Understanding Control Voltages vs. Midi

[Larrio]

Hi all,

So, I'm running a TD-12 with an array of pads. I know the pads send control voltages to the TD-12 module.

My question is, does the module automatically convert them to a midi value (0-127) and then trigger the samples with that value, or is it using the more sensitive control voltage value to trigger the samples? Any idea how many recordings make up one sample?

What I'm getting at here, is... do I have greater control over samples I'm triggering from my module vs. sending midi out of the module to a different sampling device, such as a laptop?

Thank you for any insight you can lend on this one!

[Michael Render]

The modules use the MIDI value. That way, recording and playing back the MIDI stream will result in the same exact performance as live.

Roland does not disclose how they do the sampling in the TD-12, so all we can do is take guesses. But their sounds go beyond mere sampling as there is a lot of COSM modelling going on at the same time.

Using some kind of sampler will give you a different kind of control. I wouldn't say greater. Some of the BFD kits, for example, not only have 127 distinct samples, but have multiple layers of distinct samples that can be blended. But to switch between coated and uncoated heads, you need a whole new sample set. With the TD-12 its a push of a button.

[Larrio]

I see.

Well, has anyone experimented with using a control voltage interface such as the dumb controller to trigger samples off a computer that have higher bit depth. One way to do this would be with open sound control, which is supported by about a dozen programs, including a few by Native Instruments.

[SuperPuss]

That is a very interesting looking device Larrio.
I am not familiar with open sound control but perhaps it is something I should investigate. My studio is full of analog gear with CV ins and outs and I have used audio outputs from Protools to trigger analog sounds and sequences directly without using midi which then come back into Protools on another track once processed by my analog synths.
I also have a Frostwave Midi - CV - Midi converter for interfacing the analog world with midi.

I have also experimented with connecting my System 700 analog sequencer to my TD12 module and that worked surprisingly well.

Here's a video of the results of my first experiment doing that.

[Larrio]

That's a very cool video, superpuss. I think you would be into learning more about this open sound stuff. I look forward to hearing what you think.

You can learn more about open sound control here and here.

[Michael Render]

Wow, OSC sounds interesting. There has to come a day when MIDI is discarded for something more flexible. This might be it.

[SuperPuss]

Agreed Michael.

I might get a chance to look through the protocol down the track but right now, I have plenty going on to take up my time. I am interested to see whether this protocol starts to gain widespread acceptance though. I think for that to happen, it would need the support of some major companies with a large inventory of new products.

[Hercules]

Good info page my friends - nice to see another Australian invention especially one where analog meets digital.

[JimFiore]

The voltages that come off of pads aren't really "control voltages", at least not in the sense of the old analog synth CVs. A standard drum transducer relies on the piezo-electric effect. Basically, it says that if you apply a voltage to a crystal, you can get it to vibrate, and if you vibrate the crystal, it will produce a voltage. The voltage is directly proportional to the physical vibration. Thus, if you hit the crystal softly, you get a small voltage and if you give it a solid wack, you get a big voltage. Now here's the fun part: The precise waveform that you get (the speed of the initial transient, damping, etc.) is a function of the size of the crystal and the surface to which it is mounted, as well as the proximity of the strike. A very basic trigger to MIDI device will simply digitize this waveform with a 7 bit ADC and just grab the highest value that occurs within a specified timing window. This 7 bit value then becomes the volume data for a note-on instruction. A more sophisticated device may perform a little DSP on the incoming waveform before determining the volume bits (such as performing absolute value, filtering, or applying a non-linear volume map). Some of this can be done upstream from the ADC to reduce computational load. Analog CV, in contrast, has a trigger voltage (AKA the "gate voltage" that basically says "note on" and triggers the ADSR) and a very precise voltage mapping onto pitch, the actual "CV" (it's been a long time since I monkeyed with that but I seem to remember one volt per octave was pretty common).

Getting a little closer to your question, if I were designing a module and had to keep my eye on costs, I'd make the sound engine respond to MIDI data, design the trigger to output MIDI data, and connect them internally. This way I incur negligible cost overhead when triggering the module from another MIDI source, or using the triggers to create a MIDI data stream. Roland hasn't made the guts of their modules public, but I think it's reasonable to assume they did it this way.

Regarding OSC, I gave this a quick once-over. My reaction as a computer science person is "Oh, that's very cool and flexible". My reaction as an electrical engineer is "Oh, this will be expensive to implement on designed-to-a-purpose gear (like sound modules)". One of the nice things about MIDI was that it was a low cost communication protocol, even 20 years ago (yes, it has been that long). It's simple to implement and the hardware to do it is inexpensive. Writing the firmware for that is tiny compared to OSC.

MIDI has its limitations to be sure, but it's not like no one ever came up with alternatives. For a while there was the MADI protocol (never went anywhere). There was also an idea to use MIDI-like commands over SCSI. (This last one actually saw the light of day with SMDI, which was a SCSI-based sample transfer protocol. I wrote a SMDI driver for the Peavey DPM-SX and DPM-SP rack mount sample units about 15 years ago. It was a speed demon for its day.)