Osmose Expressive E and the Uncanny Valley

The internet has been buzzing with demos of the Osmose Expressive E since they started arriving to VIPs studios earlier this year. I have been fascinated by it since 3D renders of it showed up in November of 2019. Four years later, I finally have it and now that I’ve had a day or two to allow my brain to reassemble itself I’m ready to say something about it.

There are many directions that artists will steer this machine. One is by leveraging physical modeling to emulate acoustic instruments. Doing this requires developing the techniques and having the knowledge to work the Osmose into matching the range and textures of the target instrument. Secondly it requires expertly designed patches that can translate the subtleties of the player’s expression into the expected nuances. Benn Jordan has a great video here that goes into detail about how this can be done. I do not intend to address the debate regarding “should this be done?” in this article, other than to state that there is an ongoing debate (perhaps since music was electrically amplified) along with far reaching consequences to musicians and the music industry at large of which we all ought to be aware. Continue reading

ISSTA Live Recording, September 2017

This recording was made during my appearance at the International Sound in Science Technology and the Arts (ISSTA) Conference in Ireland on September 8, 2017. The piece is a rendition of my composition, Vocalise Sintetica, first performed at Echofluxx14 in Prague. This piece is written so that it is allowed to evolve in a number of ways each time it is performed. Here’s how it changed this time around.

First of all it uses the Audiovisual Grain Machine (AVGM), which I update frequently. This time the updates were minor improvements to speed and efficiency. However, I did add some new audiovisual content. Secondly, in order to travel lightly, I limited the AVGM accompaniment to a single Novation Circuit. I loaded the Circuit with custom patches and samples, and used my Minifooger Delay on the AVGM (I usually leave it dry), but that was it, sonically. Other than that the Arturia KeyStep helped add a few tricks (mainly arps and one drone) to the mix.

Novation Circuit Randomized Patches

my_circuit

In my mind, sound design is at its best when it is a process of discovery. At its worst it can be an unfortunate exercise in mimicry. I am fascinated by the process of discovering sound through happy accidents. One of the techniques I have exploited frequently in this regard is synthesizer patch randomization. For example, the Yamaha TX81Z sounds great when randomized, or better yet, “degraded” with shuffled parameter values interpolated based on a time unit or clock division. The PreenFM2 has patch randomization built directly into the instrument!

So, it wasn’t long after picking up a Novation Circuit that I had the urge to use a similar shortcut to mine fantastic and otherworldly sounds from the unit. Full MIDI specification for the Circuit is available so that development of a standalone randomizer is possible, but Isotonik Studios published a free Max for Live editor in partnership with Novation. Max for Live patches are inherently editable so I decided to start there.

Send Random Values

It took me a couple of hours to get into the guts of the editor and setup a drop down menu for randomization. The drop down has choices to either “randomize all” (not quite all parameters), or randomize one of seven sets of grouped parameters like the oscillator section, mod matrix, or LFOs. At his stage I haven’t included the EQ section, voice controls, or macro controls. I probably won’t add the EQ, but the macro controls might offer some interesting possibilities. The image above shows a simple subpatch I made that takes a bang and outputs the random values for the oscillator section. Unfortunately, I can not legally share my mods based on Isotonik’s and Novation’s EULAs. However, you’ll need little more than a basic understanding of Max to do this yourself. Checkout the video and let me know what you think in the comments.

Cymatics Test: Mustard Seed and Max

After a few attempts and creating cymatics with the WSG, we switched to creating a simple Max patch that we used to generate the frequencies. This allowed us to isolate specific frequencies that worked well to excite the mustard seed on the platform. This time it is much easier to hear the hissing sound of the mustard seed as is vibrates on the platform. It sounds a little bit like white noise, but brighter and less consistent

Cymatics Test: Mustard Seed and Max

.

Glassy Pad from the Roland D-50

Before I continue I must admit that I have always had a problem with the term “pad” as a catch-all term for sustained synthesizer textures. I also have an aversion to using sounds that are described in that manner. I’m not exactly sure why, but it might have to do with the idea of padding. Padding is an unnecessary stuffing use to protect things from hard edges, or prevent delicate items from being broken. I’ll stop there with the metaphors, but my aversion comes down to not wanting to produce work where any sound is considered filler. However, I realize that this term is impossible to escape, so over the years I have tried to embrace it, but it still doesn’t sit right for me. In any case here’s a “sustained synthesizer texture”, produced by the Roland D-50, that I’m quite fond of.

Glassy D50 Pad