set exact point rates on a cheap 5:1 unit either. On the pro stuff, that's easy, you just set the sample rate and the driver and hardware take care of the exact timing.



Steve-o, exactly.. motor (muscle) memory is a big part of music making, having to do everything through a limited interface makes such poor use of that especially when one person is composing for multiple parts. We're not all Beethoven, we can't all hold the various lines and rhythms of many parts in mind while we note any one of them down. I'm suprised that more effort hasn't been put into this in the decades people have wanted it. It seems that all the things that can do it are either toys, or too expensive to consider. And sometimes both. The best controllers for electronic solo lines (Onde Martinot, Theremin) were invented at the end of the 19th century, or the beginning of the next, and no-one's improved on them. Apart from a few rare wind instrument emulators, there isn't anything new, let alone better.

[steve-o]

Yeah we need a direct brain-to-audio converter