Klevgrand introduce a new take on wavetable synthesis.
Available for Mac OS and Windows (AU/VST/AAX), the Tomofon plug‑in might best be described as ‘a wavetable synth, Jim, but not as we know it’. Klevgrand say that the instrument employs a new ‘Audio Model’ format, containing thousands of waves based on real audio as its sound sources. It is not a sampler, though you can feed it with samples, and it isn’t your usual wavetable synth either. As I understand the process, Tomofon analyses a monophonic sample, which can contain different notes, and then spreads the pitches across the keyboard in keygroups populated by separate wavetables.
Sounds are created by repeating single‑cycle waveforms in a similar manner to the more familiar wavetable synthesis. It is possible to modulate the position (Depth) within the wavetable from which waves are taken, which is what gives the sounds movement. Klevgrand explain that when playing a note at a certain pitch and at a certain Depth, the engine will find the nearest two pitch zones and inside those two pitch zones find the nearest two waves according to the Depth value, then it will render a new momentary wave. Maybe I’ve oversimplified what is actually going on, but that seems to be the gist of it!
Diving In
Tomofon comes with 124 Audio Models and over 180 presets based on multiple expressions of several instruments that include strings, vocals, brass, woodwind, guitars and so on. The way the notes are spread across the keyboard is depicted in the Audio Model display at the bottom of the GUI, showing the pitch zones and their frequency range. When playing, Depth modulation is shown by yellow dots moving vertically over the Audio Model.
Above the envelope display are tabs for Gain, Depth, Pitch and Filter, allowing them each to be controlled by their own envelope or LFO. There’s a controller matrix at the left of the GUI, below which are On buttons for four voices. Delay and reverb effects are also built in. Below the ADR envelope display, which includes a loop between D and R, are controls for Time and Value, but note that the nodes on the graph can only be moved vertically, so in the horizontal plane the visual node spacing does not reflect the actual spacing as set up using the time controls.
Clicking the Editor button brings up a file management area, faders for Normalise, Smoothing and Clean Up, as well as an Import Audio box into which you can drag your own samples. Tomofon can create an Audio Model instrument from just one sample or combine multiple samples for greater complexity, but it is important that the imported samples are monophonic so as not to confuse the analysis process. After doing a few experiments I concluded that the samples that work best are of a nominally consistent level with no gaps, and that they include some significant timbral changes. For example, if you just import a sample of somebody singing a sustained ‘Ahh’ sound, the extracted waves won’t have much variety regardless of the position from which they are taken. Import a complete sung or spoken word, however, and the instrument has more formant information to play with.
Tomofon is actually very easy to use and, to my ear, produces sounds reminiscent of traditional wavetable synths, occupying that twilight zone between realistic samples and sounds that are entirely synthesized.
Conclusion
Despite the initial unfamiliar operational paradigm, Tomofon is actually very easy to use and, to my ear, produces sounds reminiscent of traditional wavetable synths, occupying that twilight zone between realistic samples and sounds that are entirely synthesized. Some of the pad sounds are particularly appealing. The options to control various parameters by LFO or envelope can add a lot of movement to the sounds too, but you do have to choose your own imported sounds carefully to avoid the results sounding too similar.
Summary
Tomofon offers a new slant on wavetable synthesis that produces some interesting results.