The No.1 Website for Pro Audio
How does a fully anolog poly work?
Old 7th October 2011
  #1
Gear Addict
 

How does a fully anolog poly work?

Lets just let me begin with a disclaimer of some sort. Up until a few weeks ago I felt that analog or digital (computer) its just the same crap, but after listening to older fully analog synths I would now say that fully analog synths actually does not sound the same, the sound more fat to me, more "bite" more badassness, and dirt. Not as polished and perfect in a kind of way.

So I know that there are not that many poly analogs around, so it must be hard or expensive to get it working.

I have heard many synths and for me personally I want something that I can buy today, something thats made today, the best I have heard up until now are Oberheim SEM and Mr Macbeths creations.

These are really expensive if you want to buy like 4 of them to have a 4 voice synth, but in a few years maybe I could afford it.

I just want to know how this is all supposed to work, I keep reading synth x has 2 osc and 10 voices how does that work, doesn't it need 10 oscs for 10 voices (or 20).

Is it possible today to build a real polyphonic analog synth, polyphonic as in vst-polyphonic with the gear thats out there?

How much would this cost? It has to be new.
Old 7th October 2011
  #2
Lives for gear
 
Llitsor's Avatar
 

It's hardly convenient to explain the inner workings of an analogue synthesizer in a single forum post. Entire books have been written about that.. You'll have to hit up Google and do some research yourself.

But, voices vs oscillators.. If you look at some mono synths.. They don't always have 1 oscillator. A TB303 has 1 oscillator. But something like a Waldorf Pulse, which is also an analogue mono synth has 3 oscillators.
It's not that different for a poly synth, each voice/note can use 1 oscillator, or it can have more than that. It depends how the synth was designed.
Having more oscillators gives more scope for varying, and thickening up the sound.

So if you wanted to make an 8 note polyphonic analogue synthesiser, using, say, 2 oscillators per voice/note. You need 16 oscillators. And 8 filters, and 8 or more VCA's and and, lots more besides. Hence, complicated, and expensive.
Old 7th October 2011
  #3
Lives for gear
 

Yes! Analog polys are actaully several mono synths in one. An 8 voice synth needs 8 filters, 8 vca's, 8 envelopes, 16 vco's etc. A good example is the Oberheim 4 voice. It consists of 4 monophonic synthesizers (each has its own box/case) that are connected to play chords. The cheapest way to imitate an analog poly is to buy a mono synth and to record every note of the chord separately. It may be a pain in the ass, but with todays DAWs and sequencers we have boundless possibilities, everything is possible!
Old 7th October 2011
  #4
Lives for gear
 
BTByrd's Avatar
DSI Prophet 8 and Studio Electronics CODE are both polyphonic analogs available today. The P08 operates by using integrated circuits, essentially "synth voices on a chip." Using ICs enables the P08 to be small and relatively inexpensive. The CODE's design, by contrast, is largely discrete and has a full circuit board for each voice. If you want to expand its polyphony, buy another voice board.

I honestly couldn't imagine trying to use something like the four voice (though it sounds amazing awesome). Having separate interfaces for each voice seems like it would get pretty exhausting pretty quickly.
Old 7th October 2011
  #5
Gear Addict
 

Thanks guys this is exacty what I meant. Its a lot of monos. Going to be expensive.

What do people usually do when they want a real analog poly these days, and how much would this actually cost/voice? I know about overdubbing and such but what a hassle, you can't even hear what you play until the overdub is done since you can only play one single note at a time.
Old 7th October 2011
  #6
Gear Addict
 

Quote:
Originally Posted by BTByrd View Post

I honestly couldn't imagine trying to use something like the four voice (though it sounds amazing awesome). Having separate interfaces for each voice seems like it would get pretty exhausting pretty quickly.
I think this a bit cooler because now you can alter the individual voices in some way, it will sound "unique" every time you use it. But yeah no presets there, not for any of the 4 voices...

But still there has to be something that says to one of the voices "its your turn now" how does this work? If it works with the 4-voice it has to works with separate monos too right?
Old 7th October 2011
  #7
Lives for gear
 

Quote:
Originally Posted by ._. View Post
I know about overdubbing and such but what a hassle, you can't even hear what you play until the overdub is done since you can only play one single note at a time.
That's why you have to have something like a VSTi or VA like a Nordlead or Virus etc. You create a similar sound in the digital world, play your chords and then you record one voice after the other with your analog mono synth. As I said, it's a pain in the ass, but it's cheap and sounds unique and fat without the crappy, digital, aliasing, plastic sound!
Old 7th October 2011
  #8
Lives for gear
 
BTByrd's Avatar
Quote:
Originally Posted by ._. View Post

But still there has to be something that says to one of the voices "its your turn now" how does this work? If it works with the 4-voice it has to works with separate monos too right?
One way to do this would be to use a voltage controlled switch so that each time a gate signal is sent from the keyboard, it routes that gate and pitch CV to a new voice. But you really don't need to worry about it, because any polysynth you buy will be able to do this automatically.

