An introduction to Max
Dom Aversano
What is Max / Jitter / Max for Live / RNBO?
Max is a visual programming language originally developed at the catchily named Institut de Recherche et Coordination Acoustique/Musique in Paris, better known as IRCAM. Now developed and maintained in San Francisco by the company Cycling ‘74, it is widely regarded as one of the quintessential tools for people who desire a deeper level of control over their sound and visuals. In short, if you can imagine it, you can build it.
Max/MSP is a standalone visual programming language that allows you to build and compose beautiful-looking instruments, effects, sequencers, and anything else you can dream up, without having to get into the intricacies of code. Instead, you use patch cords to connect different modules, in a manner that more resembles the physical world. Max is also capable of creating standalone programs.
Jitter is the visual part of Max/MSP, not a separate program. It allows you to create visuals, edit videos, or control lighting in the same patching environment in which sound occurs. There’s no need to find elaborate ways to make distinct sound and visual environments talk to each other — it’s built into this system. Your sound and visual environment are all under one roof.
Max for Live integrates the above into the highly flexible digital audio workstation (DAW) Ableton Live. This means you have audio programming, visual programming, and a world-class sequencer complete with effects and instruments combined into one program. It’s incredibly powerful, which helps to explain why it’s so popular. Whether you want to make an album, an installation, or a live show combining audio and visuals, it’s all contained within Max for Live.
RNBO
RNBO is one of the most powerful aspects of Max. Without having to write a line of code it allows you to export your Max patches for use outside of Max. Using RNBO you can create audio plugins for DAWs, programs that run on Raspberry Pi, and export to C++ or Web Assembly for building desktop, mobile, or web applications.
This video gives a neat overview of its capabilities.
Who uses Max?
Many well-known artists use Max such as Autechre, Holly Herndon, Pauline Oliveros, and Björk. All of these artists are pioneers who benefit from being able to customise and design their software, using an elegantly designed that makes it intuitive and enjoyable to do so.
Why learn Max?
Most people learn Max/MSP because they have a project that they want to create that cannot be realised in standard commercial software. While the thought of learning to program might be intimidating, Max makes it far more approachable and achievable for those who do not have years of their life to dedicate to learning to code in programming languages like Supercollider, Python, or C++.
How long does it take to learn?
As with all things, it depends to what level. If you have a basic understanding of synthesis and sound design then making a simple synth or effect should not take very long — perhaps a few weeks. To gain a solid foundation in Max takes time, practice, and some dedication, making 3-6 months a ballpark time-estimate for getting to grasp with the fundamentals.
How to get started!
First, you need a bit of motivation. The best of all is if you have a project in mind that you want to build. To witness something that you previously only imagined starting to take form is a great way to stay motivated and learn. Piece by piece you can start building your bespoke musical universe, inspiring you to push further and develop more of your ideas.
Finally, another great tutorial for exploring the potential of Max is Umut Elder’s course on stochastic music. Beyond the fancy name, stochastic music simply refers to music that contains various random processes. By having a degree of randomness enter into music it helps give one’s music and visuals a sense of unpredictability and spontaneity.
Finally…
Remember to have fun! If you enjoy what you’re doing you’ll do more of it. Happy building!
A Q&A with Max for Live expert Phelan Kane
Dom Aversano
Music Hackspace tutor Phelan is a Berlin & London-based music producer, engineer, artist, developer and educator. He currently runs the electronic music record label Meta Junction Recordings and the audio software development company Meta Function. As the course leader for our new Certified Courses program Learn Max for Live, we asked him to answer a few questions about Max for Live.
We’re delighted to be featuring your Certified Course on Max For Live, tell us a bit about who you are and how you’ve developed as a producer and sound designer?
