Pages - Menu






wtorek, 14 stycznia 2020

Matthew Skala North Coast Synthesis Interview ''I try to make every module unique''

Matthew Skala is the founder of North Coast Synthesis Ltd. from Toronto in Canada.
Ph.D. at the University of Waterloo Specialization in multidimensional geometry and its applications in areas such as security, database search and computational linguistics. He spent about 15 years on academic research and teaching at universities in Canada and Denmark. In 2017 he founded his own company in Toronto-North Coast Synthesis.
North Coast Synthesis introduces the art of hand electronics to Eurorack modular synthesizers. His products are intended for serious hobbyists and DIYers, with an emphasis on top-quality components; excellent technical openness and documentation; and long-term ease of maintenance.

For this Matthew leads powerful entries on his website about the components used in eurorack their use and security, and extensive arguments about modular synthesis and its use in ambient music and more. I am excited that I was able to ask Matthew about a few issues related to the production of his modules, his inspiration and how his North Coast Synthesis Ltd. company looks like today.

Can you say something about yourself North Coast Synthesis?

North Coast Synthesis is a one-man synthesizer manufacturing and consulting
company, so I do everything, but when I get the choice I usually say my job
title is "Module Designer."  That's the role I most enjoy.

Your articles are extensive and read very engagingly, what inspires you to convey this knowledge?

This market is competitive and I'm always looking for an edge - what can I
do that other companies aren't doing, or that I think I can do better?
Since I have some technical writing and teaching background from my previous
life in academic computer science, writing technical articles for the Web
site seems like something that my competitors can't easily imitate.  There
are other, more personal motivations, too.  I always wanted to have a
widely-read Web log, and I more or less had to give up on ever gaining a
large audience for my personal site.  People are reading the articles on my
business Web site in a more serious way and that's very satisfying.

A lot of articles are focused on SDIY, for example the title how to get homemade fluxes, is it because you do many DIY Kits yourself that you can buy from you and you want to develop DIY even more than assembled modules in the future?

I see myself as a hobbyist who has "gone pro," and the customers I'm
interested in selling to are hobbyists too, whether they're electronics
hobbyists or music hobbyists or somewhere in between.  DIY fits naturally
into that.  But I want to sell assembled modules, too; lots of musicians
don't want to do electronics themselves.

There are of course many analog circuits you mentioned about describing 14 parts about synthesis and its elements in a modular system. but what inspires you as electronics, are they your own imaginary circuits, are there still synthesizers that captivate you and modify their circuit?

I like many of the early digital synths - the Roland D-50 being one of my
all time favourites - but I'm not much interested in directly imitating or
modifying their circuitry.  I try hard to avoid having any of my products be
"clones" or closely similar to other things that have already been done.

So as you see the future of DIY modules, do they have the chance to grow and get more attention?

It's difficult because of the limitations of what people can build at home.
Present-day electronics is very heavily oriented toward robotic
manufacturing on a large scale.  It's hard to even get basic parts like
through-hole transistors anymore, and that's only getting worse over time.
Some people are going to be disappointed if they want to build something
just like the latest mass-produced product and they discover that the parts
for it, if you can get them at all, aren't suitable for hand assembly.  So
we need to push DIY as its own thing, not as a way of getting products just
like what come out of the big factories, and especially not as a way to
"save money."  It's more than that.

Of course there is the whole "maker" thing going on, and that may be
growing, but "maker" is usually more about people doing their own
programming of mass-produced complete products instead of really building
circuits at the level of single components.  If custom small-run IC
manufacturing, like the old MOSIS, becomes available in a format like
today's PCB services, there could be some very exciting hobbyist
opportunities there.  And I think that's possible.

The DIY environment is constantly evolving but what do you think of the knowledge they have, do you see any definitive errors in the fcompany starters or just DIY guys?

I think many people underestimate the difficulty of stuff like mechanical
design, sourcing the parts, and convincing people to actually pay for the
product.  That last one, the marketing, has been much harder for me in
particular than I expected, and I'm worried about running out of money as I
try to build sales to the point where I can pay for food and rent.  I'm not
there yet.

There's a perception that developing a product begins and ends with drawing
a schematic - and as a consequence the PCB embodying the schematic is the
only part on which a manufacturer deserves to make a profit - and there's a
lot more to it than that.  But one of the good things about modular synths
as a business is that there's not much cost to get started.  It's easy for a
small player to start up and make some mistakes before they need to be
really serious about making money.

By developing the topic of the modules themselves and their components, you focus on the highest quality selected parts and even urge you to create some elements yourself, does it have a big impact on how long the modules last without failure?

Ask me again in 30 years!  So far as I know none of my products have ever
worn out, but none of them are very old yet either.  I certainly hope that
my modules will last, and it's one reason for using expensive high-quality
potentiometers, which are a point of wear for many modules.  But quality
components is also a business decision:  because I don't have the huge
factories and cheap labour that a big company with overseas manufacturing
would have, it's clear to me that I can't compete on price against the
big companies' products.  I have to target the higher end of the market in
order to survive at all; and then using the best components is part of how I
justify my prices.

