I particularly enjoyed the Cuckoo overview of FM principles that I came across in the Audulus forum thread on AM/FM/PM modulation.

I’ve since been wanting to take a closer look at the difference between linear and exponential frequency modulation and came across this video (also featuring an old ARP Oddysey Mk I) by Synthesizer Keith.

There’s a follow-up Reddit thread also clarifying that it’s linear modulation at play in the Yamaha DX style FM – the phase type of frequency modulation that Cuckoo refers to in his video – as well as a similar thread at KVR.

Keith refers to the Yamaha style linear FM as being “more musical” in his video, and there’s a Reddit thread, leading to a Muff Wiggler thread, on that question as well.

I found it interesting to look through the Wikipedia entry on Frequency modulation synthesis. I always associate FM with the 80s (the DX7!) but as the article points out Yamaha was already developing (and patenting) it (after famously licensing it from John Chowning¹ ) in the mid 70s. It was only in 1980 that they’d developed the technology to a degree stable enough to release it commercially in the form of the curious Yamaha GS-1.

All of which was preceeded by Buchla with his own developments and implementations in the mid-60s.

I went back and listened to the tryouts of the self-made TTSH that Jesper Pedersen posted on SoundCloud a while ago.¹ (The Two Thousand Six Hundred is an unusual DIY clone of the ARP 2600.)

It’s interesting to think that while preceeding it by a year, the modular 2600’s production run is more or less parallel to that of the Oddysey. It’s a very particular sound those ARP instruments have!

In contrast to the punch of the first sequencer tryout that Jesper posted there’s also a warm drone and a more mellow sequence with some nice intonation and modulation details that appear as it progresses.

Scratch the measurements from the last two days. I’ve been puzzled by the frequency readings I’ve gotten in Audulus – made using the Mic (External Input) ¹ and ZeroCross nodes. (Calculating the deviations in cents was also done in with a little patch using the Audulus Hz2o (Hz to octave) module.)

I found a little tuner app that displays deviations in cents nice and clearly, and tried making some measurements with it directly via my iPhone’s microphone. I could set the ODDYSEi master tuning back to 440 and have the frequency register as a perfectly tuned A4 on the app – which is what one whould expect. The ODDYSEi Equal tuning matches equal temperament perfectly according to the tuner – which makes sense, also aurally! It turns out that Analogue tuning is calibrated to a near perfect A3 (220Hz) and gradually gets sharper as one moves up over the next 3 octaves. Inversely, moving down gradually gets flatter.

ARP ODDYSEi Analog tuning deviations in cents over 6 octaves:

(The measurements for the lower 2 octaves fluctuated a lot and were therefore more tricky to notate precisely.)

Here are yesterday’s measurements of the ARP ODDYSEi Equal and Analog tunings with the deviations from equal temperament indicated in cents.

Here’s a quick comparison of the differences between the Equal and Analog tunings on the ARP ODDYSEi – and how they compare to strict equal temperament. I’ve only measured one octave up from A = 440 Hz and am curious to measure how the tuning behaves in the other octaves as well. Interestingly I had to set the master tuning to 437.6 Hz on the ARP in order to get a measurement of 440 in Audulus. And the analyzer in MC Studio is showing pitches roughly 4 Hz below the equal measurements taken with Audulus. I’m not sure what’s going on there. Something to look into.

The next step is to calculate the deviations in cents.

I‘ve been taking a look at putting together a quantizer for custom tunings in Audulus – there’s been a little activity on that front in the forum. I have in mind something that builds on the small ratio modules that I was using to tune a sequencer in a previous post. Something very flexible that can be changed on the fly with the possibility of additional ‘hand-tuned’ deviations from the basic ratios. A kind of hybrid between systematic and empirical tuning.