My name is Phelan Kane and I am a Berlin and London-based music producer, engineer, artist, developer and educator. For almost thirty years I have been active in both the music industry and the contemporary music education sector, with a focus on electronic music and alternative bands. I have worked on music for artists such as Placebo, I have supported Depeche Mode on a European tour as part of the Fad Gadget band and I have taught music tech to members of Radiohead. I am one of the first Ableton Certified Trainers in the world and since 2021 I have been one of only three dual Ableton and Max Certified Trainers. I have released music on the labels Bulletdodge and Meta Junction Recordings and I currently run the audio software development company Meta Function.
Where did it all begin for you and how did you get your first break?
Back in 1994, my Tutor at Audio Engineer College recommended me for an assistant engineer position at a studio owned by a band member of his. From there I went to work on lots and lots of music with projects at big studios in the UK such as Townhouse, Mayfair, Livingston, Metropolis, Rockfield, Monnow Valley and for labels such as Universal, WEA, Polygram, Mute etc etc.
So tell us about this course what will you be covering?
This course will explore how to make your own sound design, synthesis and sampling devices in Max For Live via Ableton Live Suite. With these tools, you can create novel devices/plug-ins that no one else has access to which will empower you to create unique electronic sound for music, game and media projects.
What are some of the most salient bits of career advice you’ve been given along the way?
Explore as many tools and techniques as you can. Be open-minded and absorb all types of music and media.
Why did you choose to teach a live online course over a click-and-watch pre-recorded one?
Live online courses are much more rewarding for the students. They can ask questions and receive mentor support through their learning journey.
So what can students expect to learn from the course and the experience of studying with other peers online?
They will learn the necessary latest skills, techniques and first-hand industry insight into the methodology, workflow and tips and tricks to create unique devices and patches in Max For Live.
Check out Phelan’s certified course Learn Max for Live
Eduardo Pesole’s top 10 tips for Wwise and Unreal Engine 5
Dom Aversano
Music Hackspace tutor Eduardo Pesole is a Berlin-based sound designer, sound artist and composer He is a graduate in Sound Design from IED – Istituto Europeo di Design, Rome and his clients include Point Blank Games, 505 Games and Oneoone Games. As the course leader for our new Certified Courses program Master Game Audio with Wwise and Unreal Engine 5 we asked him to share his top 10 tips for Dolby Atmos mixing.
- Organize your hierarchy structure: A well-organized project is the key! Build a solid hierarchy structure using a proper naming convention for your files into the project, categorize sounds with names and color coding, that would help you to easily navigate into the project and quickly find what you need.
Reach out to a community to ask for help: Game sound designers have a nice community of people who help each other solve problems, give advice, share knowledge, etc… In my personal experience, it has always been a great resource for problem-solving and inspiration. here are some link ( For Italian speakers: https://discord.gg/ZZdDuCjY; For internationals: https://www.facebook.com/share/NRFdhVMcp68HFXZx/ ; https://www.airwiggles.com/c/general?post_login_redirect=https%3A%2F%2Fwww.airwiggles.com%2F )
Read the Audiokinetic blog and watch their Twitch channel: These 2 are other powerful resources, they’re full of articles and videos of professionals sharing their projects, workflows and tricks. There is a lot to learn from them.
Mind your soundbanks: If you’re working with soundbanks in Wwise you don’t want to mess them up! Soundbanks that are too heavy will take more time to be loaded from the game engine. Dividing them into categories, and loading and unloading them when required will substantially improve the performance.
Randomize assets: randomization is a great friend, randomize pitch, volume, triggering time of a sound, and everything you need to make the game sound as natural as possible enhancing immersivity.
Recycle your assets: Recycling your assets is a good way to save resources and time. sometimes you can just change pitch, add some in-engine effects, play with filters to make footsteps sound bigger or smaller for example and use them for different creatures in the game.
Use Multiposition: Another way to improve performance is using multiposition emitters, which allows you to place a sound effect in different positions on the map (for example lakes, torches, waterfalls) as many times as you want without using too many resources it will always count only as 1 physical voice playing.
