The SN-Voice Project: Sound Samples

The SN-Voice In Action

SN-Voice Main Sound Samples SN-Voice Schematics SN-Voice Construction SN-Voice Modifications SN-Voice Operation SN-Voice Parts SN-Voice PCB

This is where the SN-Voice puts SN76477 where it's mouth is. Eschewing my usual practice of peppering samples willy-nilly here and there, I've elected to consolidate all of the samples in one spot.

I've divided the samples up into two groups. The "Naked Voice" section consists of samples the SN-Voice produces without having its voice modified by synthesizer modules. This is to demonstrate exactly what the SN-Voice sounds like on its own. Other than the occasional touch of reverb and excessive dollop of analog delay, these samples are of the native SN-Voice.

Other than an LFO used to gate the SN-Voice in a few of the samples, all modulation is supplied by the SN-Voice - self modulation in the 'naked' samples and self-modulation and modulation of external modules in the "SN-Voice with other modules" section. All samples are single track (no overdubbing) except where noted.

The Naked Voice

This first sample is Thomas Henry's definitive SN-Voice demo, found on the opening page of this site. If you haven't already downloaded it, I urge you to do so. It's In a piece by Bach, Thomas uses two voices in the demo - one is an organ voice from Thomas' sound card, the other is the SN-Voice. There's a slight touch of reverb added, but this is the raw LFO controlled PWM sound of the SN-Voice singing.


Next up is a sample of the various pulse outputs of the SN-Voice, controlled by a keyboard with a slight bit of portamento. The first part of it uses the narrow pulse output, which is a modification detailed on the "Modifications and Addtions" section of this site. The sample starts out with a sequence of notes using the narrow pulse, then another sequence of notes occurs using the 50% duty cycle wave, followed by a sequence of notes using EG controlled PWM, and finishing up with a sequence of notes using LFO controlled PWM.

The next sample is similar to the previous sample - it uses the three standard pulse output options of the original design, plus the extra narrow pulse option. This sample begins with LFO controlled PWM, moves on to EG controlled PWM, then 50% duty cycle square wave, and, finally, the narrow pulse setting.

Here's a short sample of the narrow pulse (from the modification added to the PWM Select switch), this time in a lower register.

This brings us to the triangle wave output. This wave is not a normal output of the SN76477 - it's something Thomas snuck away from it while it was off doing other things =0). If you download this sample, you may be surprised at how un-76477-like it sounds. Ms PacMan never sounded this languid =-D. The sample starts out with the triangle wave alone. When making these samples, I ran both audio outputs of the SN-Voice into a mixer, so it was quite easy for me to mix the two signals together. In this sample, towards the end, I mix in the LFO controlled PWM signal, which provides a nice articulation to the triangle wave output.

Next up is a kind of off-the-wall sample using the triangle output and the different pulse outputs of the SN-Voice at various times (like all of the samples up to this point, it's a single track - I just switched from waveform to waveform when the EG wasn't kicked in). A keyboard with lots of portamento was used to create the wailing SN76477 lament. Lots of analog delay is used on this one.

Overlooked up until now - a demo of the noise generator. I start out with a manual sweep of the generator from high to low, then move the clock up a bit and control it with the keyboard CV for a while at different octaves. Note that the EG is the original response of the SN-Voice until the 00:56 mark in the sample. Here I use the EG response control (a modification discussed in the modifications and additions section of this site) to sharpen the response of the EG. I simply crank the control all the way CCW - notice the attack is much crisper. This modification plays particularly well with the noise output.

The next sample is another demonstration of the EG Response control modification in action. In this sample, the pitch is controlled by the keyboard, but the SN-Voice is gated by a narrow pulse LFO (Ray Wilson's "Cool New LFO") from my modular. The sample starts out with the 'normal' SN-Voice EG response. As the sample progresses, I turn the control CCW towards the sharper response. As a result, the general volume of the sample increases - the EG rises more quickly and has more 'on time'. Note the gradual change in the EG response. The EG controlled PWM setting is used.

A lot is going on in this next sample. Again, a narrow pulse LFO is gating the SN-Voice. The triangle voice is mixed in as well. The SN-Voice LFO is set for a very rapid rate and is modulaing the EXP FM input of the SN Voice. Moreover, the first part is using the LFO controlled PWM. The combination of frequency modulation and pulse width modulation of the two voices mixed together produces a fairly interesting voice. The sample starts off with subaudio modulation gating the SN Voice while I hit notes on the keyboard, which is controlling both the VCO and the noise generator. Towards the middle of the sample, I flip the range switch on the gating LFO up into the audio range - now the effect is audio rate pulse wave AM. From that point on, I begin manually varying the SN-Voice LFO frequency and EXP FM Amount Control, as well as the frequency of the gating LFO. At some point I switch the pulse/noise output to noise, and the sample draws to a close.

SN-Voice With Other Modules

I originally posted the next three samples on the Electro-Music forum.

On this next sample the pulse output of the SN-Voice is patched straight into my 2040 clone filter. The triangle wave is patched into the triangle input of the wave multiplier, then out the folds output to a second input of the same filter.

The EG output of the SN-Voice is going to the folds CV input of the wave multiplier and one CV input of the filter. The LFO output of the SN-Voice is going to the offset control of the wave multiplier and to a second CV input of the filter. The output of the filter is going straight to the amplifier, and the recorder is just recording from the line-in of the amplifier.

The SN-Voice is set to modulate the PW of the pulse wave with the SN-Voice's EG.

The sample starts out with just the pulse wave heard - I have the output of the wave multiplier turned all the way down. The CV inputs of the filter are turned all the way down, too - the SN-Voice EG and LFO are not modulating the filter.

Then the sample pauses and I turn up the input attenuator of the second input to the filter, the wave multiplier signal. Because the triangle wave is not controlled by the SN-Voice EG, it's constantly on, so even if I don't hit the keyboard, you can still hear it as a sort of drone, modulated by the SN-Voice's LFO.

Then, when I do start playing notes, a mix of the wave multiplier and square wave can be heard (the SN-voice is set so the square wave passes through the internal VCA of the SN-Voice). The EG also is modulating the 'folds' of the wave multiplier. Then, as I'm playing the two lower notes (badly) I first bring up the LFO CV on the filter and then the EG level on the filter (the lower notes display my inability to accomplish two tasks at the same time).  Then I play some higher notes again, and the sample fades out.

So, in addition to the SN-Voice, this sample uses a filter and the wave multiplier. And the analog delay on the output.

The tone the wave multiplier provides, to me, is just pretty cool - it can do so much, it just amazes me.

The first part of the next sample uses the Wave Multiplier (it's the same patch as the first sample). On this one, I'm modulating the expo input with the output of the pulse, so it bends all over the place. Then, the second part is the same thing, but no wave multiplier

The next sample is just a hodge-podge collage of various sound (I.E. it's multi-tracked). The sounds were on different tracks of the D8, I just sort of mixed them in and out. It's a fair demonstration of how the SN-Voice can do musically-intervalled sounds as well as some pretty cool sound (these are more DSC2000-like sounds). I used the same filter and a VCA with these sounds, but no wave multiplier. I used a Ray Wilson LFO to auto gate it on a couple of tracks, but all modulation and voices heard are provided by the SN-Voice (many sounds are accomplished by self-modulating the SN-Voice, using the noise output to modulate the triangle wave, things like that).


The Thomas Henry SN-Voice design is for personal use only and may not be published without permission of Thomas Henry or Scott Stites. This site copyright (c) 2010 Scott Stites. All rights reserved.