Prism used “diode mixing” to combine two or three input waves in such a way that the sound changes depending on which note is played, in a more gradual way than the Splice Combiner. This not only allow new unique single tones or drones, but also the key organic effect of gradual variation in tone up and down the keyboard or wind controller. In addition, it can be used for clangerous inharmonic sounds, like a gentler ring modulator.
The three inputs are “diode mixed” (unbalanced frequency mixer) using the MinMaxMix approach (see this post at Electro-Music.com). If the inputs are derived from the same oscillator, only harmonic content is generated.
However, if the inputs are signals tuned to different frequencies, other frequencies will be also generated in an effect similar to (but not identical to) ring modulation.
There are three signal input sockets, with corresponding knobs
- Main input is unfiltered: the knob is an “attenuverter” allowing positive or negative levels
- High input has a simple high-pass filter whose frequency is controlled by the knob,
- Low input has a simple low-pass filter whose frequency is controlled by the knob.
So the tone will change with the note played (because of the static filters) and can be modulated by modulating the inputs.
In usual operation, you would plug the wave that is the fallback wave in the Main input (such as a sine or triangle wave at a slightly low level). Then plug in the wave that will influence the highs and high notes into the High input, and then, if you choose, a wave that will influence the lows of lower notes in the Low input.
Operation
The basic MinMaxMix diode mixing method of combining waves can be seen in these charts.
- If there is one wave active then it will be reproduced on the output (with some mild zero-crossing distortion.
- If there are two waves active then:
- if both are positive, then the most positive value will be selected;
- if both are negative, then the most negative value will be selected;
- otherwise just the average will be used
- If there are three waves active then:
- if all are positive, then the most positive value will be used;
- if one is negative, then the average of it and the most positive of the positive two will be used;
- and similarly for the negative.
So the combined wave will have the extreme bits of the shapes of the incoming waves. In the first example below, you can see two waves of about an octave apart being combined. The output wave will have additional harmonics to the inputs.
In the second chart, above, the two inputs shown are different frequencies. This gives an effect like a softer kind of ring modulation.
Because Prism operates geometrically, on the levels and shapes of the incoming waves not their harmonic content per se, the best waves to use are those with many bumps, such as the Shell and Rasp waveshapers provide, or which have been put through some high resonance filtering, such as the Vox H. resonator or some VCF.
Audio Example
This unpleasant audio is intended to show how you can mix together different components using Prism, how the tone will vary across notes and octaves, and how the distortion adds a grungy sheen. The sweep allows you to hear the changes in tone at different frequencies as rhythmic beats.
The Patch: An LFO sweeps two synced VCOs:
- VCO 1 sine out is patched to the Prism Main input;
- VCO 1 rectangular outis patched to Prism Low input; the VCO has DC blockers;
- VCO 2 (which is running several octaves higher than its master, VCO1) sawtooth out is patched to Prism High input;
- Prism out is going directly to output.
This gives a feel for the different sounds and tones. The Prism controls and VCO pulse with and LFO rate are adjusted during the run.
Lets look at the waves, to see how Prism allows you to construct quite complex waves.
Wave at Lowest Frequency of Sweep
At the lowest section of the LFO sweep, VCO2’s sawtooth wave has been, for all intents and purpose, completely filtered out, and makes no contribution.
Instead, what we see is a combination of the sine input and the 90% rectangle wave. (Because of the DC blocking on the Rectangle out (performed by the VCO), the positive “mark” signal level is about 1/9 of the negative “space” portion’s signal level.)
The slope in the rectangle is a combination of two factors: first, when the sine is positive and the rectangle is negative, you get their average (the “Mix”); second, because the frequency of the note is low, the DC-blocking capacitors are operating, creating a skewed rectangle. You can see the shoulder on the sine: this is because the sine wave goes through a diode drop so that it has some period at 0V: during this sort period, the low-level “mark” of the rectangle wave is the maximum, so it is used.
Wave at Highest Point of Sweep
For the following, we change the pulse width to about 40%.
At the highest point of the sweep, we see a big contribution from the VCO2 sawtooth. But now the Rectangular wave in the Low input has almost completely been filtered out by its low-pass filter but is still having enough effect that it blocks the saw waves during its “space” portion.
Because the sawtooth has been high-pass filtered, it shows as a series of spikes during the “mark” portion,
Waves in Between
In between these extremes, the wave does not simply crossfade. Instead, as signal levels change because of filtering, the extremes of the signals that “win” the diode mixing (or get mixed) gradually take over.
You can see that as the VCOs sweep higher, first only one of the little saw waves “wins”, then two, then more.
Physical: 4HP wide, 37 mm deep, 16 pin power socket (10 pin can be used).
Board: Blue ROHS board marked “Rev 2”
Status: RELEASED
More Info:
Freak Out! 