learning to talk

I just installed a new module yesterday called the Flame Talking Synth. I’d been eyeballing this thing for some time, first being intrigued by a standalone MIDI version Flame has had out for some time, but turned off by the price tag and, well, the fact that it was MIDI-based. I love talking synth sounds and it’s always fun to find ways of making stuff sound like a screwed up robot. This eurorack version wasn’t exactly cheap either, so when it was first released a few months back I decided to get the E350 Morphing Terrarium instead, knowing it had formant sounds in its wavetables, and believing I could get something close with that. Well, the formant sounds are lovely on the E350, but it’s not a screwed up robot. So in the recent purge, where I traded out or replaced seven modules from my synth, I sold the E350 and went ahead and grabbed this Flame Talky thing.

The Flame Talking Synth is based around a digital chip called the Speakjet. It’s sophisticated in interesting ways, and it’s got some interesting limitations as well. The module has three modes that each produce quite different sounds. These tracks focus on the “Phoneme” mode and “word” mode (the third is “synth” which is not about the speech but has it’s own sound and nuances. “Phonem” mode has dozens of simple speech sounds (for example, “tu,” “eyrr,” “uh,” “aw,” as well as sounds labeled things like “biological 2” and “Pistol Shot” which you can hear quite a lot in the carnival track below. “Word” mode allows the synth to say actual words like “techno” and, yes of course, “robot.” This is cool and all, but what’s fun is that these words are selected using CV, so they are playable the way a note on a keyboard is playable. For instance, G2 on a keyboard would “play” the word “robot.” But since using a sequencer on a modular synth like mine is not an exact science, a lot of what happens is, let’s say, gibberish-like. In the carnival track you can hear a couple of spots where it leaps into the words mode, but I can’t understand a thing it’s saying.

These tracks were recorded in the first twenty minutes after I installed the module. Basically, it’s random sounds created by running the Noisering and the Choices joystick into various CV inputs, controlling the pitch, the speed in which the thing “speaks,” the bend of the phonemes, and the actual words and sounds it makes. It’s just heaven. It came with a nice detailed manual that I’ve since read and I’m looking forward to attempts to actual get it to say things, and maybe even sing.

These two tracks are, as mentioned earlier, the Flame Talking Synth controlled with the Noisering, the Choices joystick, and a little bit of Pressure Points. The first track is fed directly into the also-new Pittsburgh Analog Delay module, which I’ll get a little deeper into real soon. The second track is run first through a ring modulator (µMod by Intellijel) and then to the delay. The third is self-explanatory. Ha ha.

Leave a Reply