There are LEDs on the Omega/Code to indicate the active voices, which can be kind of hypnotic to watch.

Old 7th October 2011
  #9
Lives for gear
 
Llitsor's Avatar
 

Quote:
Originally Posted by ._. View Post
What do people usually do when they want a real analog poly these days
Just take a look around! They get Jupiters, Andromedas, DSI prophets amongst several others.

Quote:
Originally Posted by ._. View Post
and how much would this actually cost/voice?
As I mentioned earlier. Google. Look around the forum. Search for the synths listed so far and you'll get an idea of what they cost.

Quote:
Originally Posted by ._. View Post
I know about overdubbing and such but what a hassle, you can't even hear what you play until the overdub is done since you can only play one single note at a time.
There are a few ways to do it, and some modern DAW's make it a lot easier too. For example, you could use some other, maybe soft poly synth to compose with. Then explode the chords into their own note lanes/tracks. Send those midi tracks to your analogue mono synth however many times is needed, record each track, and there you have it.
I tend to just play it in by hand (and edit later). So just do what works for you.
Old 8th October 2011
  #10
Lives for gear
 
pointsource's Avatar
 

When i want to make pads with my slim phatty, i do that as well.

After you compose the track with a soft synth, take the midi clips and record your analog synth, replacing track by track.

If i only use pads, let's say, from C3 to C4, i record 20 seconds of the raw waveform from the SP (from C1 to C4), and then play tem using Ableton's sampler.

But if you're using just a few chords, you can multitrack your synth. Eg C minor:

1st track will be C
2nd track will be E#
3rd track will be G

From here the possibilities are endless, just be creative. You can use this same chord to make other effects or whatever, applying different efffects, etc...

To make the screen less cluttered, you can set up another track and record the audio coming from those 3 tracks (I use the resampling option), then you'll have 1 track with the chord.
Old 8th October 2011
  #11
Gear Guru
 
Yoozer's Avatar
Quote:
Originally Posted by ._. View Post
But still there has to be something that says to one of the voices "its your turn now" how does this work?


The following is a simplified description based on the little stuff I know about electronics and digital engineering. It is probably not entirely correct.

If you have 8 voices, you have 8 memory registers that can be set to on/off. If you have a monophonic synth, there's no problem; any new incoming signal overwrites the previous one so you only have to keep track of pitch and whether the key's pressed.

You press a key. A signal is sent to the processor, which then looks up in the memory which register is set to off; simply loop through them to find the first free one. It then marks it as "occupied" and stores the 7-bits code of which key you pressed. Then, it generates a corresponding control voltage depending on the pitch of the key. It also generates a gate signal, which opens up the amplifier and starts the envelope trajectory.

This goes at incredibly high speeds - it's not like you're going to be able to press two keys at the same time, nevermind the fact that they're scanned in series, too - so one will always be later than the other.

You press the next key. The processor looks up "ok, slot 1 is occupied, but 2 is still free!". It switches it to occupied and repeats the whole thing.

When you release a key, the processor receives a signal "key nr. 66 out of 128 - (the aforementioned 7-bits code) has been released. You loop through the slots, look up which one that was, mark it as free and drop the gate signal.

The lights on the CODE demonstrate exactly that; all you need is a little loop that scans those registers 50 times a second or so and copies the results to a bunch of LEDs.

You can implement this in a very basic processor with minimal memory. and a low speed already; this is why synths from the 80s could already do this.

There are several "modes" for this. For instance, you could make it so that it looks for the first unoccupied register. That way, if you would be playing 3-note chords with a very short envelope release on a 6-voice polyphonic synth, it'd never use more than the first three registers. This is not necessarily desirable, so you can also keep in mind what the last voice played was and always find the next unoccupied spot. If you have a set of lights like on the CODE or Prophet '08, you can test this; hit a note on a sound with a short release time repeatedly. If it scrolls through the lights, it always finds the next unoccupied spot, if only light nr. 1 lights up every time, it finds the first unoccupied spot.
Old 8th October 2011
  #12
Gear Addict
 

Thanks for the explanation. I guess this function can be had when hooking up 4 different synths too? Like 4 Sems. Are there separate stand alone boxes that handle the "voice-administration"?
Old 8th October 2011
  #13
dont buy 4 mono synths - look at the second hand vintage synth market.

