Modcan Vocoder 71

[dude]

"Vocoder is 10 band 4 pole filter design coded for DSP by the mighty Eric Brombaugh."

this is what we know so far direct from bruce. he has said the firmware should be available in about three weeks from now. power requirements should be about 100ma on the + side and less on the -. the price will be about the same as the digi delay. you can go get in line to buy if you like or haven't already bruce asked for this to be posted:


"The output mode switch selects between normal summed out with all 10 filter bands effecting the one out. In other words mono in mono out.
Split takes input and splits odd filters left and even filters right to give a stereo spread.
Invert maps the highest frequency carrier to the lowest modulator frequency and vice-versa. It inverts the filter responses so high frequency voice for example will
activate envelope detection/VCAs on low frequency carrier synth notes or whatever is being used as carrier and the opposite for low frequency voice.
Center frequencies are also crossed-over except for dead center obviously.
The other switch adds compression curve to the modulation envelope detection inputs so that vocal signals have more oomph if selected.
It brings up low volume transients and flattens the louder one."

-Bruce

i will post any more info i find/get and of course anybody who gets one and is willing/able, please post audio video and cats!

[dude]

i've asked about the squelch knob. my guess is that it is related to the compression switch? we shall see.

rereading the description has my heart beating faster.

[rezzn8r]

and not just for robot voices. also functions as a vc filter bank.
should be one in my cabinet by the end of the year

[pugix]

"power requirements should be about 100ma on the + side"

At my request Bruce built two modules for me (VCDO and CV Recorder) to use an external +5V power supply. It's quite simple. The PC boards are ready for it. During assembly, the +5V regulator is left off and a different power connector is used. Bruce included a power cable with one end having a 5-pin connector for the module and no connector on the other end. I added a 6-pin MTA connector for the MOTM-950 3-voltage power supply. (The extra pin is explained by the fact that MOTM uses two ground wires for the +/-15V instead of one on Modcan. The +5V supply has its own ground wire on both systems.) This will greatly reduce the current draw on the +15V supply, though I haven't measured it. I noticed that the Dual Quantizer 55B has the same board design, but I didn't know it when I ordered. I'll bet that this option for external +5 power is available for the new Vocoder module, and probably all the digital modules.

[dude]

"No the squelch is a noise gate. The knob sets the threshold level that it becomes active. It helps clean up input signal that have
quiet sections that could end up sounding like noise or rumble on the output."

-Bruce

[dude]

also functions as a vc filter bank.

can you explain what you mean? as far as i can figure, that wouldn't work. there is no dedicated control control over the bands. i suppose you could tailor your carrier input to influence the follower/vcas to open only certain spots on the 10 bands?

[emdot_ambient]

also functions as a vc filter bank.

can you explain what you mean? as far as i can figure, that wouldn't work. there is no dedicated control control over the bands. i suppose you could tailor your carrier input to influence the follower/vcas to open only certain spots on the 10 bands?

I'm not sure exactly what he meant either, but every vocoder is really FC filter bank (Frequency Controlled).

For example, I used to run a guitar into my PAiA vocoder as the carrier, and then run one channel from my Ensoniq EPS sampler to the modulation input. I'd have the EPS playing a sequence where I had recorded pitch bends on a simple waveform. The pitch bend wheel had been set to control the sampler's pitch over 4 or 5 octaves.

The frequency of the pitch bent modulator signal imposed dramatic filter sweeps on the guitar. One could do the same thing with a Modcan CV Recorder and have it control a triangle or sawtooth waveform and impose user-defined filter sweeps that you'd normally have to do manually (and with very little likelihood of being able to repeat the same motion).

...but there is a Tone CV input on this modual, so?

[dude]

sure sure but not a voltage controlled filter bank. has anybody ever made a fully vc filter bank? you would needs tons of modulators but i bet that would be fun as hell in a system that could handle it.

[emdot_ambient]

I've seen VC Filter Banks discussed as dreams but haven't seen any modules.

You could still simulate it with a vocoder, again in the frequency domain, by running a VC mixer into the modulation input...but you would need 1 VCO and one VC Mixer channel per frequency.

OR...run a square wave into a MOTM 730 VC Pulse Divider and then its outputs to a VC Mixer...

[discomicke]

has anybody ever made a fully vc filter bank?

http://www.ear-group.net/model_16.html tada!

[rezzn8r]

also functions as a vc filter bank.

can you explain what you mean? as far as i can figure, that wouldn't work. there is no dedicated control control over the bands. i suppose you could tailor your carrier input to influence the follower/vcas to open only certain spots on the 10 bands?

not sure exactly how its going to work. it was something Bruce said to me on the phone. again, not sure how/if you can set the levels for each band, but apparently one of the cv inputs will sweep the bands. looking forward to getting my hands on one to see how this works!

[dude]

right on! it will be very interesting to see how this one turns out. a little more mystery than usual

[ach_gott]

not sure exactly how its going to work. it was something Bruce said to me on the phone. again, not sure how/if you can set the levels for each band, but apparently one of the cv inputs will sweep the bands. looking forward to getting my hands on one to see how this works!

