Instrument Design Journey: Natural Gate, Part 4 - Materials

There was an extra bit of work required to get the new envelopes fast enough.  It turns out that of all the op amps I tried, they simply didn’t have enough drive current to charge up they heavy instantaneous load of the new envelope circuits.  And so I wasn’t getting the hardest of materials from NG even though the curves were very close.  Zoom in and I found an unwanted rolling off of the very fast attack portion of the envelopes.  And it was absolutely audible!

So I had to build in a current driver that boosts the available drive current of the HIT/CTRL stage.  This worked rather well and just like that the characteristic ‘hitting something hard... hard’ could now be heard.

If you use one shape, you have one material.  And the material I had was a hard one.  But what about softer sounds?  NG can be very sharp and cutting (BTW, we find NG is easy to mix around - it cuts quite well in a mix!) but sometimes you do not need or want this.  Sometimes you want things that are a bit more subtle. Most of this lies in the attack portion of the envelope. And so we decided that altering the shape of these envelopes further would give us some different material options.  This includes rounding over the attack portion as well as reducing the overall height of the envelope signal.  Because NG is very 'filtery,' this does indeed mean softer where 'soft' means both quieter (volume) and roll-off (spectral content).  We were tempted, initially to normalise the volume so that there wasn't such a big jump between materials but after a quick test decided this was a bad idea and that a more filtered signal, even at the peak of the softest material EG, sounded like it should - soft.

On that note, keep in mind that the CTRL level also affects the vertical positioning of the envelope signal inside NG.  So you can get some in-between settings.  This is why CTRL input is normalised to a DC voltage.  We sometimes bias it a little higher and sometimes we pull it back for very dynamic velocity-based strikes.  Try it out!  In fact, NG is calibrated such that the CTRL input isn't fully 'zero' but since the closure of NG is "a lot," we still have some room to play with how the tail can be biased up or down.  And this is exposed to you via CTRL.  Play with CTRL bias -  subtly - to see how it affects the sound.

Stay tuned for Part 5 "Memories of Closure" which will be the final installment of the NG Instrument Design Journey!

Copy blog RSS feed url here


Share this post