Loads of interesting poly analogues...
Old 8th October 2011
  #14
Lives for gear
 

Quote:
But still there has to be something that says to one of the voices "its your turn now" how does this work?
As Yoozer stated, the CPU decides. On most poly-synths the method of voice allocation is somewhat programmable.

For example, here are the modes for a Jupiter 6:

Quote:
Solo, Unison, Solo Unison, Poly 1, and Poly 2 modes. Solo allows you to only play one voice at a time (ie. one 2-oscillator voice). Solo Unison mode limits the voice allocation to one note at a time, but stacks all six voices on that note (as opposed to solo, which uses just a single voice) so that the Jupiter becomes a 12 oscillator, monophonic synthesizer! Unison mode plays all of the voices, all of the time, dividing them as evenly as possible between all sounding notes. That means that if you played a single note, you would hear all 6 voices stacked together. Playing two notes would get three voices apiece, playing three notes would get two, and more than that would get you one or two notes with two voices, and then the rest with a single voice. This offers the flexibility of having a polyphonic keyboard when you need it and the ability to get a fatter sound out of one or two notes. The amount of detuning in Unison mode is programmable for each side of a split. Poly 1 is the normal 6-voice polyphonic mode, and Poly 2 uses normal polyphonic allocation as long as you play legato. Once you lift your hand off the keyboard, any new notes will cut off the releases of all currently sounding notes. This is recommended for use with portamento.
On something more complex like an Andromeda, CODE 8, or Xpander, the voice allocation modes are completely user-programmable, with such things as round-robin, high/low note priority, splits/layers, MIDI channel, etc.
Old 8th October 2011
  #15
Gear Addict
 

Are there any cheaper poly synths that sound like that Omega CODE 8? I really like that entire pad demo.
Old 9th October 2011
  #16
Gear Guru
 
Yoozer's Avatar
Omega 4-voice?

It's awesome, right? heh
Old 9th October 2011
  #17
Lives for gear
 
crufty's Avatar
vermona perfourmer ?

Analog osc + fltr has magic. No question.
Old 9th October 2011
  #18
Lives for gear
 
pinkerton's Avatar
 

is there any sort of thing anybody makes that can round robin notes over different cv/gate or midi outputs? would love something like this ... sory for the hijack but it seemed relevant.

actually im pretty sure you could do it with a sequencer that had gate outputs at each stage, so nevermind.
Old 9th October 2011
  #19
Gear Addict
 

Quote:
Originally Posted by crufty View Post
vermona perfourmer ?

Analog osc + fltr has magic. No question.
From what I understand its a 4osc mono, unless something can administer the polyphony.


Imagin how cool it would be to have a 10 voice 20 osc Macbeth, Cwejman or SEM. Or Even a 2 voice 4 osc one. If I just had some disposable millions I would have 3 synths, the 3 I just mentioned. heh
Old 9th October 2011
  #20
Lives for gear
 

Quote:
Originally Posted by pinkerton View Post
is there any sort of thing anybody makes that can round robin notes over different cv/gate or midi outputs? would love something like this ... sory for the hijack but it seemed relevant.

actually im pretty sure you could do it with a sequencer that had gate outputs at each stage, so nevermind.
The old Roland 4 channel MPU-101 and the Encore Expressionist could do it but I'd look at the PolyDAC.
Old 9th October 2011
  #21
Lives for gear
 
crufty's Avatar
Quote:
Originally Posted by ._. View Post
From what I understand its a 4osc mono, unless something can administer the polyphony.
Sort of


I think logic or reaper could do the round robin midi note thing

Probably pretty easy to coax an Adruino to do it too

Well maybe not easy but easier...
Old 9th October 2011
  #22
Someone elase sould probably explain this better, but...
There is also some sort of volume output compensation in the circuitry for a poly synth. The VCA on a mono is designed/calibrated for a single voice. In order for a poly to have acceptible levels when playing more than one note, the VCA would need to output each voice at lower levels in order to prevent clipping/distortion. Imagine playing 4 minimoogs at the same time, with the output set appropriately for one voice, across each minimoog. A single note played on a mono synth usually sounds louder, or bigger than a single note played on a poly. On a old Oberheim X Voice synth, there is an output module that allows you to adjust the output for each SEM because those synths are basically 2,4,8 mono synths chained together.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump