Expressive Physical Modeling Synthesis With The LinnStrument

Composer and electronic music instrument pioneer Roger Linn shared this video, exploring the combination of physical modeling synthesis and his LinnStrument expressive controller.

He demonstrates some of the possibilities using Audio Modeling’s The Cello and the Sculpture synthesizer included in Apple’s Logic and MainStage.

Linn demonstrates how an expressive controller like the LinnStrument can be used to create expressive performances with physical modeling instruments. But he also goes beyond imitative synthesis, showing how the combination can be used to create performances that extend familiar sounds into unfamiliar territory.

10 thoughts on “Expressive Physical Modeling Synthesis With The LinnStrument

  1. What this really shows is how the conventional on/off (SPST switch) keyboard has held back the synthesizer. Yes, an organ-type keyboard is a cheap, ubiquitous device, but as a method of allowing a musician to interact with the actual expresivity of the musical instrument, it stinks. The pitch/mod wheels are, honestly, awful. There is so much that is just bad design because everyone has to work around the keyboard.

  2. @Brian You make a valid point that a standard velocity keyboard with channel pressure is quite limited in it’s expressive capability. Add wheels, breath-control, and/or even polyphonic-aftertouch, and you get more dimensions of control.

    TouchKeys is a very impressive way of adding expression to the traditional keyboard.

    There are two basic design challenges: 1. Note layout, 2. Expression on played notes.

    With note layout, the upside of traditional keys is that it makes entry-level playing VERY easy– key of C, all white notes. The downside is that there are 12 different key patterns for 12 different roots– for every scale. Of course, alternate scale layouts can be effective, like the Jankó layout. But those layouts present other challenges– not least of which is cost and risk– as you have to get people to adopt them. Non-key devices are great for some kinds of playing, but velocity is perhaps more difficult to control.

    With expression input, the first challenge is to have the motion be easy to control– putting 127 steps of control into a pressure sensor or a small dimension of movement means that the control will be difficult to harness. This is where the ribbon, B.C., or mod wheel are pretty easy to use with intention & finesse. The second challenge is that many of these control sources must start at zero, then have some intentional non-zero values, then back to zero again.

    If I had to choose an alternative, I suppose I’d choose a Jankó key layout with three “levels” per key. And some way of making it so it plays more like a traditional keyboard– less like a button accordion. PolyAT for sure. B.C. for sure. And option control via touch on the keys– or perhaps some opto sensors per key.

    1. @stub

      I come from an initial musical background of trombone and tuba. “Note layout” on the trombone is, literally, more tubing or less tubing, and pucker factor. For expression, the vast majority of traditional instruments outperform the synthesizer. Really, a wood box, strung with tanned sheep guts, rubbed with horse hair that’s been coated with tree pitch, has been used to entertain people for nearly 500 years, with no end in sight. How many synthesizers out there can approach, let alone beat, a violin for expression?

      By and large, the synthesizer falls in line with the harpsichord and organ for expression. If there were synthesizer keyboards that actually followed the real design of the grand piano mechanism, that would be great leap forward. Unfortunately, you’d also have a base $4000 for the keybed before the synth electronics were added on. Capturing the final dynamics of the hammer just needs a deflection sensor where the string would be.

      The Jankó layout on Wikipedia shows an upright piano. The upright mechanism has to be fully released before it can strike the string again. Given that, though, I can still do presto pianissimo with forte accents on it. It’s still better that pretty much everything out there for synth keyboards. I certainly can’t do that on my Korg Prologue.

      We use envelope generators because of the SPST keyboard. We have hardwired pitch and mod wheels when we should have two pads similar to the Expressive E Touche. Synth controls haven’t really changed in decades. Everybody does the same thing again and again, because it’s so cheap.

      1. I’m a trombone player, too. That’s why I like breath controllers so much. I don’t have as bleak an assessment of synths, though. But sort of bleak. Manufacturers could start by building instruments with proper velocity and polyAT. That would help. I think weighted-hammer actions do a decent job of emulating a piano.

  3. It’s nice to see a respectful “tribute” to modeling.

    Sculpture is one of my fave modelers. Would love to see a wind-based physical modeling synth– in plugin format.

      1. Thanks, @MrMidi

        I’m glad you reminded me of it. I’m pretty sure I heard of Respiro here on Synthtopia. What probably made me pass on it before was that I got the impression it was (for all intents and purposes) a preset-only instrument. Controls are “macros” and no under-the-hood tinkering is possible.

        However, this time around, as I listen to the demos, I really like what I’m hearing. At first, it’s difficult to listen to the music because I’m almost distracted by the “uncanny valley of the winds”. But then, as I listen to the tones, and imagine how it would feel to have that degree of control over the various qualities of sound, I’m convinced it is probably the right product for me, now. It has most of what I’m looking for, and the addition of convolution processing is a nice bonus.

  4. Another good modeler (wind included) is the AAS Chromaphone 2. It offers several approaches utilizing tubes, membranes and plucked elements. You can also blend two models, which opens it up a lot. Its clean GUI makes it far more approachable than Sculpture, which almost has a mad scientist vibe.

    Roger makes a compelling argument for the expressiveness of PM, but I have a lot of hours logged on a traditional keyboard. I’m considering the smallest ROLI, as I have a couple of MPE-ready synths, but commiting to a largely new playing method is no small thing.

Leave a Reply