New Free Open Source Synth, Terrain, By Aaron Anderson

[_Tmas_]

Microsoft Defender virus alert on Win11.

[HiEnergy]

Is there a slight mismatch in the envelope parameters? To me it sounds as if even the longest possible decay is shorter than a medium release.

BTW: I've suggested Terrain as a candidate for an OSC. It's too quirky (yet still manageable and beautiful-sounding) to pass on it.
Awesome work, Aaron!

[El°HYM]

Hi!

I like the synergy effect in this thread.

Very cool.

[aaron_anderson]

I actually hadn't heard of ADAA. This is really interesting, thank you for bringing it to my attention. It would seem that a 'trivial' implementation would be possible in Terrain; take the first or second order antiderivative of both the trajectory and terrain function independently. Each successive antiderivative order seems to increase the computational cost; likely not near as much as the oversampling however. It would be a nice feature to provide both orders of antiderivatives as options in addition to oversampling. I would really have to brush up on my calculus to pull that off though.

We use it quite a lot in surge. Both in vcos and waveshaoers. Feel free to fire me questions on discord or in pm if you want examples but it works really well.

Key insight is that the numerical derivative of the analytic integral is equivalent form to alias supressing filters but happy to share papers code blogs etc

From playing with terrain i would think a blit approach may also work (I haven’t read the code but I assume you are just putting position on surface into output and there may be better techniques to reconstruct)

Anyway as well as the tuning channel we have a dsp channel and always happy to shoot the you know what with friendly oss devs

Thank you for your offer to help on this! Frankly the mathematical/DSP implementation needs major work in Terrain; I use a trivial approach to generate all trajectories and terrains. I've suspected a BLIT approach would be possible for the trajectories but have not begun to approach this. Maybe with the help of the Surge DSP community Terrain can have a complete synthesis engine overhaul!

[aaron_anderson]

Thanks Aron for Terrain,
what is the difference to sonic-lab.com/oceanic ?

I hadn't heard of Oceanic; cool app, thanks for showing it to me!

The biggest difference is that Terrain uses a oscillating 2D Trajectory while Oceanic uses a stationary Probe (series of probes actually) while the terrain (ocean) evolves around it. Oceanic uses the reads from the probes over time to generate 2D wavetables. These 2D wavetables evolve over time as the ocean waves progress. That's about the extent I could glean from the documentation.

[aaron_anderson]

Is there a slight mismatch in the envelope parameters? To me it sounds as if even the longest possible decay is shorter than a medium release.

BTW: I've suggested Terrain as a candidate for an OSC. It's too quirky (yet still manageable and beautiful-sounding) to pass on it.
Awesome work, Aaron!

Thank you!

There is an intentional difference between the time lengths of envelope parameters:
Attack: 2.0ms - 2000.0ms
Decay: 2.0ms - 1000.0ms
Release: 10.0ms - 4000.0ms
I chose these varied ranges based on what I believe to be practical ranges for these parameters.

PS: What is an OSC?

[jbraner]

Ver 1.0.3 works great with Reaper presets now
Thanks.

[Borbolactic]

Is there a slight mismatch in the envelope parameters? To me it sounds as if even the longest possible decay is shorter than a medium release.

BTW: I've suggested Terrain as a candidate for an OSC. It's too quirky (yet still manageable and beautiful-sounding) to pass on it.
Awesome work, Aaron!

"...PS: What is an OSC?"

Just a guess that it's a Surge XT oscillator and/or oscillator algo...

"Oscillator Algorithms

Surge XT provides 12 different oscillator algorithms, each capable of generating sound in different ways with a different set of controls. They’re not just different waveforms." ~ Surge XT manual

These days, I guess a lot can be packed into one, maybe an entire synth, such as how all the controls can change once a new oscillator is selected.

[elassi]

Is there a slight mismatch in the envelope parameters? To me it sounds as if even the longest possible decay is shorter than a medium release.

BTW: I've suggested Terrain as a candidate for an OSC. It's too quirky (yet still manageable and beautiful-sounding) to pass on it.
Awesome work, Aaron!

"...PS: What is an OSC?"

Just a guess that it's a Surge XT oscillator.

My guess is KVR's "One Synth Challenge" (OSC).

[Borbolactic]

Is there a slight mismatch in the envelope parameters? To me it sounds as if even the longest possible decay is shorter than a medium release.

BTW: I've suggested Terrain as a candidate for an OSC. It's too quirky (yet still manageable and beautiful-sounding) to pass on it.
Awesome work, Aaron!

"...PS: What is an OSC?"

Just a guess that it's a Surge XT oscillator.

My guess is KVR's "One Synth Challenge" (OSC).

True enough.

[HiEnergy]

Thank you!

There is an intentional difference between the time lengths of envelope parameters:
Attack: 2.0ms - 2000.0ms
Decay: 2.0ms - 1000.0ms
Release: 10.0ms - 4000.0ms
I chose these varied ranges based on what I believe to be practical ranges for these parameters.

PS: What is an OSC?

With "OSC" I referred to the "KVR One Synth Challenge"

In my opinion the longest possible envelope decay should be as long as the longest release. In most natural instruments a sound takes longer to decay than to release.

Some examples:
A piano: Decay is while the the key is pressed and the damper is lifted, release is after the key is released and the damper falls back on the string
A wind instrument: Decay is while the player slowly runs out of air to keep the tone up, release is when the player stops blowing into the instrument
A bowed string instrument: Decay is until the length of the bow is used up, release is when the player stops bowing
A plucked string instrument: Decay is while the string rings, release is when the player stops the string by loosening the grip on the fret or touching the string with the right hand

I have yet to find an example for a natural instrument where decay is shorter than release.

[Tj Shredder]

An initial attack of a wind or brass instrument might be short and any thing after would be the sustain. Same for string instruments. But all those instruments which do not have a real sustain (plucked sounds), would need a longer decay than release - absolutely...

[foosnark]

This is very cool so far. A great drone synth when modulating parameters from Bitwig. The feedback is a cool feature, it really makes it fascinating to me as a nerdy explorer of sound

My wishlist would be:

- standard 48 semitone range for MPE pitch bend
- Assignable modulation targets/depths for velocity, pressure, etc. (I can sort of work around this with Bitwig)
- An audio rate modulation oscillator as a modulation source, even just a sine for this seems like it'd be fantastic if you could specify a ratio relative to the main frequency.

[aaron_anderson]

Hello everybody,

I just released 1.1.0 today. New features include presets, controllable pitch bend range, and CLAP support.
https://github.com/aaronaanderson/Terra ... /tag/1.1.0

[veloheinz]

thanks Aaron,
but don't make a mistake and integrate FX please.