Entry 2: Noise/Tonal Splitting

If you're using a boring and moldy sample loop or even a dull MIDI instrument, why not take advantage of tools that are at your disposal and do something different from others to enhance your track. It might turn out to be pretty great! :)

It is as easy as splitting the audio into two different sections; Noise and Tone. The Tone aspect of the sound is a recording's harmonic content. The Noise aspect is what gives an instrument its unique sound, made up of its inharmonic content.

"Musical instruments vibrate naturally at several related frequencies called harmonics. The lowest frequency of vibration, which is also usually the loudest, is called the fundamental. The higher frequency harmonics are called overtones." (Source) There is also subharmonic content which is below the fundamental.

Click HERE to see my blog entry on Scalar EQ-ing.
A technique I came up with to improve tonal definition in your audio track!


The signal chain for the following is Guitar > Interface > VST Amp. Any processing was solely carried out on the DI signal.

Once you have split the signal into its Tone and Noise components you will notice that the Noise track is attack heavy and quite harsh, and depending on the source it may not be pleasant to listen to. However it's great when mixed in with the tonal component for good transient definition.
The Tone part is very resonant, so the first thing I'd do is gently EQ-ing it of a few dB (with a high Q) to remove any harsh resonances.

If you plan on compressing the DI, try compressing the Tone track quite harshly (+10:1 Ratio), and then a 4:1 Ratio with a fast attack/release on the Noise track. That way you always have a tonal base on which the noise can go on. This should keep transient definition that could be otherwise lost in compression.

It is very subtle, but there is a difference - Both audibly and visually!

Notice the more defined attack and lesser sustain, but more natural decay in the POST track.

The Audio Sample

Even on the low-definition spectrograph of Soundcloud you can see the better transient definition.

There is no processing in the Post-splitting DI aside from very minimal EQ-ing, to decrease resonance, in the Tone track which is blended with the Noise track. Then, the PRE and POST tracks are sent to the same amp-simulator on an AUX track.

Emilio M. RizziComment