Before synths became mostly digital, there was a brief period where we polyphony was the main attraction. However, in general terms, subtractive ploys didn't offer many vectors of expression.
That isn't to say there weren't expressive analog polys. I'd point to the CS-80 as an expressive instrument, with polyphonic aftertouch, ribbon controller and lovely ring modulator. It is just that subtractive offers a fairly limited set of parameters suitable for expression. Once you've routed aftertouch or velocity to open up the filter or amplifier a bit to make things louder and/or brighter, you've pretty much exhausted much of what makes sense musically.
While it is true the DX7 justifiably earned its reputation as difficult to program, the one thing people seem to neglect to mention is Yamaha's implementation of FM synthesis was extremely musically expressive to play. The patches came alive with velocity, not simply brighter or louder with velocity. The timbre could change is ways that were musically sensible, but responded with shifts of timbre way beyond what was possible with subtractive. Aftertouch was standard. The modulation wheel could do many more interesting things beyond introducing vibrato and if that wasn't enough, there were foot and breath controller inputs. All of these could produce timberal shifts that were previously unheard of.
Every major player needed their own take on a unique digital polyphonic synthesizer and the K5 was Kawai's first real additive synth. The styling of the K5 is very late-80s Roland, bearing striking similarities to the S50 sampler.
This was the grandpa of the K5000 series which came much later. The K5 retained all the qualities that made the DX7 difficult to program, but the real time expression didn't extend beyond what was typically available on a subtractive poly.The result was a very thin, flat, glassy synth. Worse, it was fairly noisy.
I picked up a K5m in the early 90s for $150, and, despite the limitations mentioned above, I used it a ton. I had a small program that would analyze Mirage samples and send the resyntheized data to the K5 for an extremely poor man's Axcel. It had a very capable multi mode. While programming individual harmonics over time (well, there were only four envelopes available to control the harmonics) from the front panel isn't very much fun, or musically intuitive, navigating and altering the multimode was straightforward and easy to use. So, I used it in layers a lot, adding a digital sheen to other, fatter synths, like the K3m.
The K3 predated the K5 by a very short amount, and it was pretty much a standard subtractive synth. The K5 did have a few unique features for the time, including a limited waveform creation mode where you could build up a waveform, harmonic by harmonic that pointed gently at where the K5 was ultimately going. There were a total of 32 different available waveforms available and these were fed into a lovely analog SSM filter. The envelopes were also analog. In mono mode, the K3 was completely capable of gobbling your entire dynamic range. The K3m was another module that I picked up used for $150 and used for a number of years.
Both the K5m and K3m ultimately gave way to a pair of Kurzweil K2500s and Roland JV1080 that I used from the mid 90s on though a transition to software instruments and a shift to the modular as the dominant hardware synth.
Update. I managed to dig up a very early specimin that is exclusively K5 and K3. The K3m is providing the bass, everything else is K5m. ORIGINALLY RECORDED TWENTY YEARS AGO.
K3m K5m example by stretta samples
Monday, February 21, 2011
Tuesday, February 8, 2011
Tarrega Andalusian Dance
I've been avoiding taking on a musical project - even a small one - because I'm consciously keeping my schedule and psyche clear for a couple important things that require my full attention. So, this piece is sort of an accident. I started it as a study of a compositional style I want to try.
Tarrega Andalusian Dance by stretta
Using a modular feels like the most natural, organic, hands-on approach to synthesis I've ever used. My first programmable synth was a Casio CZ-101. I was brought up on menus and inc/dec buttons. With the modular, I don't feel like there are any barriers between me and what I'm trying to achieve with sound. I'm not an analog purist and I have many 'digital' modules, but those modules still interface with knobs and patch cords.
There are obvious drawbacks to using a modular, like the lack of patch storage, but tearing down a patch doesn't bother me in the slightest. I create the patch for that moment, record it, and move on. Usually when I'm recording, I'm planning the next patch in my head. It doesn't take that long to patch up the voice I want to use. A patch doesn't have to be complicated, it just has to be appropriate.
There is a lot of FM synthesis in this piece. Some from the Hertz Donut, some from an AFG driving a Cwejman VCO6. Not much filtering. Mostly plan B low pass gate on some parts. Some bits from the Piston Honda. Not using stacked detuned oscillators, either - I didn't think the orchestration called for it. One reverb (eos).

Tarrega Andalusian Dance by Matthew Davidson is licensed under a Creative Commons Attribution 3.0 Unported License.
Tarrega Andalusian Dance by stretta
Using a modular feels like the most natural, organic, hands-on approach to synthesis I've ever used. My first programmable synth was a Casio CZ-101. I was brought up on menus and inc/dec buttons. With the modular, I don't feel like there are any barriers between me and what I'm trying to achieve with sound. I'm not an analog purist and I have many 'digital' modules, but those modules still interface with knobs and patch cords.
There are obvious drawbacks to using a modular, like the lack of patch storage, but tearing down a patch doesn't bother me in the slightest. I create the patch for that moment, record it, and move on. Usually when I'm recording, I'm planning the next patch in my head. It doesn't take that long to patch up the voice I want to use. A patch doesn't have to be complicated, it just has to be appropriate.
There is a lot of FM synthesis in this piece. Some from the Hertz Donut, some from an AFG driving a Cwejman VCO6. Not much filtering. Mostly plan B low pass gate on some parts. Some bits from the Piston Honda. Not using stacked detuned oscillators, either - I didn't think the orchestration called for it. One reverb (eos).

Tarrega Andalusian Dance by Matthew Davidson is licensed under a Creative Commons Attribution 3.0 Unported License.
Tuesday, February 1, 2011
serialosc
"serialosc is designed to be invisible"
Beautiful.
I never met anyone as obsessed with burning away design impurities, as Brian Crabtree. People gape at the physical design of monome's products. Their simple beauty is almost universally acknowledged, even by their detractors. Fewer people appreciate how the PCBs themselves are beautiful. For example, the PCBs are modular, which mirrors the modularity of the devices themselves. A monome 256 has four button board PCBs, two of these same boards may be used to make a 128, or one may be used to make a 64. There is one logic board per unit, and the same board is used in all units. The same beauty extends to the deceptive simplicity of the adaptable and flexible protocol itself.
The near simultaneous arrival of variable brightness, support for the arc, the protocol, and serialosc (win/mac/linux - designed by brian and programmed by will light) are all obviously connected and I don't think people fully understand the consequences. This is the groundwork for a new class of device; laying down the literal language of musical interoperability.
Beautiful.
I never met anyone as obsessed with burning away design impurities, as Brian Crabtree. People gape at the physical design of monome's products. Their simple beauty is almost universally acknowledged, even by their detractors. Fewer people appreciate how the PCBs themselves are beautiful. For example, the PCBs are modular, which mirrors the modularity of the devices themselves. A monome 256 has four button board PCBs, two of these same boards may be used to make a 128, or one may be used to make a 64. There is one logic board per unit, and the same board is used in all units. The same beauty extends to the deceptive simplicity of the adaptable and flexible protocol itself.
The near simultaneous arrival of variable brightness, support for the arc, the protocol, and serialosc (win/mac/linux - designed by brian and programmed by will light) are all obviously connected and I don't think people fully understand the consequences. This is the groundwork for a new class of device; laying down the literal language of musical interoperability.
Subscribe to:
Posts (Atom)


