[sdiy] Digital ADSR - perceivable staircase?

[Scott Gravenhorst]

>On 15 Feb 2011, at 21:07, David G. Dixon wrote:
>
> Think about it: 16-bit accuracy for a lousy envelope!  Talk about hammering
> a nail with a pile driver!

I tend to view digital techniques as an approximation of analog techniques.  The smaller the steps
used (both amplitude and time), the better the approximation.  To the point that (as Tom said), as
the resolution is increased, the difference eventually becomes undetectable (at least it is to me).

And by now, since fast 16 and 32 bit processors are available for "walkin around money" - gone are
the days when we are forced to use slow 8 bit processors where we would add external analog
hardware (which could easily make configuration difficult) to make up for the lack of resolution. 
So using a few more bits to get better approximations is simply using the best of the selection of
tools at hand.  I'd hardly call it "hammering a nail with a pile driver".  Also, I wouldn't call an
envelope "lousy", as if it's not important enough to warrant such effort.  In my view, the shape
and accuracy of an envelope generator's output is every bit as important to the audio quality of a
synth as is it's oscillators, waveshapers and filters.

As I previously stated, my envelopes using a quite modest FPGA are generated using 28 bit
arithmetic.  And even with such "wasteful" design work, I'm able to design 8 and 16 voice
polysynths that run at sample rates at least double that of CD quality.

-- ScottG
________________________________________________________________________
-- Scott Gravenhorst
-- FPGA MIDI Synthesizer Information: home1.gte.net/res0658s/FPGA_synth/
-- FatMan: home1.gte.net/res0658s/fatman/
-- NonFatMan: home1.gte.net/res0658s/electronics/
-- When the going gets tough, the tough use the command line.