This page is intended to provide a brief outline on the basics of polyphonic synthesizers. My intention is not to dig deeply into the technical design aspects, but to try an give the musician or person new to synthesizers, the basics of how they generally work. Other pages in this series may help in getting a backgrounder:
Challenges of Polyphony in Analogue Synthesizers
Imagine implementing a analogue polyphonic synthesizer using the keyboard CV and VCO design example below, (for a backgrounder see VCOs, DCOs and Digital Oscillators). The drawing below shows the analogue synthesizer process for a monophonic synthesizer. The keyboard generates a CV (Control Voltage) for pitch and another called a Gate, which provides a signal for the length of time a key is pressed. Unlike the variable control voltage, a gate signal has a fixed voltage. This concept is known as CV/Gate. All of the sections that make up a voice require CV or Gate signals to make adjustments and modifications to the waveforms from the VCO. Note here that keyboards may or may not be integrated into the overall synthesizer, CV/Gate will still be the underlying concept (example: using patch cables in a modular).
What if you want to play chords? You will need multiple voices.
But how?
Imagine making a 4 voice polyphonic synthesizer as depicted above. You will need:
- 4 separate CV signals, one for each voice
- 4 separate gate signals, one for each voice
- A method of assigning each voice to the CV/Gate signals
- A way of sending the control signals for knobs, switches and sliders from the control panel.
This is a very complicated problem indeed. In monophonic synthesizers, you press a key, the note plays until you release the key. If you press 2, 3 or 4 keys simultaneously, only one of the voices will play a sound. For our 4 voice poly, you could press 1, 2, 3 and 4 keys (even simultaneously) and you will get the appropriate note for each key. But what if you press a 5th key? You will need some way to reassign one of the keys. Do you stop playing one of the current 4 keys and now play a new note for the 5th key? You could also ignore the 5th key and just play the 4 notes being held down. If one is released, then you could end that note and then play the note representing the 5th key.
Take for example the Korg PS-3100 or 3300s which used a complete analogue voice for each of its 64 keys (48 on the PS-3300). This was one of the first solutions, but needed a large amount of electronic components, miles of wiring and would cost thousands even back in 1977. Issues would be with tuning and calibration, heat generated from all the electronics, and size.
Other early polyphonic synthesizer manufacturers like Yamaha, turned to complicated multiplexing, which uses dozens of analogue integrated circuits and miles of wiring to provide CV, Gates and control the multiple signal paths required. They also built customized integrated circuits for voicing components. Yamaha’s efforts were rewarded with a 82 kilo (180 lb.) behemoth of an 8 voice polyphonic analogue synthesizer called the CS-80.
In trying to build it all analogue, Yamaha ended up increasing the costs tremendously and produced a machine that although sounded fantastic, still had challenges for size, tuning, calibration and maintenance. The CS-80 was just under $7,000 USD in 1977 (about $36,000 in 2024). You would also be hard pressed to lug this thing from gig to gig. It was just too large and moving it constantly meant it would break. The CS-80 had hardware based (switches) presets and a huge control panel for adjusting all of the parameters but still had no way of saving those adjustments.
Microprocessors and the advent of Polyphony.
About the same time, and realizing the issues associated with complicated multiplexing circuits, designers of the Prophet 5, turned to commercially available microprocessors, which were now becoming inexpensive and abundant. A microprocessor, when used with Analog to Digital (ADC) and Digital to Analog (DAC) devices, can encode & decode all of the CV signals, thereby supplying a stable CV source wherever needed throughout the synthesizer. It also enabled the use of programmable presets with instant recall for any new, or previously saved presets.
There are huge benefits of a microprocessor:
- MIDI can be incorporated into the design, thereby replacing older CV keyboards.
- Presets and patches can be saved and recalled.
- Polyphony is much easier to implement and more stable.
- Complex envelopes and modulations can be implemented.
- Size and complexity reduction
- Far less expensive to manufacture.
