Subscribe to Horizontalpitch via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 166 other subscribers.

Latest Articles



Design Notes on Colour Blindness – Part 2


In Part 1, I went through some of the background that lead Émilie Gillet and I to search for solutions to improve the usability of Mutable Instruments’ modules for people with colour blindness.

Let’s now see what solutions we considered, tested and eventually implemented, both to improve older modules and to design new ones.

Initially, there appeared to be a certain consistency in the perceived brightness of the panel colours. For instance, red would appear darker and turquoise lighter. So one solution that we considered was to create a separate mode where the LEDs would be brighter or darker in a way that could be linked to the labels’ perceived brightness.

Deuteranopia simulation in Adobe Illustrator

This was not a perfect solution because not all people we asked described their perception of the colours in the same way. While most people would say that purple appeared darker to them, for some it was actually brighter. Of course, purple and turquoise are quite different from red and green even if they are spectrally close, plus the most common form of colour blindness is deuteranomaly, which changes a bit from person to person.

It was a viable solution since the brightness of a LED was still much easier to see than its colour, but after a lot of experimentation, Émilie was able to find a better solution for some parameters: blinking patterns.

She explains it this way: “Sometimes, associations between the mode/parameter to be modified, and the blinking pattern could be found. For example, if the LED selects a number of polyphony voices (1, 2, 3, or 4), we can represent this parameter by the number of blinks/flashes. Or a smooth vs steppy option (or square vs triangle wave) can be represented by a smooth vs hard blinking pattern.”

On Rings, for example, this special colour-blind mode was implemented:

The polyphony selection button blinks once, twice, or four times to indicate the number of active voices.

The three resonator types are represented by three levels of brightness (normal, intense for the middle setting, and faint).

On Stages, where blinking was already in use to signal that a segment is looping, blinking and brightness was combined in this way:

RAMP: moderate brightness, with a faint brightness modulation. Smooth blinking pattern when the loop is enabled.

STEP: high brightness. Sharp blinking pattern when the loop is enabled.

HOLD: low brightness. Sharp blinking pattern when the loop is enabled.

Future Solutions

While blinking patterns and brightness modulation were an easy drop-in solution to make existing modules more usable for people affected by colour blindness, for future modules we decided to look into ways to avoid red-green LEDs altogether.

Here’s some of the solutions that came up over the course of our investigation:

Blue/red LEDs

These are similar to green/red ones but use a blue diode instead of a green one. The resulting hues are more easily visible by most colour blind people. Keep in mind that people affected by Tritanopia and Tritanomaly will not be able to see these as intended.


RGB LEDs can light up in a very wide array of hues. Colours can be chosen to be outside of the affected spectrums as much as possible and one could even design different colour schemes that the user can choose from. They do pose some challenges as well though. Again, Émilie explains this much better than I could ever do:

In the case of a digital module, RGB LEDs can be easily driven by the PWM pins of the microcontroller, allowing any colour to be displayed, with everything configured on the software side.
However, RGBs LEDs are more difficult to put into application in the case of an analogue module, where one would want, for example, to display one colour when the signal is positive, and another colour when the signal is negative. For this application, the alternative is to either use a pair of back-to-back LEDs, or limit oneself to the bicolour LEDs made available by LEDs manufacturers – which almost always means red/green.

Physical controls

Using a physical selector (switch or knob) to access modes will make LEDs redundant. These can still be used as an additional indication to display the actual parameter/mode after modulation has been applied.

This is basically how the big knob on Warps works.

Mutable Instruments Warps

Discreet LEDs

If the panel space is sufficient, nothing will beat using a LED for each mode, with icons or text clearly stating what they stand for. This is how some of the older modules work, for example, Grids and Peaks have discrete monochromatic LEDs to display the modes.

Alright Devices Chronoblob 2

Another more recent example is the Alright Devices Chronoblob 2, where each mode has its own LED and icon. Colours are being used to indicate which modes are mono and which are stereo, but that’s rather an added bonus, not necessarily an essential feature.

The recent Mutable Instruments Shades remake uses 2 LEDs for each channel, one to display the positive and one for the negative portions of the incoming signal (useful for modulations). This makes it even easier to see if the signal is positive or negative. A cyan LED is being used for the former, and a purple one for the latter.

When working with PCB material the icon or text can be incorporated in the LED display, rendering the colour an auxiliary but not essential tool in giving feedback to the player.

The Hexnverter Mindphaser would be perfectly understandable with all-white LEDs. Colour is merely used to emphasize certain aspects, positive is blue, negative is red and some of the LEDs change colour based on the waveform selected.

Hexinverter Mindphaser

Graphical displays

If everything else fails one can always use a display. But even in this case, one needs to pay attention to which colours are being used and how.


Often the space on the panel, or even more likely on the PCB, is very limited. This means that one needs to work with a combination of the above techniques and possibly combine them with blinking patterns or brightness modulation as an additional colour-blind mode.

A Practical Example: Beads

On Beads we initially wanted to have one LED for each mode/function as you can see on some of the iterations here:

Some of the iterations we went through with Émilie to figure out the final layout of Beads.

Unfortunately there wasn’t enough space on the PCB for all these LEDs, so we ended up going with an RGB LED for the quality setting. The LED lights up in the exact same colours as the labels on the panel. While turquoise and purple are still problematic to some people, they are easier to discern to many.

The CV select feature, which lets you decide which parameter the multi-purpose CV input should be routed to, has individual LEDs which light up under the chosen patermeter as you press the assign button.

Mutable Instruments Beads

Closing Thoughts

Colour Blindness is, of course, only one of many aspects that needs to be
taken into consideration, but it’s definitely one that should be
considered more by hardware manufacturers.

If you’re designing hardware interfaces, or you are thinking about
getting into this field, I hope this short insight into our design and
research practice will be useful and interesting to you.


Graphic designer, illustrator and soundmangler. He makes music as kurodama and as part of the electronic music duo kvsu. Together with his wife Elizabeth he runs Papernoise, a small design studio located in northern Italy, specialised in graphic design for the world of music.

Comments 0
There are currently no comments.

This site uses Akismet to reduce spam. Learn how your comment data is processed.