New images in the gallery

I just added some more of my fractal artwork to the gallery page. Go check them out!

Upcoming Sevish album – Morphable

My new album Morphable will be available on 5th August 2022. Nine new electronic microtonal beats from the harmony hacker.

Front cover artwork for Morphable album by SevishIf you want to hear one of the tracks today then you can! Yeah Groove can be played on youtube:

Scale Workshop 1.5 released

Scale Workshop has been updated this weekend. Let’s take a look at what this useful microtonal web tool is capable of.

Screenshot of Scale Workshop 1.5

Launch Scale Workshop in a new tab

Key functionality

  • Create your own microtonal scales by direct data input, or use one of the many generator methods to aid your discovery.
  • Export tuning files in a large variety of useful formats.
  • Hear the scale by using your qwerty keyboard as an isomorphic keyboard, or use MIDI input (requires a browser with web MIDI support).
  • Import Scala scl file and convert to various formats.
  • Collaborate and share your scale easily by copy and pasting the URL which automatically has your scale encoded.

Recently added features


With a web MIDI compatible browser, you can use Scale Workshop to enable microtonality on your hardware synths and sound modules. This is achieved by 16 channel note output with pitch bend on each channel.

Rotate modifier

Rotate your scale so that a different interval becomes the new 1/1.

Improvements to the virtual keyboard

When using the mouse cursor, you can now play the virtual keyboard stylophone style, i.e. click and drag across the keyboard to hit a sequence of notes.

Export REAPER Note Name Map

REAPER supports custom piano roll layouts. You can now export a txt file from Scale Workshop to import directly into REAPER.

Export tuning to Korg Monologue/Minilogue XD

Export your tuning to Korg Librarian format. This can be imported into the Korg Librarian software so you can write it to your synth.

Better precision

Scale Workshop now has better handling of large numbers and ratios.

Better synth

Issues with the synth audio dropout are resolved. Various new waveforms are added. The default waveform is changed to semisine which is more ideal for auditioning tunings than the previous triangle.

User guide updated for 1.5

As this version is a major milestone for the app, we have made sure to update the documentation to cover all the new features.

The future of Scale Workshop

Early development work has started for Scale Workshop 2. This involves a complete rewrite from scratch and a new UI. The project will remain on the permissive MIT License so that synth developers can re-use parts of Scale Workshop’s code to add microtonal functionality to their own projects. Scale Workshop 2 is intended to be released when it reaches feature parity with Scale Workshop 1.5. Scale URLs will remain backwards-compatible. Scale Workshop 1.x will receive no new features except for bug fixes.

Get involved

Watch Lumi’s music theory analysis of Gleam

Gleam is perhaps the iconic Sevish track (or so I am told), but have you ever wondered how it was made? Lumi’s recent video analyses Gleam from a music theory perspective, explaining how I used 22-tone equal temperament (22ed2) to make a catchy piece of music.

Sevish’s scale impressions

Here are some of my thoughts on various microtonal scales. These thoughts are my own subjective impressions and there’s no need to take them seriously. Enjoy!

Equal-step tunings

  • 1edo – wide-open featureless space
  • 2edo – dual opposition
  • 3edo – dreamy flashback, bells
  • 4edo – cartoony horror, staircase
  • 5edo – open, bright colour, ease
  • 6edo – cartoony wonder, tension
  • 7edo – flat, comfort, bright colour
  • 8edo – moist shit
  • 9edo – bells, chimes
  • 10edo – vivid, open
  • 11edo – pure, warped, intuitive
  • 12edo – nostalgia, comfort, jazzy, transparent
  • 13edo – warped liminal space
  • 14edo – multicoloured
  • 15edo – wonky
  • 16edo – expressive
  • 17edo – angular, intuitive, colourful
  • 18edo – bright colour, I still haven’t explored this one
  • 19edo – flat, wonky, uncanny, pale
  • 22edo – chill, trippy, hyper-jazz
  • 23edo – bright colour, unstable
  • 24edo – something you think should be familiar is now wretched, hyper-jazz
  • 29edo – expanded wonky wonder
  • 31edo – rest, clarity, transparency, and once you get this far out, each edo has the capacity for extreme variety
  • 314edo – way too many notes. this is the point where you’ve gone too far

Bonus, here are some temperament impressions too

  • blackwood – jazzy, dirty, trippy barbershop pole effects
  • machine – sparkly, pure, unusual
  • mavila – face swap, expectation and surprise
  • porcupine – an otonal- and Arabian- influenced diatonic without being either of those things

What are your own impressions?

Sevish music at the British Library sound archive

The British Library sound archive seeks to collect sound recordings that can be preserved for generations to come. They are home to one of the largest collections of vinyl recordings in the world. They also curate a library of digital recordings. The collection includes not just music but also drama, literature, field recordings, and oral history.

A photograph of shelves filled with sound recordings

A representative from the library reached out to ask if I would donate some of my recordings to the sound archive. Of course I was happy to do that. It’s fun to wonder what some futuristic person would think if they discovered early 2000s xenharmonic bangers.

Collection of tuning and microtonality related bookmarks

It’s a bit oldschool, it’s a bit Web 1.0, but I think it could be quite helpful – I’ve started collecting a list of bookmark links! The links are on-topic for Sevish music: tuning practice, tuning theory, fractals, music playlists, etc.

