Sonification Suite Good Practise Guidelines

Our team have performed research to understand good approaches for sound designs. Based on the team leadership's background, most of our research has focussed on astronomy communications; although members of our wider team come from other domains. Much of what we have learnt could be applied to sonification of data from any domain.

Please be advised, that there are many benefits for using sound with Special Educational Needs and Disability (SEND) audiences, for example, for those with low vision. However, there are extra considerations when preparing shows and events for SEND Audiences, which are still learning for ourselves.   

Our pre-recorded shows and The Sonification Suite are designed to enable sound as another modality of communication, beyond the standard visuals and narration used in most communications. This provides an added benefit to enhance the learning experience of various phenomena. However, without care, sound can also be uncomfortable, or at worst, distressing for audiences. For example:

• Surprises with sounds can be particularly distressing. Therefore, it is very important to provide an expectation of what is about to happen before playing a sound. Unless your goal is to make some sort of scary experience (please be very careful with this for SEND audiences)!

• Long length pure tone sounds are not comfortable for listeners. This is why the Sonification Suite does not provide a sine wave sound as a standard sonficiation approach. More harmonically rich sounds are generally more comfortable to listen to.

• Although human hearing can in principle detect frequencies of 20 Hz to 20,000 Hz, this is not the range we should use! Note that you can use the Spectrogram option in the Suite to visualise the power at different frequencies in your sounds. Some main points:

  • The full range is not perceivable by many audiences (e.g., older people loose higher frequency sensitivity above about 12,000 Hz). 

  • Lots of high frequencies are perceived louder by the listener (for the same amplitude) and can be very uncomfortable unless used very carefully. 

  • Below around 60 Hz we can just barely perceive as a pitch. The lowest frequencies can be an uncomfortable experience for many as they are cause a vibrating/thudding feeling. Use with care!

  • We suggest focussing on a range of around 250 Hz and 6,000 Hz.

• Multiple sounds that are in conflict with one another can be very difficult and can cause sensory overload. 

• All audiences can be be negatively affected by poor sound design choices, but these challenges and negative feelings/responses can be even greater for SEND audiences. 

We have provided some example "sound styles" along with the Suite, which have been tested with audiences. These can be a good starting point, but might not be suitable for all your goals and applications. We generally encourage you to test your own sound designs with representative audience members before rolling out into public shows.

Here are some pointers from our own research with family audiences when testing the Sonification Suite for Planetaria:

• Simple sounds, such as the 'Japanese Harp'* example style for astronomical light curve data, can be the most effective for more detailed learning goals. Using pitch to map to brightness is effective for audiences to hear subtle features, such as the different depths of planet transits. However, such styles are often not the favourite during out research from an audience experience perspective (when asked to choose favourites).
  *In our research we used flute notes instead of a harp notes for creating this sort of pitch-mapping. However, the flute was not popular, and therefore we used our own discression to replace this with the softer harp sound.

• For astronomical light curves our research found that the Sci-Fi style we provide was the most effective considering both learning and experience criteria together. However, the Twinkle and Power Hum styles were also often chosen as audience's favourites. 

Most of this guide focusses on using sounds in an educational context; i.e., if you want to use the sounds to communicate information. However, remember you can also use the sounds to create a sensory experience without needing to explain everything. 

If you are focussing on learning goals: Firstly it is important to explain the fundamental concept of representing data with sound. In several contexts this can be justified because so much is invisible to human's (e.g., most of the electromagnetic spectrum is outside of the visible range, gravitational waves are not light, dark matter does not interact with light, atoms are too small to be seen etc.). Therefore we are regularly converting data into something we can experience. Usually this is light we can see (e.g., we might convert X-rays into a blue picture or represent dark matter with purple blobs). For sonification, we are converting signals or data into sounds. Otherwise, we can justify sonification because: it can be interesting and fun; it allows us to experience data in a different way, which might help us understand; etc..

For astronomy shows: remember to state that sound itself does not travel through the vacuum of space!

For an idea of how you might introduce the concept of data sonification, you can watch the first few minutes of Audio Universe: Tour of the Solar System, which was extensively tested with audiences.

Whatever sound designs you choose: if your aim is to communicate some information it is important to explain to the audience what they are listening to and what features they should be listening for. For example, what do higher notes represent? Why is the sound getting more or less intense? 

It is better to introduce the concepts slowly, especially at the beginning when the audience are getting used to using their ears to learn information. Audiences are usually not used to using their ears to extract information! Later in shows we find audiences get more and more used to this. Therefore, we recommend using simpler sounds at the start with simple mappings to get the audience used listening. If you want to have the audience listen to more complex, multi-layered sounds, build these up slowly. 

Although explanations are important, avoid talking over the sounds as they are playing. Do this before, and/or after, the sounds. Playing sounds multiple sounds with explanations in between can be very beneficial.

If you are focussing on experience, and are less interested in learning objectives, you can focus less on the explanation, and simply let the audiences enjoy the sounds that have been created from the data!

Check out this dedicated page on using the Sonification Suite, and preparing related shows and workshops: with SEND audiences.