[sdiy] Log and Lin Controls in a DIY synth - Where to use?

megaohm megaohm1 at gmail.com
Fri Feb 3 16:44:05 CET 2006


Disclaimer: (same as Jure's)

Actually, most common EGs DO use log pots for A,D,R. I have built one with
linear pots and the feel is not-so-good. Most of the range will lie in the
first quarter or half throw.

Whether a VCA is lin or exp has no bearing on the style of pot to use.

Linear and Exp inputs on VCOs? Again, those names have NO bearing on what
pot to use (use linear for both...at least I do).

Common places to use a log pots are:
A, D, and R of an envelope generator
Slew amount of glide/slew/porto module
Audio input attenuators

You can use linear (or log) pots for everything. It will only effect the
"feel" of the controls.
It is awfully hard to get smooth feeling audio fade-ins with linear pots,
but it's not a huge deal.
In fact, there are only a handfull of log pots in my synth (and none of
these are for audio!).

You can also shape the response of a pot with clever use of resistors.
There's a great write up on the Elby website (I think that's where I got
it?) , may be called "tailoring pots" or something like that. Can be very
useful for shaping bipolar controls so you can actually find the "zero
point", or center... as in no modulation. It can be a bitch to do
otherwise.If it can't be found I'll send it to you.

Best idea, don't rely much on what is told to you. Buy some of each kind and
swap them in and out while prototyping your circuit. That's the best way to
figure out which gives you the best feel.

peng



On 2/3/06, jure zitnik <kokoon at gmail.com> wrote:
>
> disclaimer: everything i'll write is purely by common sense. i might
> be wrong with something and the terminology is probably completely
> whacked since english isn't my first language.
>
> i don't think you'd like A, D and R controls logarithmic - logarithmic
> means that the function value grows faster at higher input values - in
> practice - if you had logarithmic attack it would mean that the
> resulting time difference between 0% and 50% of attack setting would
> be MUCH smaller than between 50% and 100%. it would make setting
> longer attacks accurately much more difficult than shorter ones.
> however it might just be what you're looking for. maybe you care about
> precise setting of short attacks and not for long ones.
>
> i hope i helped. i'm not 100% sure about VCAs being typically linear
> but my common sense says so. the deal with tone pitch as percieved by
> human ear is that with higher frequencies you need bigger frequency
> differences to *hear* the difference (in pitch).
>
> but that's not the case with amplification of audio signal versus the
> perception of loudness - in fact it's the opposite. with louder sounds
> you need to change the amplitude (amplification) LESS to notice the
> same relative difference in loudness than with lower amplitudes.
> that's why audio mixers have non-linear channel volume but inverse
> logarithmic sliders... which could ofcourse be normal logarithmic
> sliders if used for attenuation instead of amplification...
>
> but probably none of these subjects are new to this list and i'm
> probably excersizing it only for myself :) i'm sure someone will
> respond and write exactly what is the most common configuration.
>
> cheers,
> jure
>
> On 2/3/06, Chris Manders <wight446 at yahoo.com> wrote:
> > Hi all
> >
> > Thanks Jure and Yves for your replies.
> >
> > I am understanding this now. I was under the mistaken impression that
> VCAs
> > only had Log inputs. I will be building a Linear VCA so won't need to
> fiddle
> > about.
> >
> > One final question - In a basic ADSR circuit, I assume that the Sustain
> > control would be Linear (to describe the voltage level at Sustain) but
> the
> > other three controls are Rate controls and therefore would be
> Logarithmic.
> > Is this correct?
> >
> > Thanks
> > Chris
> >
> >
> > jure zitnik <kokoon at gmail.com> wrote:
> > i'd say linear for VCA. but log for anything pitch-related.
> >
> > most modulars have 2 separate inputs for VCO modulation - lin and log.
> >
> > jure
> >
> > On 2/2/06, Chris Manders wrote:
> > > Hi folks
> > >
> > > I am going back to my old love, Synths after a brief absence.
> > > About 6-7 years ago (seems like yesterday) I successfully built a Moog
> > > ladder filter and a VCO using help from some of you guys out there.
> > > I am going to start collecting up bits and bobs for continuing this
> > project
> > > of building my first Analog(ue) synth.
> > >
> > > This will probably be a real Newbie question to you guys out there,
> but in
> > a
> > > typical synth which Potentiometer controls should be Linear, and which
> > > Logarithmic?
> > >
> > > I understand what Log and Lin are, and this may be something I just
> need
> > to
> > > get my head around to fully understand it.
> > >
> > > Am I right in saying that if a control describes a voltage level (CV
> to
> > > Pitch being 1V/Octave), then that control should be Linear. All the
> rest
> > > should be Log (as for example Amplitude being a Log scaling)?
> > >
> > > My head starts to ache when I think of other routings. If a Linear LFO
> > level
> > > control is set to sweep up and down by 1 Volt, when fed into the VCO
> it
> > will
> > > modulate the pitch up and down by an octave. Would this LFO signal
> then
> > have
> > > to be converted to a Log scaling if one wanted to feed it to the VCA
> for
> > an
> > > even tremolo effect?
> > >
> > > Thanks
> > > Chris
> > >
> >
> >  ________________________________
> >
> >  What are the most popular cars? Find out at Yahoo! Autos
> >
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20060203/6595afdc/attachment.htm>


More information about the Synth-diy mailing list