Use convolution: Reverbs are essential to the immersiveness of a game, and the best thing is to use convolution reverbs to recreate the acoustics of a space.
Use built-in game syncs: Built-in game syncs can be very useful to drive real-time parameters for the 3D spatialization of game objects, enhancing the mix and immersivity of the game.
Be creative: Last but not least, creativity. Be creative while designing your systems, think out of the box, and explore the thousands of possibilities that these tools are offering you. Don’t be afraid to try and experiment while researching.
Hugh Neal’s top 10 tips for Dolby Atmos mixing
Dom Aversano
Music Hackspace tutor Hugh Neal has 20 years of industry experience, having mixed at top London studios such as Abbey Road and Metropolis, working with renowned artists like James and Paul McCartney. As the course leader for our new Certified Courses program Complete Guide to Dolby Atmos for Commercial Music Production, we asked him to share his top 10 tips for Dolby Atmos mixing.
1 – File Management and organisation
Get your files and audio properly organised and copied to the relevant folders. Make sure you know where all the relevant audio for the upcoming session is to be located should you need to add any additional audio later. Also, BACK IT UP, if you are not in the habit yet, you should really get this into a regular habit. Don’t just wait till you’re finished to back up! I’ve seen this go very badly wrong before.
2 – Get your session organised
I have a tendency to do all the “admin” organisation work before any mixing begins. So all my possible busses, routing, sends etc that I think I potentially will need, can be set up ready to go. I have a specific project I can import session data from which covers every possible variable I might encounter in a mix. I also have a colour format and specific order of audio tracks to help efficiency within the mix. Doing this even the day before, I find frees my mind up to allow for complete creative flow when I actually start to mix the project.
3 – Be familiar with your session’s audio stems for the mix
Take the time to familiarise yourself with all of the audio that has been provided for the mix (NB this is really only mainly applicable if you are mixing from an outside source, ie audio tracks prepared for mix, although it can be useful to do on your own productions if you haven’t gone through everything in a while) I find this process helps me to identify any rogue audio tracks that may have slipped into the session or split tracks up that might be better suited if they are processed separately. It also can be that you can potetially delete any audio tracks that may have accidently doubled up in the export or production process. Either way, you are now more prepared to start building the mix more confidently than having sounds coming from all over the place when you hit play.
4 – Set out an idea of the level of your special mix environment
This can be a very useful step, Identifying the potential special environment you are going to mix toward, plus now being familiar with your audio will allow you to begin to build up your mix without getting cluttered or overwhelmed. It also enables you to keep checking back that you are on your intended path but be aware that all good plans can change. You may find that as you begin to build your special mix, the special sentiment may change so don’t get too hung up on the initial plan.
5 – LFE management in the session
Be mindful of this bad boy, it’s easy to get carried away. LFE can really help your low end and weight, but good control over this is crucial. I advise you adopt an LFE approach you can easily adjust and control as the session develops. Having the LFE as an AUX send in the session can be very useful as you can adjust and even automate should you require.
6 – Be mindful of your headroom!
In Dolby Atmos, we will not know what the final playback format our end user will use, so as a result, we need to be very mindful of the headroom we occupy in the mix. We have strict delivery requirements that we cannot overshoot when it comes to final submission. In a full 7.1.4 mix, we will have multiple streams of audio available to enable lots of perceptible headroom space but as we fold down to 5.1 and finally down to Binaural 2.0 the headroom is very different at each stage, so to be mindful of what audio we place in and around space would be a wise consideration.
NB: the Renderer uses a 5.1 fold-down to monitor and track the LUFS level, so really we are only fully considering the headroom in LUFS from 5.1 downwards.
7 – When using speakers, check the different fold downs.
This is tricky obviously if you are only using headphones, although a very large portion of Atmos listeners are currently listening exclusively on headphones, there are more and more Atmos-ready devices coming onto the market. So like when mixing in stereo and checking your mono mix, it is strongly advised to have someone or yourself check the mix fold-downs to see if your mix is translating according to your overall vision.
