I've been doing some experimentations with DC blocking caps on an old USB soundcard to see whether I could get Silentway working on it.
The soundcard in questions is a M-Audio Transit which is a simple 2 in 2 out usb sound card. It's pretty simple schematic with the output of the DAC feeding an opamp with a couple of DC blocking electrolytic capacitors on the output.
So I dug out my soldering Iron, bridged these capacitors and measured the voltage of the audio output, sure enough there was a constant DC of around 2.3 volts. Firing up the silentwaydc vst app I had a go at varying the voltage at this ouput and managed to get a constant voltage that varied from about 0.8 volts to 3.6 volts. The voltage appeared to be constant and did not drift.
So with a bit of level shifting and amplification I'm sure I can get this to work with the Silentway VST. My main problem is that the sound card is a simple 2in 2out affair and as such I would need to find a host that supported more than one ASIO soundcard, in order that I could dedicate this one to CV duties.
Bridging soundcard DC caps
-
- KVRAF
- 6323 posts since 30 Dec, 2004 from London uk
No need for all that - just get a MIDI to CV unit. Job done :
http://www.philrees.co.uk/products/miditocv.htm
http://www.philrees.co.uk/products/miditocv.htm
-
- KVRAF
- 6323 posts since 30 Dec, 2004 from London uk
It may not be the MIDI to CV unit timing, it may be how its being used. Adding more software +extra ASIO in order to correct the timing will only add to the fun.
-
- KVRian
- 1417 posts since 22 Mar, 2002 from UK
I think the main point here is that once you've got any latency/timing questions sorted out, the timing on a soundcard based system (e.g. Silent Way) *will* be better than any MIDI/CV based system.
With MIDI/CV you've got, at least, the following factors, all of which may be variable:
* the DAW generating the MIDI events (this really should be tight)
* the MIDI hardware driver picking up that event and turning it into something that causes the MIDI hardware to do something
* the time for the MIDI hardware to generate and send the relevant messages
* the time for the MIDI/CV converter to receive and interpret that message, and turn it into a voltage.
Once an audio interface-based solution is working, your sequenced voltages are simply generated with sample accuracy in time with whatever audio stuff your DAW is doing. End of story.
If you're using one interface for audio and CV output, life is really easy. If you have to jump through hoops to get more interfaces working together, then maybe that's hard but at least the timing will be completely deterministic and therefore correctable. With actual hardware MIDI in the chain, timing is just really hard to guarantee.
With MIDI/CV you've got, at least, the following factors, all of which may be variable:
* the DAW generating the MIDI events (this really should be tight)
* the MIDI hardware driver picking up that event and turning it into something that causes the MIDI hardware to do something
* the time for the MIDI hardware to generate and send the relevant messages
* the time for the MIDI/CV converter to receive and interpret that message, and turn it into a voltage.
Once an audio interface-based solution is working, your sequenced voltages are simply generated with sample accuracy in time with whatever audio stuff your DAW is doing. End of story.
If you're using one interface for audio and CV output, life is really easy. If you have to jump through hoops to get more interfaces working together, then maybe that's hard but at least the timing will be completely deterministic and therefore correctable. With actual hardware MIDI in the chain, timing is just really hard to guarantee.
-
- KVRAF
- 6323 posts since 30 Dec, 2004 from London uk
Id like to see the outcome of this. I think it will work wonders with one card doing it all. I think with 2 cards, its not going to be easy as Ive tried linking 2 cards and latency is different between devices. I used ASIO4ALL to link the cards together. For me, I had to run at 25ms latency to make the sounds cards work together. No good for VSTi. I think you just exchange one timing issue for another.
-
- KVRian
- 1417 posts since 22 Mar, 2002 from UK
My point is, it may be different between the devices, but as long as it's the same difference all the time you can exactly compensate for it.UltraJv wrote:its not going to be easy as Ive tried linking 2 cards and latency is different between devices.
For sure, a single interface is easier though.
-
- KVRist
- 175 posts since 1 Jul, 2009
When I want really tight....I mean tight as in down to the individual sample tight!
Silent Way gives me that tightness. If I want my midi to wander and jitter, I would like it to be because I programmed it that way and not be at the mercy of usb midi.
Silent Way is as tight as tight can be, it's not actually possible to get any tighter lol
Silent Way gives me that tightness. If I want my midi to wander and jitter, I would like it to be because I programmed it that way and not be at the mercy of usb midi.
Silent Way is as tight as tight can be, it's not actually possible to get any tighter lol
-
- KVRer
- Topic Starter
- 13 posts since 20 Jun, 2001 from Manchester
Just thought I would give an update on my current progress.
I have built a small circuit that takes the signal from the modified M-Audio Transit adds the required offset voltage and amplifies the signal. It's currently amplified to +- 5.5v.
I have had a few tentative tests with the Silentway Voicecontroller plugin and an old Roland synth and it appears to be working without fault. I get a good number of octaves out of it and it is triggering fine.
The main problem I have encountered is trying to get the 2-in 2-out card to coexist with other cards in the host software. I've ended up using the ASIO4ALL driver to enable more than one card.
I will do some more thorough tests and post the results. If I get time I might even do a video.
I have built a small circuit that takes the signal from the modified M-Audio Transit adds the required offset voltage and amplifies the signal. It's currently amplified to +- 5.5v.
I have had a few tentative tests with the Silentway Voicecontroller plugin and an old Roland synth and it appears to be working without fault. I get a good number of octaves out of it and it is triggering fine.
The main problem I have encountered is trying to get the 2-in 2-out card to coexist with other cards in the host software. I've ended up using the ASIO4ALL driver to enable more than one card.
I will do some more thorough tests and post the results. If I get time I might even do a video.