Sounds like a VC mixer function that changes the level of the output of each band similar to the Moog MuRF. Just speculating...

[rezzn8r]

I just got word from Bruce that my Vocoder is ready!
I'm going to pick it up tomorrow afternoon, so expect a demo soon!

[dude]

me too! you get to make first video as mine won't likely even powered until i can make sure everything is ok.

[decaying.sine]

[video][/video]

[rezzn8r]

[rezzn8r]

This thing sounds fucking awsome with a drum loop as the modulation signal

[dude]

This thing sounds fucking awsome with a drum loop as the modulation signal

nah you don't even have it.

MAKE VID!!! or at least audio.

pix/vids or you don't have it.

tough love mister

[rezzn8r]

This thing sounds fucking awsome with a drum loop as the modulation signal

nah you don't even have it.

MAKE VID!!! or at least audio.

pix/vids or you don't have it.

tough love mister

tomorrow. I promise.

[ndkent]

To me (and without further doc) this looks like it's meant to appeal to people wanting a vocoder both their modular and a small space. It's definitely not a no holds barred modular vocoder. It's interesting that it seems to add some extra features and is arguably missing some.

The compressor is surely useful have to have built in. Stereo out is uncommon, how well it sits is a good question, but you can always do mono. The band invert feature to me sounds like a novelty, it can't hurt but on a vocoder where you can patch bands, a valuable feature you could then access is the ability to 'transpose" the range, like have a low pitch modulator mapped onto a higher frequency range or vice versa. The attack decay feature seems pretty unique as it sounds like a separate up slew and down slew to me. Several Vocoders have slew to shape the performance of the output, but that would affect up and down at the same time. Another common feature I see missing is lack of unvoiced detection, in other words control on inserting some other sound for unpitched portions.

As for VC filter banks, it depends what the VC is applied to. When you have patchable bands (not found here) you can also get at them with control signals. Doepfer for instance makes a MIDI in and separate MIDI out module with enough channels to control each band. Sennheiser and Doepfer as an option offer enough offsets for each band, so you can essentially open them. Perhaps Decay might be the key? You would get a filter bank if you feed a signal with a specific spectrum in and then either keep it fed in or temporarily locked it somehow.

Anyway, with fixed frequency bands there are definitely some players. What isn't cheaply available is a big bank of filters with both VCAs for amplitude and VC over the frequency. Pre-built dedicated Vocoders use fixed filters to cut down on cost. One could of course make a vocoder given enough modules. But thinking about it now, with DSP there would be no cost limitation to require fixed filters so long as you had an interface to control the filters individually, I don't see that, but it's one price advantage of DSP. Then again DSP still has a cost. I can see maybe the 10 bands might be due to that rather than have more bands.

Anyway, I'm just sharing what I know, I've not heard the module, what it sounds like obviously is a huge factor

[dude]

yeah, messing with it will surely be telling. it dos appear limited in setup. but then some of the better end results i have experienced with a vocoder came from limited options. implementation will tell. i think the invert function came from another design. i don't remember where he said that was from but i think i recall him saying it at some point. whats interesting to me are the vc parameters and the dual outs with the band splitting to send half the freqs to one side and the next half to the other. the envelope looks similar to the way the ableton live vocoder works. i was surprised that 10 bands was chosen given the dsp nature. it seems like having adjustable bands would be good and maybe not too hard. the lower the bandwidth, the less intelligible the speech tends to be if one were to use it for speech applications. but then again, they may have just found a sweet spot with 10. we are talking about eric brombaugh and bruce duncan so i doubt anything was done in a fashion other than pure impeccability.

anyway mine arrives in no time and i am sure we will see and hear rezzs soon enough.

anybody else got one on order?

[NV]

i was surprised that 10 bands was chosen given the dsp nature. it seems like having adjustable bands would be good and maybe not too hard. the lower the bandwidth, the less intelligible the speech tends to be if one were to use it for speech applications. but then again, they may have just found a sweet spot with 10.

I was a bit surprised by the 10-band approach as well. I sort of figured they would have a selection switch for something like 8, 16, and 32 bands. Past that point you start getting less significant returns compared to processor needs, but at 32 bands you can get some very smooth intelligibility.

Roland's old vocoders were in the 10 and 11 band range, so that may have had something to do with their decision. A large number of vocoder bands can be asking a lot from some processors, but then again Electro-Harmonix has the V256 vocoder pedal with adjustable bands from 8 clear to 256 plus other features, so it's not out of the question.

[rezzn8r]

ok, here's a little taste. I must admit that I didn't buy this thing to make robot voices. I just want to pump beats through it.

looks like it's meant to appeal to people wanting a vocoder both their modular and a small space

yup, that pretty much describes me. While individual bands would be nice they would eat up a lot more space, and so far I have found that the simplicity of the Modcan version lets me get the sound I want quickly and easily. That said, my demands and expectations may be lower than those of other Wigglers.

Sample is a drum loop as modulator, and an analogue VCO as carrier.
More to come...

[dude]

could you by any chance do the same patch with the carrier and modulators switched? and what analog vco wave have you used?