8 – Keep an eye on the LUFS
We have two targets for LUFS levels in an Atmos Render File, one for the main speaker array fold-downs and one for the binaural headphone fold-down. It most cases these work pretty much in tandem but as mentioned before with the headroom, there can be some discrepancy between these targets than we would like if we did not get into the habit of keeping an eye on them during the mix. It can be a pain to try and level off a slightly “over ” headphone LUFS reading when the main mix levels are a solid LUFS and sounding good. It’s a lot easier to just keep an eye on these levels as we progress through the mix.
9 – Just because you can, doesn’t always mean you should!
There is a very real possibility we can overdo the “specialness” of an Atmos mix, I always work with the mantra “for the good of the song” if I feel elements are being placed or automated that distract from eth song or piece of music’s intent, maybe I should consider reigning it in little. It’s easy to get carried away when we have so much space to play with. Sometimes the intention of more is potentially better but sometimes we may also want to consider the old adage of “less is more”
10 – Don’t be afraid to keep it simple
There can be a real tendency to feel like we need to fill the space of an Atmos mix with all manner of exciting movement and effects but sometimes the simple approach can have more drama, effect and believability to the listeners perception than we may consider. There will usually be an Atmos moment/s in the arrangement where we can use the medium to really accentuate the depth, width and height but for the most part, simplicity can be so very effective. Placing the listener into a simple sonic blanket that they can meld into can be more powerful than a thousand effects wising around for the sake of it.
How an image led to an album
Dom Aversano
Curiosity takes you to new places. I arrived in such a place after contemplating what a highly complex polyrhythm might look like. As an instrumentalist, I am accustomed to placing limits on my thinking based on what is physically possible, but since the digital realm essentially removes most physical constraints, we enter a new world of disembodied possibilities. The following image — created in P5JS — is one such example.
This image depicts 750 polyrhythms juxtaposed one on top of another. The x-axis represents time, and the y-axis is the increasing division of that time. At the very top left of the image, there is a single point. The line beneath has two equidistant points — one at the top left and one at the top centre. The line beneath this has three equidistant points: then four, five, six etc. all the way to 750 divisions. To create this by hand would be painstaking, if not impossible, but coding it is simple.
When I saw the image I was astonished — it contained star-like shapes, mirror symmetry, curved lines at either edge, and a series of interesting patterns across the top. Despite finding the image fascinating, I could not find a use for it, so I shelved it and moved on.
A little while later I decided to share these images on Substack, hoping they might be of interest to someone. To bring the images to life I decided to sonify them, by building a simple custom program in Supercollider. The program quickly morphed into a rabbit hole, as when I tinkered with it I heard new sound worlds awakening. It wasn’t long before I realised I was halfway into creating an album that I had never intended to make.
What captured me about the music was the same as the images: they were humanly impossible. Performing 750 rhythms is completely beyond the capabilities of the human mind, but effortless for a computer. The result was music that was temporally organised, but with no meter or ‘one’ to resolve on. There was a logical flow of patterns, but nothing to tap one’s foot to. Using the harmonic series as a scale allowed vast clusters of tones that the chromatic scale could not accommodate. With this vast number of tones, the distinction between timbre, chords, and notes started to break down.
The idea that computers can unlock forms of artistic expression which lie beyond the capabilities of the human body was described eloquently by the late computer artist Vera Molnar.
Without the aid of a computer, it would not be possible to materialize quite so faithfully an image that previously existed only in the artist’s mind. This may sound paradoxical, but the machine, which is thought to be cold and inhuman, can help to realize what is most subjective, unattainable, and profound in a human being.
Molnar’s proposal that machines can provide access to unattainable realms of artistic expression seemed a strong counter-argument to the romantic notion that machines degrade the subtleties of human expression. Rather than having machines imitate human expression, in Molnar’s interpretation, they could express facets of the human experience that the limits of physicality prevented. The machine rather than deromanticising human expression could be a tool used to express subtle aspects of ourselves.
