[A6] What is fat anyway?

[Michael E. Caloroso]

Tuning drift is not the only attribute to a fat sound.

Fat is a delicate balance of ordered harmonics generated by overdrive to what sounds pleasing to the ear.  The distortion generated by analog circuits, however subtle or extreme, introduces harmonics to the timbre.  The degree of distortion is highly subjective.  You can overdrive an analog circuit in the oscillators, the mixer, filter, VCA, anywhere along the audio path.

Everyone knows that tube amps are great for overdrive.  The trumpet is another good example of natural overdrive.  Its timbre is dynamic with air pressure - soft timbre at low pressure, brighter with increasing pressure.  The alloys in the tubes, the flare, and the valves all play their part in the timbre of the trumpet.  The tubes are not entirely rigid and will resonate with varying pressure, which alters the basic timbre by adding harmonics.  That is why it gets brighter with increasing air pressure.

You can also introduce harmonics in synthesizers with room ambience, sound processors, multiple oscillators, or modulation.

Room ambience causes attenuation or peaking of selected harmonics with short reflections combining with the direct signals at your ears (or the mic in a recording environment).  Is your favorite synth starting to sound old?  Put it through a multitap delay with real short delay times to change the timbre.

Sound processors, which include the gamut from phasers to flangers to choruses, also have an attenuation or peaking effect on the timbre.  They don't have to be modulated either, a phaser set at a static setting will do the job.

Everyone knows how multiple oscillators add their degree of fat.  You can do this either with tuning near unison or with them tuned at selected intervals.  One little used technique is tuning them an octave+fifth apart with subtle balance to get a certain timbre (not just to simulate guitar barre chords either).  Again, with the right waveform, volume balance, and tuning intervals the combination of multiple oscillators can produce a timbre that is pleasing to the ear.  You can also get some overdrive from the mixer if that capability is there.

Now modulate those oscillators.  Pulse width modulation isn't the only trick.  One of my favorite tricks is to route just a slight amount of EG1 to only one of the VCOs for a real nice attack transient that sounds fat.  The sound resembles a trumpet section, because the pitch of a single trumpet is not constant in the attack transient it actually wavers.  Put two or more on unison lines and the attack transient has a chorus effect, which adds or reduces harmonics.

You cannot overdrive the circuits of a digital synth - it does not sound at all pretty.  Signal generation in the digital domain introduces zero overdrive, it is completely sterile.  Great for audiophile engineers but not always appropriate for music.  If you want the overdrive and addition of harmonics known to analog circuits, it has to be modeled in DSP code.  Which is not easy because the dynamic overdrive of analog circuits is not easily reduced to a mathematical formula.

MC