Since it’s taking me a little time to figure it all out here’s a little experiment with changing the tuning on the ARP ODYSSEi from Equal to Analog while running a simple sequence. The deviations are subtle but nevertheless quite noticeable (often a little flat in relation to equal temperament) – at some point it could be interesting to measure them precisely and put together a little list that could be used in other contexts.¹

Here’s the unadorned sequence – two cycles in equal temperament followed by two with the analogue tuning:

And a longer jam trying out the analogue tuning at different transpositions as well as playing with the length of the sequence.

The ARP Odyssey marks is an interesting point in the development of sythesizers with some of the flexibility of the modular systems it grew out of and the fixity of a portable performance instrument that became the norm. I’ve been thinking about the development of synthesizers over the years – the gradual change from modular systems that lent themselves to ever-evolving textures to banks of presets and large libraries of relatively short sounds. The Art + Music + Technology episode with E-Mu’s Dave Rossum provides a good example of this arc – and the spiral back to the beginning that we’ve now reached.

Benge’s 20 Systems presents 20 pieces of music created with 20 different synthesizers, one from each year between 1968 and 1988. Starting with the Moog Modular 1968, moving through ARP, Serge, Roland, Oberheim, and Yamaha in the 70s, and ending with the Fairlight and Synclavier, amongst others, in the 80s. I was thrilled to receive the CD as a gift when it first came out nearly a decade ago, feeling that it nicely mirrored the first 20 years of my own life. It too provides a good trace – a journey in sound of the arc from evolving textures to precise presets.¹

It seems to me that a large part of the increase in activity on the Audulus forum during the last half year has been due to recreations of hardware modules/instruments. The hardware models seem to provide a useful grounding – something to kick against in the open space of the software platform. In the absence of owning any hardware syths at the moment I’ve also found the fixity of app recreations – the Moog, Ripplemaker and VCS3 apps that I’ve looked at, a useful counterpart to the openess of instruments/devices created in Audulus. I do keep coming back to Audulus though – it still ends up being the place to get on with some real building after diving into the behaviour of the other apps.

A few weeks ago I had the Korg 2015 version of the ARP Odyssey under my fingers on a visit to a music shop, and now the Korg ODYSSEi also has it’s place on my iPad. While not exactly a modular synthesizer it has a wealth of routing options that make it extremely flexible, while at the same time practical to use as an instrument. GForce have created their own software version of it – the Oddity – and also provide a wonderful introduction to the original as part of their synth archive. As they mention in the video this can be a bit of an in-your-face punk instrument. The opposite of the distant spaciness of the (i)VCS3.

The iPad ODDYSEi sounds great, and especially with the white Rev. 1 face is a visual pleasure as well. The fader interface design¹ (there are no knobs on this synth) seems well suited to an iPad touch-screen. I was delighted to find that the analogue tuning has been modelled as well. One can shift between pure equal temperament and the analogue version if it and clearly hear the differences. Richard D. James talks about layering similar differences between equal temperament and the analogue tuning of the Roland SH101 in his interview with Tatsuya Takahashi.

One of the recent guests on Darwin Grosse’s Art + Music + Technology podcast was E-Mu Systems Dave Rossum. Along with tracing the beginnings of the E-Mu Modular he also mentions the modular sequencer developed for the system. The idea of a sequencer broken into smaller component modules intrigued me, but I haven’t been able to find much information on it on the web.

What I did find was a video of Benge demonstrating it.

Benge also provides a grand tour of basic sequencer techniques demonstrated on Moog, ARP, Roland, Serge, and Buchla modular systems, as well as a little diversion with the VCS3.

Following up on Benge’s sequences I came across this thorough demonstration of classic Berlin School sequencing with a Doepfer MAQ16/3 and a Synthesizers.com Q960 (a recreation of the Moog 960).

While Benge uses a quantizer after the 960 Martin Peters tunes his sequencers by ear/hand – and that’s the possibilty that intrigues me. Hand (micro)tunings, and the possibility of setting up special tunings for a sequence without having to account for a complete system as one would be inclined to when thinking in terms of a keyboard.