Author |
Message |
unshaven
Joined: Feb 18, 2012 Posts: 8 Location: Germany
|
Posted: Sat Feb 18, 2012 11:14 am Post subject:
Analog polyphonic - topology question |
|
|
Hey there!
My first-time post here. From what I've seen this would be the sub forum best suited to my question, correct me if I'm wrong.
Since it's customary to introduce oneself when new in a forum, I guess I do that here, briefly.
Just jump to "The actual question", if you don't care
I'm a software developer, and it would be kind of natural to build software synths, heh, but besides the math for *really good* softsynths being in the way of me achieving results anywhere soon, and I really dig the sounds of good old analog and some "hybrid" synths, I recently started to reasearch a bit about this subject.
I have pretty basic electronics knowledge, done DIY stuff in my youth, and just recently acquired some gear again to be able to revive this forgotten hobby of mine, the main motivator being synths, and the fact that I'm working at an embedded SW/HW company for a while now, which kind of sparked the interest in getting the hands dirty in hardware again.
I'm making music, recording with a computer, for myself, since I'm 14 or so. I have no musical eduation nor am I a synth expert not even from the user side, although I have some understanding of how components achieve their effects in the frame of subtractive synthesis.
I have made a modular soft synth already, but not state of the art, and I'm especially dissatisfied with the filter sections - implementations which I copied from free internet resources and which at best sound okay, since as mentioned, I lack the math skills to do much more.
I never owned a hardware synth, but the last 80's analog synths and hybrids like Juno series seemed especially attractive to me.
My plans, so far
I am now planning to first experiment with different topologies and points on the digital---analog line, and when gained some experience with different approaches, I want to build a bigger, complete hybrid synth.
Hybrid because, it saves space & time to build everything that does not critically contribute to the sound characteristics, in software - for me at least, since software is my game.
Hybrid can mean anything from Ensoniq digital osc + rest analog style, Juno style, to all-analog VCO & remaining signal path but digital for envelopes and all controlling.
The actual question
As I understand, the principal topology of all monophonic analog synths I've had a glimpse at, dictates that the VCA comes as last part in the chain, fed by the VCF output, to create the envelope.
I sometimes wonder why - is it to reduce noise, as opposed to, e.g. VCF behind VCA? Since that way around, it would more closely emulate natural instruments. Say, a guitar, has the sound source, the string, and the body as sort of formant filter. The envelope is shaped by the decay of the vibrating string itself, which is clearly before the filter.
This has implications, of course, and may be the reason why synth sounds like synth, not like acoustic instrument. Things come to mind like, a sound being always saturated (when the filter is saturated), no matter of point on the envelope curve (unlike an electric guitar, which will distort less as the envelope is near its end).
What's now confusing me is:
When it comes to polyphonic synths - from what I recall, polyphonic synths, at least cheaper ones like Poly 800 or so, do not have an own VCF for each of the voices that can be played in parallel. But for true polyphony, each key would have to be able to start with a different "velocity" than other keys, right? Which implies that each key needs an own VCA? But VCAs come usually after the VCF - but if there is only one, this does not work, and there it comes, I'm confused
The only answer I see is that those synths simply do not follow the VCO->VCF->VCA route - if that's so, I wonder why it's so pevalent in mono synths (or is that impression mistaken?).
Thanks in advance for enlightening comments,
- Steve |
|
Back to top
|
|
|
Dougster
Joined: Sep 20, 2005 Posts: 272 Location: Tucson, AZ, USA
|
Posted: Sun Feb 19, 2012 11:22 am Post subject:
|
|
|
As we say in car racing, "Speed costs money. How fast do you want to go?"
To change that phrase for this particular discussion, polyphony costs money...
A long time ago, some people (Bob Moog and others) were trying a lot of different topologies, but the VCO -> VCF -> VCA architecture turned out to be the most generally useful for most musicians, so that's what most manufacturers started making. You don't have to limit yourself to that design though...
Regards,
Doug _________________ Once you start down the modular path, forever will it dominate your destiny!