Complicated internal wiring is replaced by the processor, which handles all CV & Gate signals. In the design above, voices are still completely analog, as they are using VCOs, so there can still be some drift within each voice and between voices. In fact you could still use discrete analogue designs for all of the sections in a voice. Keeping them tuned and calibrated with 4 or 5 voices was still somewhat complex and would still be less stable with temperature swings, This would really be noticeable as more voices are added. It also became more expensive to add voices.
The solution was for most manufacturers to replace discrete circuits with specialized integrated circuits (ICs) for the VCO, VCF, VCA and Envelope Generators. Companies like Curtis, Roland, Yamaha, Korg and Solid State Music provided many different designs. These specialized ICs had a massive impact on music synthesizers because they are very stable and consistent in performance. Most polyphonic analogue synthesizers made since the late 1970s up to today’s models, utilize a combination of VCO or DCO with some specialized ICs for voicing.
Voices grew from 5 to 8, 12 and even 16 (or more). As voices increased, so did some of the issues with using VCOs, namely, drift in tone and tuning between voices. Many consumers wanted a tighter sound between voices, especially when playing strings and long pads. Consumers also wanted new sounds and modulations types. Using a DCO can solve the tuning problem and implementing a digital oscillator can provide other types of synthesis. Many new models of polyphonic synthesizers have also replaced the internal analog components for CV & keyboard control with other forms of digital communication, or eliminating the concept if CV/Gate altogether. This enabled designers to add functionality such as MPE and other forms of expressive behaviour to the voices.
But wait a minute! If all this control stuff is implemented with a computer and software, is it really analogue? Yes, and the reason is that you do not hear any of the CV signals, you only hear their effect. CV signal are just voltages, that increase or decrease over some time element. CV signals are used to control an audio signal (which is what you hear). CV can be used to adjust the pitch of an oscillator, open or close a filter or provide an envelope (voltage profile) for an amplifier. In fact, computer or digitally generated CV can model discrete designs or integrated circuit designs exactly, even adding instability and skewing to the CV to mimic fully analogue designs. Remember, VSTs and Virtual synthesizers are fully implemented in software and converted to analogue for the output amplifier. See Digital Vs Analogue.
I love polyphonic synthesizers and I find them far more versatile than mono synths. Contrary to many arguments about sound differences, the simple fact is that many modern analogue monophonic designs are built basically as a single voice poly. They incorporate microprocessors in the exact same way as polyphonic synths do (depicted above). Many monophonic synths today even use integrated voicing circuits, DCOs or even digital oscillators. This is the primary reason I chose polys, as I can get larger phatter sounds by simply adding voices in unison mode. That’s not to say I don’t use mono synthesizers. They certainly do have a place, but I tend to use and keep only those mono synths that are pure analogue (CV/Gate driven & no MIDI), without any microprocessor or digital elements in the design.
Some of the polyphonic synthesizers I use in studio are below. I have linked to reviews from others, and expressed my opinions about them.
Roland Integra 7 (128 Voices)
- DSP ROM Sampler
- Digitally created filters, envelopes, modulation and effects
- Analogue VCA
- MIDI DIN or USB
- 16 Part timbral
- 19″ Rack Mountable
If I had to choose one synth to compose soundtracks for documentaries or films, it would be this sound machine from Roland. Everything from keyboards to drums to orchestral instruments are available. Hundreds of instruments from around the world are stored in high quality digital samples stored in ROM. I can easily play 7 or even 8 separate MIDI tracks without any delay, Phenomenal machine for composition, and used on almost every piece of music I compose. As of May 2024, these are still made but you can also get these used for under $1500 USD. Sound on Sound review.