Have a link to suggest? Send me a quick message on my contact form.

On “equal” divisions

In discussion of musical tuning there is often talk about “equal divisions” of an octave or other interval. There is potential here for confusion as there are two different ways of equally dividing an octave.

Wait, two ways? You’d think there was only one way to divide an octave equally. You make all the notes the same size! But I’m being serious. You can equally divide an interval either arithmetically or logarithmically.

Arithmetically equal = each step is the same Hz difference from the next
Logarithmically equal = each step is the same ratio difference from the next

Arithmetically equal division of an octave

Let’s say you have two notes, A 440Hz and A’ 880Hz. The interval between these two notes is an octave. If I want to divide this octave into arithmetically equal notes then we simply split it in a way that the difference in frequencies are equal.

For example we could split it into 8 intervals. We get the difference between the two notes by subtracting the frequencies. 880 – 440 = 440. Then divide this into 8. 440 / 8 = 55. So the base step size is 55Hz.

Constructing the scale is done by starting from 440Hz and then adding 55Hz each time until we reach the octave at 880Hz:


Each note is 55 Hz apart from the next. You can move up and down the scale by adding/subtracting 55 Hz. The result is the harmonic series segment 8:9:10:11:12:13:14:15:16 with a fundamental frequency of 55 Hz. This scale sounds nothing at all like the equal tempered scale, in fact it sounds similar to the scales you get by playing harmonics on a guitar string.

To hear what it sounds like, you can follow this Scale Workshop link – it will open a page where you can press qwerty keys to play in this tuning.

Logarithmic equal division of an octave

Let’s take the octave from A 440Hz to A’ 880Hz and look at it as a ratio. 880/440. This can be simplified to 2/1.

To divide 2/1 into 12 logarithmically equal steps we need to find the step size. 2/11/12 = 1.05946309436 (approx).

Constructing the scale is done by starting from 440 Hz and then multiplying this value by the step size 1.05946309436 twelve times until we reach 880 Hz:


These steps don’t look equal in terms of frequency – the Hz values get larger with every step. Maybe by now you have noticed one way in which the steps are equal… Each step is an equal ratio difference from the next.

To hear what it sounds like (I mean it’s just 12edo at A440Hz), you can follow this Scale Workshop link – it will open a page where you can press qwerty keys to play in this tuning.

Comparing these two ways of dividing an octave equally

In an equal temperament, you can modulate between keys and every key will sound equally in-tune. Whereas arithmetic divisions will only give you perfectly-tuned harmonics of a single fundamental with no concept of tonal modulation.

Which of these two methods sound equal to the ear?

It’s the logarithmic version – also known as equal temperament or EDO (equal division of the octave).

When tuning theory folks mention “equal” scales, safely assume that they’re probably talking about this method.

Equal temperaments sound equal because our perception of pitch is logarithmic itself. The hearing system isn’t listening out for equal difference in Hz, it’s listening out for equal difference in step size ratio. This isn’t immediately obvious until you’ve compared the two for yourself.

So this was just a quick and simple post to explain something that has caused a little confusion in the past. Hope someone will find it useful.

Project OutFox rhythm game features Sevish track “Some Things Must”

Project OutFox is a rhythm game simulator that brings together over 15 styles of gameplay. The OutFox team have just released Serenity Volume I, a community pack of songs which you can now download and play for free. The pack includes the Sevish track Some Things Must from the album Bubble. It also contains music from Aspid Cat, Drazil, Jack5, Matduke, mmry, PizeroFox and Seo.

Newcomers to OutFox/Stepmania are advised to check online for setup instructions. A (probably incomplete) summary for first installation is below:

  • Download and install Project OutFox
  • Download Serenity and extract the files to the Songs directory where OutFox is installed
  • In the options screen, run the option to sync your machine
  • Select a game mode and then start playing

It might surprise some to know that I got into Konami’s Dance Dance Revolution and Andamiro’s Pump It Up some time around 2001 and it was some folks on the DDRUK IRC channel around 2003 who convinced me to start making my own electronic music. I quit playing rhythm games for a long time but I got a high quality DDR pad in early 2021 and the obsession has well and truly returned.

“Re-Tuned” demonstrates the aesthetic characteristics of musical tuning systems

I started a small project where I take existing Sevish tracks and adapt them to different musical tuning systems. It is called Re-Tuned and is available from my Bandcamp page to stream or as a free download. In this project, you’ll hear many tracks with altered tunings. Many of these are originally written as microtonal pieces that are re-tuned to standard 12 equal!

Why have I done this‽ The project originally started out as trolling youtube videos which rendered familiar microtonal tunes in 12edo. It turned out to be an effective demonstration that 12edo is just another tuning with its own character, and that this kind of retuning subtly or dramatically altered the feeling of the music. I then started to produce microtonal retunings, for example Desert Island Rain (originally in 313edo) is rendered in 24edo, 19edo, 14edo and 9edo.

My musical philosophy is that music is a fun craft to be involved with and I’m glad it exists. When expressing yourself through music there are various tools in the toolbox – for example dynamics, timbre, tempo, tuning – though tuning is often forgotten about. I hope Re-Tuned makes the case that deliberate use of tuning is one more powerfully expressive and aesthetic parameter that musicians can use to hone their craft further.

Cover artwork by Juha Penttinen.

A new original Sevish album is coming out next month…