Extracts the lowest note from a chord progression to make a bass line. Take your chord progression, and copy it into a clip on your bass track, then run this transformation! This device uses the same chord detection logic as Chord Inverter, with an Attraction control that allows you to dial in the sensitivity of the chord detector.

You can adjust the velocity of the bass notes and also octave-transpose them!

This device is born straight from the process of Ableton wizard dnksaus. The idea is to be able to automate the process of creating alternate voicings of chords in your chord progression. This device is able to detect chords in the MIDI clip by looking for overlapping notes, then can modify them using two separate transformation processes: inversion and stretching.

Inversion allows the chord to perform a sort of rolling crab walk up or down the piano roll:

Stretching makes the chord wider by shifting notes both upward and downward:

This device grew out of a desire that I had for Ableton’s native Seed MIDI Tool. Seed generates a random cloud of notes, which is a useful way to get an idea started. The problem (in my opinion) is that the distribution of these notes in time can be very uneven. There might be a lot of notes at the beginning of the clip, and nothing at the end. So, I made Even Seed to solve this problem. It does so by first slicing the clip into segments, then generating notes within each of the segments. This ensures that the notes are more evenly distributed.

Even Seed also allows you to generate chords by forcing the notes that line up along the time axis. You can also control the pitches and velocities more finely by using a bank of sliders that acts as a probability distribution. Like all of my MIDI Generators, Even Seed respects the MIDI Clip’s scale, so the generated notes will always be in key.

An idiosyncratic rhythm generator that builds patterns from negative space. The initial inspiration for this device was my desire to create a rhythm generation algorithm that would be able to produce the regular, evenly-distributed patterns that the Euclidean algorithm is known for, as well as lopsided patterns that are unevenly distributed, with clusters of notes in a short amount of time. What ended up coming out was an algorithm that could do exactly what I hoped AND way more: a one-to-four track rhythm generator that can produce a range of interlocking polymetric patterns. It works great for percussion, obviously, but also Mask Pattern 2 delivers surprising melodic discoveries through the interplay of patterns running at different pitches.

I strongly recommend pairing this one with the Reduce transformer, which allows you to generate a crazy polyrhythmic pattern, then systematically reduce its density to arrive at something appealing.

Technical Explanation

This is one of those devices that is much easier to grasp through play than through a written explanation, but here’s my attempt at a written explanation:

There are two euclidean patterns: the “underlying pattern” and the “masking pattern”. To be honest, these two names are a bit misnomer-y, but it’s fine… bear with me. The interaction of these two patterns allows us to great four separate resulting patterns A, B, C, and D.

Put as succinctly as possible, these patterns are the following:

• Pattern A is the expression of the underlying pattern, masked by the masking pattern
• Pattern B is the inverse of pattern A
• Pattern C is the expression of the masking pattern, masked by the underlying pattern.
• Pattern D is the inverse of pattern C.

Maybe that makes sense. But I doubt it! If you’re still curious, read on. Of course, you might want to stop reading now and just go use the device. Or you could keep reading! Up to you…

Underlying Pattern

The underlying pattern is a euclidean pattern whose number of steps is equal to the length of the clip selection divided by the grid interval. The number of events in the underlying pattern is set by the Distribution parameter of the device.

Masking Pattern

The masking pattern has two parameters: Masks and Mask Length . Masks sets the number of steps in the pattern. These steps will be stretched such that the masking pattern’s length is the same as the underlying pattern. The Mask Length sets what proportion of each step is “filled in”. In the example above, Masks is 4 and Mask Length is 50%

With the underlying and masking patterns, we can create four output patterns using boolean logic as follows:

• Pattern A: Underlying and Masking are both TRUE
• Pattern B: Underlying is TRUE, Masking is FALSE
• Pattern C: Underlying is FALSE, Masking is TRUE
• Pattern D: Underlying and Masking are both FALSE.

The result is a set of four patterns, each of which fills the negative space left by the other patterns. Another way to think about it is that the device is distributing the available events across the four voices. The Distribution , Masks , and Mask Length parameters can be used to modify this distribution.

The Rotation parameter rotates Pattern A in increments of +1, Pattern B in increments of -1, Pattern C in increments of +1.5, and Pattern D in increments of Mask Pattern 4 -1.5, which allows notes in different patterns to overlap and makes lopsided patterns possible.

Each of these 4 patterns can be assigned to a separate MIDI note, can be enabled or disabled, and can have an independent velocity.

Instantly and effortlessly rearrange blocks of notes in a MIDI Clip. Absolutely incredible for resequencing your favorite breaks (mine’s Apache, mainly because of A Made Up Sound, which is just a perfect track), but there are so many other possibilities too.

Phrase Shuffler Pro 2 This device is a collaboration with fellow Max Certified Trainer Ned Rush. Ned is a prolific maker of YouTube videos about electronic music making in Ableton Live and Max. One of the things that Ned does incredibly well in Live is chop up breaks using Live’s audio-to-MIDI feature. To get a sense for what I’m talking about, check out his A to Z of Classic Breaks series. So Ned sent me a note with the idea of making a device that would automate the resequencing of the break slices, and explained one very important detail: jungle producers very often slice breaks into groups of 3 eighth notes, and then reshuffle the order of those slices. So we made a device called Phrase Shuffler that let you create these groupings of notes and then randomize the order, and holy moly, it works amazingly well. The 3 eighth notes thing is absolutely crucial – it just simply grooves much harder when you preserve the groups of 3.

Phrase Shuffler was free and pretty rough and prototype-y. The idea was that we’d get that one out, then think a little bit harder about a more fully-featured version. And that’s what Phrase Shuffler Pro is.

A quick breakdown of the features of PS Pro:

• Create groupings of multiples of time intervals (e.g. 3 eighth notes, or 5 sixteenth notes, or whatever), then reorder those groupings.
• Randomize the order of the groupings (like the original Phrase Shuffler)
• Manually and precisely resequence the groupings (after randomizing if you like!)
• Control the randomization with a probability distribution that sets the likelihood of each grouping will appear in the output.
• A “variance” control that allows you to achieve controlled subtle variation by mixing between the randomized output and the original clip.

The Reduce device subtracts notes from your clip by muting them. It is a quick and easy way to reduce the busy-ness of a clip. I made it after finding that I wanted a density knob on all of my generators and realized that density reduction is a great general-purpose transformation. Truthfully, this is something that Pattern Transform can do (along with a whole lot of other things), but this transformation deserves a simple, easy-to-use dedicated device.

The device uses a modified version of the Euclidean algorithm to control which notes are muted. With the Clustering dial at 0%, the device uses the Euclidean Reduce 2 algorithm and selects the notes evenly. Clustering allows you to shift the muting pattern so that notes that neighbor one another are more likely to be muted.

Stretch, compress, shift, wrap, fold, or clip ranges of pitch, velocity, and chance. This one started with my desire to be able to quickly and easily shift or stretch a whole batch of notes to match up with the notes I’m using in a drum rack. From there, it became clear to me how useful these range operations are for creating variations on clips. I also threw in a shaper function like the one in Phase Pattern for even more power.

This device gives you the option of five different range operations plus the shaper for pitch, velocity, and chance. Here’s an explanation of the range operations

• Shift: Slides the whole range of values up or down, keeping the relative distances between the values the same Range 2
• Stretch: Allows you to change the distance between the highest and the lowest value and stretches or compresses the range to fit.
• Wrap, Fold, and Clip: Reduces the difference between the highest and lowest values while wrapping, folding, or clipping the overflow.