Ursa DSP Releases Lagrange Granular Delay VST (KVRDC 2016)

Dave Elton, developer and founder of Enginehouse Studios and Ursa DSP, has announced the release of Lagrange, his entry in this year’s KVR Developer Challenge.

Lagrange is a highly sophisticated stereo delay that uses granular methods – but don’t be mislead by that terminology! This is not a “grain delay” that will loop a set point in the delay buffer. Instead, the audio stream within the delay buffer is processed in very much the same way that a granular sampler engine would cherry pick “grains” (i.e. small bits of audio) within the sample and then play them back in accordance with a control signal of some kind.

Basically, the same thing is happening here, except within Lagrange, it’s all happening within the delay buffer in real time. The result is a very spontaneous, organic cluster of echoes that you can process in a variety of interesting ways to produce surprisingly natural (or unnatural) sounding spaces, and with a bit of feedback combined with the use of intuitive limiting controls, you can generate huge atmospheres and otherworldly soundscapes almost from nothing.

A brief list of Algorithms appear in a small menu within the “Echoes/Taps” section near the top of the interface, along with a group of controls that determine how Lagrange will process the grain stream. We’ll get to those algorithms in due course, but for now, let’s focus on the “Grain Rate”, “Density”, “Fixed” and “Variable” Delay controls, which may seem like simple delay parameters, but there’s a lot more going on here than you might have anticipated.

The Grain Rate determines how quickly the grains “swap”. With the Grain Rate turned all the way down, the grains are much longer and fade in & out very slowly – but that doesn’t mean that the delay time is slow! The Fixed Delay time might be be very fast! The Grain Rate just changes the rate of the granular behavior in the delay buffer. With the Grain Rate turned all the way up, each of the grains become much shorter and varied in length. The “Density” control can be used to adjust the space in between the grains, or at higher values, overlap each of the grains significantly.

The Fixed Delay control (obviously) has a fixed delay time, which is the minimum amount that each grain will be delayed by in milliseconds. Simple enough. But the Variable Delay has a direct relationship with the grain stream, basically randomizing the space between each of the grains in accordance with the selected algorithm; while lower values allow for more traditional fixed delay times, larger values make the delay grains sputter and dance around in fun ways.

Each of the delay grain algorithms (some of which are still in development) determine the overall behavior of the grains, among other things. By default, the “Clean Random” algorithm places the delay grains at a random point that is always at least as long as the Fixed and Variable Delay times combined. The result is a diffuse albeit very clear output.

The “Smooth Random” algorithm does something similar but starts new grains in an offset fashion between left and right stereo channels and also uses a modulated all-pass filter to help avoid harsh resonances with high feedback settings. I especially like this algorithm for ambient sound design.

Both “Flangoid” and “Phasoid” have a chorus-like effect by selecting grains in accordance with a moving pointer in the delay buffer. Even though the grains do not chorus or detune, the first and subsequent delay taps seem to have that general character.

“Marching Random” is in many ways a combination of the above and yet the odd one out. Here, each grain has a longer delay than the previous one, that is until it reaches the maximum Variable Delay sum and then snaps back to the original Fixed Delay time, which will attenuate the amount of randomization applied to each subsequent delay grain. Smaller values often produce strange artifacts and bizarre inharmonic overtones ideal for experimental sound design.

The feedback section is where a lot of the magic happens, often resulting in beautiful and unusual washes of stereo sound, especially with the feedback controls set to high values in tandem with the “Low/High Cut” EQ controls and two different limiters, which are essential for keeping infinite delay times from getting out of hand with a “Peak Limiter” for more immediate taming of transients and an RMS based limiter that provides more gentle control. One of the coolest features so far is a bright red ring that lights up around the knob to let you know when the gain reduction is applied.

Another awesome feature is a built-in “Side-chain Limiter” that ducks the wet level of the output based on the level of the dry signal input, which is something that I usually have to set up with an external envelope follower on the send bus, so it’s a huge time-saver. “Release” and “Threshold” controls allow you to fine tune the Side-chain Limiter’s overall behavior. Also, there’s a “Peak EQ” placed within the feedback loop with a small handful of controls to help you cut harsh resonant frequencies that tend to crop up all throughout the frequency spectrum in high feedback situations.

On the whole, Lagrange is one of the most experimental delay plugins I’ve ever used! The controls are a bit difficult to wrap your head around at first, but if you study the documentation provided within the “Help” window when you move your mouse cursor over any one of the controls, you will learn to appreciate just how special Lagrange is after you’ve managed to get some idea of what’s happening under the hood. Sure, you can just turn knobs and see what happens, but if you want to get the most out of what Lagrange has on offer, you should really get to know it.

I’ve used more delay plugins than I care to mention, so believe me when I tell you that Lagrange is one of a kind. You can really push the boundaries of creative delay processing with the different algorithms, and once you get used to the controls, it doesn’t take a lot of programming expertise in order to create polished delay sounds ranging from minimal ambiance and basic room effects to cascading walls of feedback. Also, the fully vectorized GUI makes it easy to work with, and text fields enable you to enter exact values for all of the controls.

I’ve already managed to create a small handful of presets and a few demo tracks providing a few brief examples of what Lagrange can do having spent just a few days fiddling with it, even though I haven’t even scratched the surface in regard to the Flangoid and Phasoid algorithms. In any case, I’m really looking forward to Lagrange’s overall progress and hopefully a much-deserved high ranking lead in this year’s KVR Developer Challenge!

More info: Lagrange (2.94 MB download size, ZIP archive, 64-bit VST/VST3 plugin format for Windows)