[sdiy] User interfaces (was: bunch of other stuff)

[David Cornutt]

On Apr 22, 2006, at 5:44 AM, Rainer Buchty wrote:

> Maybe it's because they are lacking the proper user interface?
> Personally, I'd for instance rather use a Korg Polysix than the
> Poly61.

Good point.  Part of the bad rap that digital synths get is from the 
crummy
one-knob user interfaces.  I've had an E-mu Proteus/2 for years, but up
until a couple of years ago I had never done much with it because
of the difficulty of paging through the menus.  Then, one day, I sat 
down
and built up a set of controller definitions in Metro so I could edit 
most of
the parameters on-screen in Metro's mixer window.  Once I started 
playing
with that, I found out that the Proteus/2 wasn't just a rompler and it 
could
actually do some things.  (Although it doesn't have any filters...)

>
> Although I'm not sure how a proper user interface for a digital synth
> should look like -- citing DX7 again, there was the Jellinghaus
> programmer. Far away from being a commercial success.
>

There is the JD-800 as a good example of how it can be done.
(And as a bad example on an analog synth, there's the Matrix-6R.)
Of course the big deal with that is knobs and buttons cost money.
Once synths had microprocessors on board anyway, manufacturers
eventually figured out that the processor could do layered menus
on a one-line display, and they could save a bunch of money on
the user interface.  In a way, it became its own worst enemy: it made
the synths a lot less expensive, so more people could buy them,
But then the buyers found out that they were really difficult to work 
with,
and since a lot of them were first-time buyers anyway, it became
overwhelming and they got a bad rep.

Of course, the JD-800, although not really intended to be a VA,
is still based on the subtractive-analog paradigm.  And that brings
up the other difficulty with the DX7: figuring out what the parameters
actually do.  We all know from experience that if we move the cutoff
point on an LPF, we have a pretty good idea what it will sound like
before we do it, and so if we're going for a particular sound, we know
how to get there. When I started with Yamaha FM synthesis (with a CX5M
actually), I had to sit down and generate literally hundreds of patches
and variations before I started to understand what, sonically, the
parameters did.  It took me quite a while to get to the point of
"if I'm using algorithm 4 and I set the index on operator 1 to this,
it will sound like this".  I get the impression that the synth community
at large never really got over this hump with the DX7, and that's
one reason that Yamaha eventually turned away from it.