With this idea in mind, I delved deeper into the visual dimension. One day, it occurred to me that the original polyrhythm image could be visualised circularly. In this case, the rhythms would be represented as rotating divisions in space that could be layered one on top of another. The result was an image distinct from the previous one.
The process for generating this image: Draw one radial line at 0°. Then add two lines equidistant in rotation. Then add three lines equidistant in rotation. Then add four lines, and so on.
The new image looked organic, almost botanical. The big division at the top was matched by half the size at 180° and two more half the size again at 90° and 270°. The dark lines represented points of convergence. The lighter areas are spaces of less density.
I chose the image as the cover for the album since having artwork and music derived from the same algorithm felt satisfying and aesthetically appropriate. Had I not made the initial image I would not have made the music, or at least I would have arrived at it in another way at another time. That this process occurred at all remains a surprise to me, which I treat as a testament to the capacity for curiosity to take us to unknown places.
An interview with Interaction Designer Arthur Carabott Part II
Dom Aversano
This is Part II of an interview with interaction designer Arthur Carabott. In Part I Arthur discussed how after studying music technology at Sussex University he found a job working on the Coca-Cola Beatbox Pavilion in the 2012 Olympic Games. What follows is his description of how the work evolved.
Did you conceive that project in isolation or collaboration?
The idea had already been sold and the architects had won the competition. What was known was there would be something musical because Mark Ronson was going to be making a song. So the idea was to build a giant instrument from a building, which everyone could play by waving their hands over giant pads. They wanted to use sports sounds and turn them into music while having a heartbeat play throughout the building, tying everything together.
Then it came down to me playing with ideas, trying things out, and them liking things or not liking things. We knew that we had five or six athletes and a certain number of interactive points on the building.
So it was like, okay, let’s break it down into sections. We can start with running or with archery or table tennis. That was the broad structure, which helped a lot because we could say we have 40 interactive points, and therefore roughly eight interactions per sport.
Did you feel you were capable of doing this? How would you advise someone in a similar position?
Yeah, I was 25 when this started. While it’s difficult to give career advice, one thing I hold onto is saying yes to things that you’ve never done before but you kind of feel that you could probably do. If someone said we want you to work on a spaceship I’d say that’s probably a bad idea, but this felt like a much bigger version of things that I’d already done.
There were new things I had to learn, especially working at that scale. For instance, making the system run fast enough and building a backup system. I’d never done a backup system. I had just used my laptop in front of my class or for an installation. So I definitely learning things.
If I have any natural talent it’s for being pretty stubborn about solving problems and sticking at it like a dog with a bone. Knowing that I can, if I work hard at this thing, pull it off. That was the feeling.
How did you get in contact with Apple?
I was a resident in the Music Hackspace then and rented a desk in Somerset House. Apple approached Music Hackspace about doing a talk for their Today at Apple series.
I already had a concept for a guerrilla art piece, where the idea was to make a piece of software where I could play music in sync across lots of physical devices. The idea was to go around the Apple store and get a bunch of people to load up this page on as many devices as we could, and then play a big choir piece by treating each device as a voice.
Kind of like a flash mob?
Yeah, sort of. It was inspired by an artist who used to be based in New York called Kyle McDonald, who made a piece called People Staring at Computers. His program would detect faces and then take a photo of them and email it to him. He installed this in the New York Apple stores and got them to send him photos. He ended up being investigated by the Secret Service, who came to his house and took away his computers.
However, for my thing, I wanted to bring a musician into it. Chagall was a very natural choice for the Hackspace. For the music I made an app where people could play with the timbre parameters of a synth, but with a quite playful interface which had faces on it.
How did you end up working with the composer Anna Meredith? You built an app with her, right?