I do think that using analog and through-hole makes my modules easier to
maintain than a lot of the current generation of digital stuff.  It's
possible to desolder and replace a bad through-hole component.  So in the
long term, even if it does break, I hope one of my modules can be repaired
and not thrown away.  Publishing the designs contributes to that, too.

It seems your flagship products have two filters MSK 007 Leapfrog VCF and MSK 009 Coiler VCF, can you tell both of them and tell the difference between them?

In both those modules I wanted to create filters that would be really new
and different.  That's a design challenge because people have been building
analog synthesizer filters for a long time and many ideas have already been
thought of.

The Leapfrog VCF is based on the leapfrog topology - which wasn't invented
until after the synthesizer industry had already started to go digital, so
it was never a big thing in mass-produced analog synths although it's well
known to analog designers who work on other electronics.  My module in
particular makes use of leapfrog topology to create a very sharp filter
cutoff, much sharper than we usually expect for a synth filter.  That gives
it a unique sound that I think of as smooth, or rounded.  But the design
originated in a hobby project that I was building for myself, before I
started the company and had to think about making my designs suitable for
commercial manufacture.  As a result, the Leapfrog VCF is not what I would
call a very "commercial" design - it requires many different component
values and a lot of adjustments and the resulting module is large and deep.
Those modules take a lot of time to build and I have to charge a price for
that time.

The Coiler is my attempt to design another unique-sounding VCF that will be
easier to build, both for myself and for my DIY customers.  Since it was
designed as a commercial product right from the start, it's a simpler
circuit, and I can afford to sell it at a lower price.  It's basically a
two-pole multimode state variable VCF, just like many well-known synth
filters.  Its cutoff slope is much shallower than the Leapfrog's and more
like what people expect from a classic 1970s analog synth.  But it uses
inductors as the main integrating components, with capacitors also
used at some frequencies, and that gives it a flavour of its own.

Also the other modules are quite impressive, for example your MSK 010 Fixed Sine Bank which is the first LFO I see without knobs, can you tell me more about this module?

That was the first module I started selling commercially, and its origin was
simply that it was the module I wanted for myself.  When I first started as
a modular hobbyist I was making a lot of drones, and using a Pittsburgh LFO
module with a triangle output to generate control voltages, and I found that
I could always hear the corner when it would change direction.  The triangle
wave was no good for a drone parameter that I wanted to fade into the
background.  I really wanted sine wave LFOs - and preferably many of them,
to control multiple parameters.

I'd gotten some good advice about not spending all of my budget in my first
synth purchase, so I had some money set aside to buy the module that I would
decide was missing once I started playing.  For me, that module was an
Intellijel Dr. Octature II.  It's a nice filter, but more importantly it
can run as a sine-wave LFO with eight outputs.  The disadvantage is that
those eight outputs are just eight phases of the same sine wave, so if it's
generating control voltages, it doesn't really generate as many combinations
as we might think.  This was a step up for my drones but still not perfect.

That led me to developing the Fixed Sine Bank.  It generates clean smooth
sine waves with no sharp corners, so it's easy for things it controls to
fade into the backgound.  It also generates eight waves independently; so it
can keep things moving in even a large patch.  But you can't change the
frequencies.  You can choose between eight frequencies by plugging into the
diffferent outputs, but those are all you get.  There are no knobs, no CV
inputs, and no trimmers on the back.

This module really upsets some people.  They seem almost personally offended
that I would presume to offer an LFO for sale with no frequency control.
But the fact is that it's one of my best-selling modules.  It was what I
wanted for my own rack and I guess other people also want to do the same
kinds of things I want to do.

The issue isn't that I really wanted to specifically not be able to control
my LFO frequencies.  Sure, if it cost nothing I'd love to have a little knob
for adjusting each output.  But when I looked at the things I wanted in my
multi-LFO module, I realised that frequency control wasn't a priority.  If I
wanted frequency control, it would mean big compromises on other things.  I
wanted sine waves; I wanted to do it with analog through-hole construction;
and I wanted a lot of them.  Each of those goals is much harder to achieve
if I also want to be able to change the frequency.

An analog sine oscillator with an adjustable frequency is a much more
complicated circuit.  I was looking at either adding an expensive dual-gang
pot to each oscillator, or building them as oscillating filters (more than
doubling the complexity) with voltage control.  It's easier in the digital
world but that has its own tradeoffs.  By giving up on variable
frequency I was able to get all my other wish-list items.

It's certainly not a module for everyone, nor the only LFO module someone
would ever need, but I use mine a lot and I've been pleased that others have
found uses for them, too.

