[sdiy] Control Interface Angst </ Wakeman>

[phillip m gallo]

Both this thread and the prev. "microcontroller for synth" thread caused
another philosophical eruption of the consideration for "instrumentalist
interface".

perhaps it's my chronological age, or just a historical familiarity, but i
too consider the rotary "knob" a very efficient form of programming language
for synthesizers.  As noted, it does provide a good measure of
self-documentation.  Rotaries seem to be "idiomatically" efficient,
intellectually coupling mechanical action to end result.  Also a row of
rotaries can be manually "wiped clean" in a single simple (if somewhat
dramatic) gesture.  Now this is not to say i do not like other control
mechanisms (though little 50mm sliders are not high on the list, apologies
to my MG-1).

Beyond the generic suitability of rotary controls i must also include a
personal failing which is a lack of intellectual and visual acuity which
make ranks and banks of knob's, indicators and switches, meld into a "vague
structure of un-identifiable landscape which i find confusing and not all
that useful from an instrumental point of view.  Those instruments which
partition and off-set position controls (vs. grid line ups) are far easier
for me personally to use.

As recent technology is brought to bear upon synthesizers, high levels of
integration allow the condensation of what where multiple modules into
single modules.  However the self-documenting nature of rotaries gets lost
in the shuffle.

Imagine a multi-VCO module.  Do you really want multiple FZ/MOD/PWM/Wave
controls or do you select Osc A, set the control's and either "dupe" these
controls to the companion devices or select Osc B and perform the same
action.  In either case do you provide an LCD with graphical indication of
the current setting of the OSC in consideration or do you implement
motorized faders and auto set the control's. If you use a LCD where do you
stop and what do you represent the parameter (using a 128x64 pixel LCD) or
the rotary position?

Perhaps you've worked with control surfaces which do not used motorized
control's, often when you use a control correlated to a device parameter you
first move the control to "match" the current setting then move to the new
setting.  Others you just move and accept the new parameter and audibly
match which is frustration when you just want to "bump" the setting.

The PC GUI mouse driven pseudo-knob is personally frustrating and not again
"idiomatic" or necessarily "effortless".

regards,
p








Rude 66 :

> heh.. and a knob is still both the easiest way to modify data, and also
very
> important, you can see in one eye what all your parameters' levels are.
one
> close look at an analogue synth and you kinda know what the sound is.
>

Well, this is partly true... As long as data to be modified is ADSR
slope or filter cutoff freq. or FM parameter (in other words : one
parameter at a time), that's true.
But what about granular reverb, granular FFT, VC additive synthesis
and other weird / complex things requiring several parameters to be changed
simultaneously ?
Yesterday (or so) we tend to agree that adding complexity could be a way
to inject new blood into existing synthesis techniques. In that case, are
knobs still efficient ?


Rainer :

> Whatever synthesis you go for, you'll finally come to a point how to enter
> the data in a most user friendly way. Which brings you back to the knobs
> being arranged in a (hopefully) most intuitive way.

> Assuming, that you actually can map the synthesis parameter to terms which
> the average musician understands (I actually doubt that e.g. more people
> would have gotten into deeper understanding of FM if Yamaha delivered the
> DX7 with the Jellinghaus programmer...)

Well, let's put this discussion back on tracks : I'm not searching for a
striking new synthesis (don't have the level / knowledge in maths nor
physics)...
My original approach was : there are quite a few dev. tools (sound, 3D)
available to DIYers these days. So instead of moaning about the lack of
innovation among big synth companies, why not use this list, our knowledges
and experiences, and our enthusiasm as DIYers to brainstorm about new
ways to approach electronic instruments design...


> When searching for a striking new synthesis this is IMO the point which
> has to be kept in mind: how useable is it for the non-technical oriented
> musician?

Good question. And actually could it be that the lack of success of synths
such as the K5000 or FSR1 was due to a lack of useability for
non-technicians ?

Does a musician think in terms such as : now I'm going to combine FM
with formant synthesis ?
Hence my suggestion to provide end-users with a set of tools to build rules
for their own musical world (that would combine bits of all existing
synthesis
techniques, but users won't care about which technique they use).
Yes, this has more to do with AI (not the Spielberg's movie) and computer
aided design. But the nice thing is that such techniques can escape labs and
are now available to DIYers...

JB


---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.497 / Virus Database: 296 - Release Date: 7/4/03

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.497 / Virus Database: 296 - Release Date: 7/4/03