This month, we get a wonderful master class on sound creation from one of the “secret weapons” that I and other veterans of the synth industry have employed over the years. Scott Frankfurt is an artist, engineer, sound designer and producer par excellence, who worked with me at Ensoniq for many years, later doing patch design and demos for Korg, E-mu, and Yamaha. He is currently the Vice President of Design at Spectrasonics, working closely with founder Eric Persing and the team developing their synths and sounds while running a world-class studio. [In his spare time? —Ed.] This month, we’ll examine Scott’s views on all the successful marks a lead synth solo sound needs to hit. Next month, we’ll look at a specific application of these principles from a patch he designed for Omnisphere.For more about Scott Frankfurt, visit scottfrankfurtstudio.com.
Fig. 1. Routing velocity to sample start time in Omnisphere.
Fig. 2. Delays with built-in filters.
1. A lead synth sound needs to “cut through” the sonic space in which it’s set. Keep in mind how focused the bandwidth of a lead guitar is. You don’t want a frequency-bloated sound that masks everything, or worse, gets masked by everything else. It needs to play nicely with others. That said, there’s room for a lot of diversity here. A smooth lead sound perfect for instrumental jazz is not going to work for hard progressive rock, for example.
2. It should include a satisfying attack characteristic that gives tactile feedback to the player. Percussive or not, the attack needs to be interesting in its own right (think about repeated short notes), and developing interest on the velocity axis is crucial. In the case of a sample-based sound, try modulating the sample start time via velocity. For samples with a prominent attack transient, start the sample playing just after the attack portion for a softer attack, and then modulate the start point backwards via velocity so the harder you play, the attack portion gets used. For other samples, routing velocity to start point can add slight randomness to each attack, which sounds more interesting and less repetitive (see Figure 1).
The ear and brain make a decision about what a sound is and whether it’s pleasing within about 500ms, so it’s important to work some character into this crucial time period. In the case of a synthesis-based oscillator, an amplitude or pitch “bump” or “blip” using envelopes can add character to the attack period.
Another example of this would be to modulate an engine parameter (like a bit-crusher, waveshaper, small amount of FM, grain shifter, fast-rate pulse width modulation, or even bringing in another oscillator that provides a short transient) to get the attack to “speak,” and then use velocity to control how much of this is heard. Anything you can employ to bring variance to the initial 40ms is desirable.
3. Related to the attack, legato/solo mode is typically desirable for leads. This gives the player more articulation options by letting intuitive legato playing technique determine envelope re-triggering.
4. The sustain segment must be compelling. Harmonic activity is important for sustain—this is where the emotion is. This sonic movement, even if it’s subtle, enhances held notes, allowing them to “sing” for quite some time. Synth soloing is similar to singing from a phrasing point of view, so you should dial in slow but steady modulation of harmonics to create the same effect a good singer naturally employs, subtly varying the sustained tone.
The personality of a sound can come from a myriad of places, so start by listening to what the waveform is already telling you. Is it squawking a certain way? Then, try to dig that out or enhance it even further, maybe with a filter or effect, and lightly modulate that. Is it super fat already? If so, then perhaps thin it out slightly it over time. Is it really thin and edgy? Go in the other direction and introduce some fatness over time.
5. Performance tools must be created for the player to interact with. Nothing is more satisfying to a soloist then reaching for the modulation wheel and hearing something awesome happen as it turns. An ideal scenario is where the initial patch sounds great, but this is only the tip of the iceberg. Performance controls (wheels, pedals, ribbons, and the like) should reveal the rest. In Spectrasonics Omnisphere, a great way to do this is to modulate the “Harmonia Depth” feature to reveal more voices tuned to various octaves, fifths, or other intervals. You can simulate this in any synth that has enough oscillators.
Frankfurt also likes to simultaneously modulate an effect so that the sound gets fuller harmonically, while also getting dirtier via sample rate modulation or distortion. Another trick is finding a key parameter that has two sonic sweet spots, and limiting a performance controller (slider, knob) to target those spots as modulation boundaries. Alternately, you could use two switches to offer dedicated variations on a modulation possibility.
6. The finish. A solo lead sound should get out of the way quickly when you let up off the keys, so a fast release works great. This not only enhances playability but also makes room for the effects to speak. Here, Frankfurt prefers delay to reverb as it’s more transparent and doesn’t disappear as easily as reverb does. He’ll even band-limit the delays to create more space if possible, by thinning out the delays. Omnisphere, and many cool plug-ins such as SoundToys Echo Boy and Delay Designer in Logic Pro, offer integrated high-pass or band-pass filters to achieve this (see Figure 2). Or you can put the delay effect on a send bus and insert the filter onto the return channel strip.
7. Release velocity. If available on your controller keyboard, try using release velocity to modulate envelope release time so that a fast key-up causes a fast release time; and a slow key up causes a slow one. Details like this increase playability.
8. Don’t overdo the amplitude dynamics. Lead sounds are far more playable if the volume “hangs in there.” The typical velocity-to-amp modulation is not as useful as modulations that alter the tone of the sound. So when programming synth lead patches, use minimal velocity control over the amp level or amp envelope. Instead, let velocity control the filter, oscillator modulation, effects, and such.
9. Make vibrato like an electric guitar’s. Frankfurt prefers a triangle- or sine-wave vibrato that modulates in the positive phase only, with a maximum pitch depth of not more than a minor third. Years ago, he took a close look at Steve Lukather’s guitar vibrato style on a pitch scope, since Lukather among the most recorded guitarists alive. The positive phase over a minor third at the furthest position of the wheel most emulates his natural bending of a fretted note (notwithstanding whammy bar technique of course, which can hit both positive and negative phases). If this was a violin sound, we would approach the vibrato differently, but for synthesizer lead sounds, electric-guitar style vibrato sounds great.