I own your octave switch myself and when you look at this module it seems like a regular octave switch but it is a multifunctional module for cv and for example a quantizer, what determines how you develop a module which has hidden functions under a simple name?

I go over each module idea many times looking for opportunities to get more
value out of it.  With the octave switch in particular, it started out with
the focus on precision adding because that was something I wanted for my own
rack.  I built a precision adder and slew rate limiter module for myself,
but I wasn't happy with the way the slew rate section worked, and I became
aware that octave switching is an issue for many people who don't like
controlling oscillators with just knobs, and so on.  I combined them and
then "octave switch" ended up being the name of the module.

The rectifier input on the Coiler VCF is another interesting example.  I
designed the filter core and found that it needed seven op amp sections.
They come in sets of four, and I didn't want to just put in two chips and
leave one amplifier unused.  So then I had a design challenge: what can I do
with one op amp and not much other circuitry, that would create a nice extra
for the multimode filter module?

Many companies I talk to often say that they create their own system, and they make modules that they want in their system, they fill the gap ... are you also going to make a bigger line North Coast Synthesis , your own MSK line for the full eurorack system?

I'd really like to have a complete line so that someone could build a
synthesizer from only my modules.  That would be a big achievement for the
company, if I can remain in business long enough to do the development.  I
figure the three things missing are a VCO, a VCA, and some kind of sequencer
or controller.  I hope to do a VCO as my next product; I'm at the breadboard
stage for that now.

The VCA, I'm not sure.  I have a complete VCA design posted as a plans-only
DIY project on my Web site, so the diehards can already download that and
have a North Coast-designed VCA module if they want.  But it's not a very
commercial design.  It has exponential-only voltage control, which is not
what I think people most want; and I'd want to re-do it if I were making a
real product out of it.  One issue is that it's very hard to convince people
a boutique VCA is worth more than a mass-produced one.  Oscillators and
filters are thought of as having their own unique sounds, so people will pay
extra for a good VCO or VCF, but VCAs are perceived as all the same, and
then everybody just wants the cheapest one and I can't really make mine be
the cheapest one.  So it has to have some kind of gimmick that makes it
obviously special.  My existing VCA design has the gimmick of being all done
with discrete transistors, but the disappointing sales of my other
discrete-transistor modules have convinced me that discrete transistors are
not actually something many people want to pay extra for.

As for the sequencer or controller, I've thought about some ideas but have
not found one I really want to run with yet.

That's great, so how long does the research and tests at your level take to get the module you are completely satisfied with, sometimes I see young companies praying from prototype to the final version in 4-5 months, how does it work for you?

I do a lot of development before I ever build a prototype.  Usually, I'll
think about an idea for a long time before doing even a sketch of a design
on paper; then I may make several sketches, and play with transferring them
into the computer schematic-capture program.  Often, after I have a rough
schematic, I'll get interested in something else, set those computer files
aside, and then if I come back to that module idea I start over from

Once I get more serious about a given module design - in particular,
deciding that I do actually want to make a product out of it - I'll probably
run some simulations, and build it on a breadboard, before doing a
prototype.  To actually do a prototype I need to design a PCB and panel and
get them manufactured, as well as sourcing parts, so it's a fairly big
commitment of money.  Before that point, development doesdn't cost much
except my time, and the time is easy to spend because it's my favourite kind
of work to do.

So depending on what counts as the start of the project, it might be six
months to a few years between when I start on a module design and when I
actually build the first prototype.  But to answer your question about the
step from prototype to final version, that could be four or five months but
it would usually be more like one or two, because I hope to have already
done most of the development before getting to the prototype.  From the
first prototype I only need to deal with any issues that arise in testing
it, write the manual, do some mechanical drawings, and then I can go into
production by building more like the prototype.  I think companies that
outsource their manufacturing, as most do, probably need more time at this

What you think makes the North Coast Synthesis modules different?

I try to make every module unique - nothing a clone of anybody else's
product - but technical openness is another important point.  I publish
detailed plans of each module including a circuit explanation so that
whoever buys it can hope to understand how it works.

In your entries there is also often a mention of ambient music is it because you are potentially involved in this genre?

Yes, a lot of the music I like to make myself is of an ambient or New Age
style.  Some of those sounds are difficult to do with analog.  That provides
a source of design challenges, but I also use digital modules from other
people, and software synthesis on a PC.

Are you planning new modules for 2020?

My next project is a VCO and I hope to have that available this year.  I'm
breadboarding parts of it now.  The feature set is not solidly decided yet
but I'm imagining it as a sort of "complex" oscillator, with two analog
oscillator cores in the module, a shaper based on the Gilbert
triangle-to-sine circuit for a number of wavefolding and through-zero
effects, and some kind of digital tuning feature which might also allow for
special sync modes.

More info about modules and You can find many great articles about modular synthesis and DIY construction on the official website North Coast Synthesis here:

Designed By Blokotek