Every DIY person should own a copy of Electronotes: http://electronotes.netfirms.com
Blue LEDs are evil. |
|
Back to top
|
|
|
unshaven
Joined: Feb 18, 2012 Posts: 8 Location: Germany
|
Posted: Sun Feb 19, 2012 11:49 am Post subject:
VCF ICs - china sellers any good? |
|
|
Dougster wrote: | As we say in car racing, "Speed costs money. How fast do you want to go?"
To change that phrase for this particular discussion, polyphony costs money...
A long time ago, some people (Bob Moog and others) were trying a lot of different topologies, but the VCO -> VCF -> VCA architecture turned out to be the most generally useful for most musicians, so that's what most manufacturers started making. You don't have to limit yourself to that design though...
Regards,
Doug |
Heh, it just dawned on me that, synths which do not have one VCF per voice, cannot trigger an own filter envelope on every depressed key in poly mode.
I'm not a "real musician" (tm), and my experience using synths is recording a few things with soft synths, so not everything of such matters is obvious at a first glance, for me.
Does this have to be bad? Not if one is content with what e.g. the Poly 800 can do, I guess.
But yeah, I guess it boils down to a question of price and building effort, and how flexible I want my synth to be.
Filter ICs, china sellers
Btw, are those 2044 chips sold by chinese sellers any good?
I've seen several different versions of them, I mean the number below the part no., is this the revision?
E.g. 8333, 8354, 8730, 8740, - what are the differences?
I figure those things should save a lot of perfboard space
I'll surely build one or the other discrete VCF, but for 8 or more voices? I'm not too inclined currently |
|
Back to top
|
|
|
Cynosure
Site Admin
Joined: Dec 11, 2010 Posts: 966 Location: Toronto, Ontario - Canada
Audio files: 82
|
Posted: Sun Feb 19, 2012 12:14 pm Post subject:
|
|
|
I think polyphonic synths do have a filter for each voice. They are just controlled with one common pot that sends a cv level control to each filter, but each voice has its own filter and envelope.
Old organs only had one envelope, so even though they were polyphonic, you needed to release all notes before hitting the next note that you wanted to produce the percussion sound. _________________ JacobWatters.com |
|
Back to top
|
|
|
ian-s
Joined: Apr 01, 2004 Posts: 2669 Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 626
|
Posted: Sun Feb 19, 2012 12:55 pm Post subject:
|
|
|
Cynosure wrote: | I think polyphonic synths do have a filter for each voice. |
That is true for your P5/OB8/JP8 etc but there were a few hybrid polysynths like the the poly800, arp omni that had a single vcf. The polymoog had a on chip vcf per key but also a single moog filter on the main output. It's better than nothing (but not much). |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Sun Feb 19, 2012 1:43 pm Post subject:
|
|
|
unshaven wrote: | ...As I understand, the principal topology of all monophonic analog synths I've had a glimpse at, dictates that the VCA comes as last part in the chain, fed by the VCF output, to create the envelope.
I sometimes wonder why - is it to reduce noise, as opposed to, e.g. VCF behind VCA?... |
That may be one reason. Another reason is because some filters are designed to oscillate when the resonance is maxed. It would probably be desirable to control the volume of that tone with an EG and VCA. _________________ varice |
|
Back to top
|
|
|
reve
Joined: Feb 23, 2008 Posts: 149 Location: USA
|
Posted: Tue Feb 28, 2012 2:20 am Post subject:
Re: Analog polyphonic - topology question |
|
|
Hey Steve.
Previous posters were correct; most analog polysynths followed the o-f-a topology... The beauty of modular is that you can do this any way you so choose. the poly 800 you mentioned actually controlled the amplitude of the individual voices in software. So while it was technically still o-f-a, it was effectively more like o-a-f-a, with individual amps on each voice and a final filter and amp at the end.. The Roland em 101 worked pretty much the same way.
Hth! _________________ - r. mosquito
SynthDIY Synthpop -> http://www.reverbnation.com/photovoltaik |
|
Back to top
|
|
|
|