Yamaha DX-7 FDII (16 Voice)
- DSP Oscillators (FM Synthesis)
- Digitally controlled envelopes
- Analogue VCFs (Yamaha specialized ICs)
- Analogue VCAs (Yamaha specialized ICs)
- MIDI DIN only
- Bi-timbral
Yamaha TX-802 (16 Voice)
- 19″ Rack mounted version of the DX7-II
- DSP Oscillators (FM Synthesis)
- Digitally controlled envelopes
- Analogue VCFs (Yamaha specialized ICs)
- Analogue VCAs (Yamaha specialized ICs)
- MIDI DIN only
- 8 part-timbral
The sound of the 80s. The DX7 out sold every other synthesizer back in the day, and I do not even know if many modern synthesizers have come close to the sales numbers from this synth. It is very versatile, well built and portable to be used on stage. My DX7-FDII finally gave up due to a blown-up power supply. But its brother TX-802 rack version is still going strong. These have not been made for over 30 years, and replacement parts are usually scrounged from other units. These sell for less than $400 on the used market. The biggest downfall is that they are a pain in the butt for programming new sounds. Menu diving along with the user manual is a must. Having said that, they are the standard model for FM synthesis. Sound on Sound review.
Prophet Rev 2
- 2 DCOs per voice (16 voices)
- Digitally implemented CV from MIDI (DIN or USB)
- Analogue wave shapers for sawtooth, Pulse Width Modulation & triangle
- CEM3320 based analogue VCFs
- Analogue VCAs
- Analogue envelopes
- Polyphonic step sequencer.
- Bi-timbral
This is without a doubt, my most used Analogue synthesizer. It has a tremendous breadth in sound palette. It can perform incredible bass parts, blistering leads and lush pads. Tons of modulation and effects enable a sound designer to produce some incredible tone. New the Rev 2 (16 voice) is over $2000, but used can be had for about $1200 USD.
Prophet 12. (12 Voices)
- 4 DSP oscillators per voice
- MIDI DIN and USB input only
- Analogue wave shapers for sine, sawtooth, square & triangle
- CEM3320 based analogue VCFs
- Analogue VCAs
- Digitally controlled envelopes
- Bi-timbral
- 19″ Rack mountable
The prophet 12 was apparently the late Dave Smith’s favourite synthesizer. Although highly underrated, this is the best hybrid poly on the market. It has incredible range of analogue sounding tone and can also perform linear FM brilliantly, and has more modulation and effects than any other sequential (Dave Smith) synthesizer. Sadly, the prophet 12 is not made anymore but a good used one can be had for about $2000 USD
Black Corporation Deckard’s Dream (8 Voices MK1, ver 1.0)
- Digitally implemented CV from MIDI (DIN or USB)
- CEM3340 analogue based VCOs
- CEM3320 analogue based VCFs
- V2164 analogue based VCAs (SSM2164 replacement. MK2 and V2 used discrete VCAs with OTAs)
- Analog envelopes
- Bi-timbral, MPE enabled
- 19″ Rack Mountable
This is in my opinion the closest thing to a Yamaha CS-80 you could possibly obtain. A very unique synth with a very unique sound. Other polys may be able to recreate one or two of the CS-80 presets, but the Deckard’s Dream sounds like one at its core, before presets are even used. This is truly a remarkable synthesizer that I use for many leads in music. New ones cost about $5000 USD, and you could get a used one for around $3500 to $4000. Be careful with used though. This synth was also offered as a DIY kit with version 1 requiring several items of rework to address a couple of issues. Many DIY builders did not implement these. If in doubt stick with a factory built version. Sound on Sound Review
Black Corporation Ise-Nin (8 Voice)
- Digitally implemented CV from MIDI (DIN or USB)
- Analogue VCOs (SSI2131)
- AS3109 Analogue VCFs (IR3109 replacements)
- SSI2164 Analogue VCAs (direct replacement for SSM2164 & AS2164)
- Digitally controlled envelopes
- Bi-timbral, MPE enabled
- 19″ rack Mountable
like the Deckard’s Dream, this is in my opinion the closest thing to a Jupiter 8 you could possibly obtain (until maybe Behringer comes up with one?). A new one will set you back about $5000 USD. Again, be careful with used though. This synth was also offered as a DIY kit with version 1 requiring several items of rework to address a few issues. Many DIY builders may or may no have implement these. If in doubt stick with a factory built version. You mat have to wait a bit to obtain a used model as they are very recent (as of 2024). Sound on Sound Review
AudioThingies MicroMonsta 1
- Digitally implemented (virtual synthesis)
- MPE enabled
- MIDI USB or DIN.
