[sdiy] VC GLIDE and ADSR

[Michael Bacich]

On Apr 21, 2006, at 8:39 AM, Cornutt, David K wrote:

>  >  Glide and slew limiter. What's the difference?
>
> As I understand it, a conventional glide circuit is basically a two- 
> pole
> low-pass filter, so its response is that of a discharging RC circuit:
> when the input changes, the output changes rapidly at first, and then
> more slowly as it approaches the input voltage.  A proper "slew  
> limiter",
> in my mind, should be linear in response: When the input changes,
> the output should change at an absolute maximum V/sec until it
> reaches the input voltage.  (Of course, if you use this on a pitch CV
> in a V/oct system, it will give you a non-linear response with respect
> to pitch.)
>

I tend to agree with you that something called a "slew limiter"  
should probably have a linear time/voltage change response curve, but  
the words "glide" and 'portamento" don't necessarily connote any  
particular type of time/shape response.  In the world of existing  
analog synth designs, you will find either of those words used to  
describe circuits that give both linear and log responses.  For  
instance, the Minimoog, which has a decidedly linear portamento  
response, calls it "glide".  Other Moog synths, such as the Moog  
Rogue and MicroMoog, which have decidedly log portamento shapes, also  
call it "Glide".  The Prophet V and Sequential Pro One also have  
linear response portamento, but call it "Glide" on the front panel.   
The TB-303 has a log glide, and the front panel calls it "slide",  
probably in reference to the way a bass guitarist sometimes "slides"  
between notes (the Roland engineers, in their wildest dreams, never  
imagined that the 303 would ever be used in the way that eventually  
made it iconic.  They were just trying to help low-budget lounge  
musicians replace the bass guitarist)

It may be interesting to note that with the human voice and with  
instruments such as trombone, the violin family, slide guitar, pedal  
steel, and fretless bass, that do not have mechanically imposed  
sharply delineated pitch points, the tendency is almost always to  
slide between notes with an approximation of a logarithmic portamento  
curve -- we slide quickly toward the target note, and then gradually  
and more slowly settle into the final pitch as we get closer.  It's  
subtle sometimes, but it's almost always there.  indeed, it's  
virtually impossible to sharply and immediately jump from one  
accurate pitch to another when singing.  Go ahead, try it -- if your  
ears are worth a damn, and you're honest with yourself, you'll see  
what I mean.  Conversely, it's almost just as impossible to  
approximate a true linear glide with your voice (if you're trying to  
start and stop exactly of specific pitches).

The fact that pitch correction software algorithms, such as Antares  
Autotune, are able to eliminate this momentary human portamento (even  
with "singers" who, unlike Britney Spears and her ilk, actually CAN  
sing in tune -- for example, Shania Twain, who is a seriously good  
vocalist, yet for some aesthetic reason still insists on using  
Autotune on every track she records -- I suspect because it somehow  
heightens her image as the postmodern, preternaturally hyper-sexual  
uber-cowgirl robot queen, not unlike something one might encounter on  
the ubiquitous video screens of the oversexed dystopia described in  
Aldous Huxley's Brave New World) -- this unnaturally eliminated  
portamento gives vocals that are recorded with the software part of  
its characteristic sound, especially if you turn the software's  
"speed" control to its fastest setting (yes, there are other factors,  
too).  It enables humans to do the impossible:  make their vocal  
tracts act like frets on a guitar.

Robin Whittle (he of Devilfish 303 fame) once speculated that part of  
the 303's natural appeal to humans is because the sound of the glide  
was very reminiscent of the sound of a baby animal's cry.  I would  
probably agree with him, but I would simplify that to say that we are  
simply naturally attuned to the sound of a non-linear, logarithmic  
pitch change.  It's just how nature sounds  Linear glide isn't, nor  
are linear-shaped envelopes, for that matter.

BTW -- Keith Emerson's iconic solo at the end of "Lucky Man"?  Log  
glide.  In my opinion, that solo would have sucked with linear  
Minimoog-type glide.

Mike B.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20060421/3bfd4394/attachment.htm>