- Bi-timbral
AudioThingies MicroMonsta 2
- Digitally implemented (virtual synthesis)
- MPE enabled
- MIDI USB or 3.5mm mini jack.
- Bi-timbral
Oberheim OB-X8 (8 Voices)
- Discrete analogue VCOs
- CEM3320 based analogue VCFs
- Discrete SEM based analogue VCF
- Analogue VCAs
- Digitally controlled Envelopes
- Digitally controlled CV
- Bi-timbral
In the late 70s, early 80s, Oberheim made some of the greatest sounding synthesizers ever made. But like many synthesizer manufacturers (Moog, Sequential, EMS, EDP), they went bankrupt in the mid to late 1980s. Roland, Korg and Yamaha dominated the market with digital synthesizers. Oberheims however, had a very distinct sound and are highly sought after in the used market today. In fact, it is not unheard of that a used OB-Xa will fetch upwards of $12,000 USD (as of May 2024). With the rebirth of Moog, Sequential and Oberheim in the 2000s, many of the original polys were recreated to offer many new features and possibilities. The Oberheim OB-X8 however, was not released as a new synthesizer design and does not come with many modern or additional modulation features.
The sole intended purpose of this synth was to provide the user with the closest sound pallet to the original set of Oberheim synthesizers: SEM-1, OB-X, OB-SX, OB-Xa, & OB-8, without worrying about the costs involved with purchasing and maintaining a 40 year old synthesizer. This is not a clone of an OB-Xa, for all intents and purposes it is the latest revision to the original.
I love this synth and the circuit designs are almost identical (see chart below). It has been modernized only through the use of surface mount devices as opposed to the older through hole technology, but that has little impact from the originals other than voltage drift from age, temperature, etc. But the OB-X8 uses discrete VCOs and VCFs just like the originals!! Compensation can also be applied through adding elements of variance to the control voltage for the IC filter version. I have included the Behringer UB-Xa as it too was built using the same basic circuitry as the original OB-Xa and OB-8 synthesizers. The only differences are that the UB-Xa does not cover the OB-X or SEM-1 filtering. The UB-Xa also has tons of additional modulation features witch can radically change the sound of the original OB-8/Xa.
There is a huge difference in price though. The Oberheim OB-X8 uses higher quality electro-mechanical components and the identical Curtis integrated circuits that the original used. They were re-released and made by OnChip systems. The OB-X8 is also built and assembled (including the printed circuit boards), in the USA. The Behringer version uses lower quality electro-mechanical components and clones of the original ICs and is made in China (see chart below). The price difference though is the real difference, with the Behringer UB-Xa coming in at $3000 USD less than the Oberheim. So, if you are a purest and value local manufacturing, go Oberheim. If you would like to experience the OB-Xa sound and perhaps try the multitude of new modulations, then Behringer UB-Xa is for you. Sound on Sound Review. If you are short on cash, just get a VST version.
A final note on the Oberheim ( applies to most of the new desktop synthesizers). The OB-X8 uses digitally generated envelope generators instead of the CEM3310 integrated circuits like most of the previous Oberheim models. In my opinion, envelope circuits can be easily modelled and most likely Oberheim chose this method as it can easily replicate the signal profiles for both the SEM and OB series of synths and provide MPE and additional modulations through MIDI. There has been so much chatter and nonsense on social media about this. Some actually equate a hardware synthesizer that implements an envelope generator digitally, with a VST. Nothing could be further from the truth. Remember, 95% of all polyphonic synthesizers produced today (especially the keyboardless, desktop versions), do not utilize an analogue implementation of CV or Gate/Trigger. Do you see any CV/Gate inputs? They rely totally on MIDI control messages and some form of processor controlled CV for the oscillators, filters, modifiers and amplifiers.
In closing I hope you, the reader has a little more knowledge to help you weed you way through much of the BS you will find online.