Yes, an augmented reality app. It came about through a conversation with my friend, Marek Bereza, who founded Elf Audio and makes the Koala sampler app. We met up for a coffee and talked about the new AR stuff for iPhones. The SDK had just come to the iPhones and it had this spatial audio component. We were just knocking around ideas of what could be done with it.
I got excited about the fact that it could give people a cheap surround sound system by placing virtual objects in their space. Then you have — for free, or for the cost of an app — a surround sound system.
There was this weekly tea and biscuits event at Somerset House where I saw Anna Meredith and said, ‘Hey, you know, I like your music and I’ve got this idea. Could I show it to you and see what you think?’ So I came to her studio and showed her the prototype and we talked it through. It was good timing because she had her album FIBS in the works. She sent me a few songs and we talked back and forth about what might work for this medium. We settled on the piece Moon Moons, which was going to be one of the singles.
It all came together quite quickly. The objects in it are actual ceramic sculptures that her sister Eleanor made for the album. So I had to teach myself how to do photogrammetry and 3D scan them, before that technology was good on phones.
You moved to LA. What has that been like?
It was the first time I moved to another country without a leaving date. London’s a great city. I could have stayed, and that would have been the default setting, but I felt like I took myself off the default setting.
So, I took a trip to LA to find work and I was trying to pull every connection I could. Finding people I could present work to, knocking on doors, trying to find people to meet. Then I found this company Output and I was like, ‘Oh, they seem like a really good match’. They’re in LA and they have two job openings. They had one software developer job and one product designer job.
I wrote an email and an application to both of these and a cover letter which said: Look, I’m not this job and I’m not that job. I’m somewhere in the middle. Do you want me to be doing your pixel-perfect UI? That’s not me. Do you want me to be writing optimized audio code? That’s not me either. However, here’s a bunch of my work and you can hear all these things that I can do.
I got nothing. Then I asked Jean Baptise from Music Hackspace if he knew any companies. He wrote an email to Output introducing me and I got a meeting.
I showed my work. The interviewer wrote my name on a notebook and underlined it. When I finished the presentation I looked at his notebook and he hadn’t written anything else. I was like, ‘Okay, that’s a very good sign or very bad sign’. But I got the job.
How do you define what you do?
One of the themes of my career is that has been a double-edged sword is it not being specifically one thing. In the recruitment process what they do is say we have a hole in our ship, and we need someone who can plug it. And very rarely are companies in a place where they think, we could take someone on who’s interesting, but we don’t have an explicit problem for them to solve right now, but we think they could benefit what we’re doing.
The good thing is I find myself doing interesting work without fitting neatly into a box that people can understand. My parents have no idea what I do really.
However, I do have a term I like, but it’s very out of fashion, which is interaction designer. What that means is to play around with interaction, almost like behaviour design.
You can’t do it well without having something to play with and test behaviours with. You can try and simulate it in your head, but generally, you’re limited to what you already know. For instance, you can imagine how a button works in your head, but if you imagine what would happen if I were to control this MIDI parameter using magnets, you can’t know what that’s like until you do it.
What are your thoughts on machine learning and AI? How that will affect music technology?
It’s getting good at doing things. I feel like people will still do music and will keep doing music. I go to a chess club and chess had a boom in popularity, especially during the pandemic. In terms of beating the best human player that has been solved for decades now, but people still play because people want to play chess, and they still play professionally. So it hasn’t killed humans wanting to play chess, but it’s definitely changed the game.
There is now a generation who have grown up playing against AIs and it’s changed how they play, and that’s an interesting dynamic. The interesting thing with music is, it has already been devalued. People barely pay anything for recorded music, but people still go to concerts though concert tickets are more expensive than ever people are willing to pay.
I think the thing that people are mostly interested in with music is the connection, the people, the personal aspect of it. Seeing someone play music, seeing someone very good at an instrument or singing is just amazing. It boosts your spirits. You see this in the world of guitar. A new guitarist comes along and does something and everyone goes, ‘Holy shit, why has no one done that before’?
Then you have artists like Squarepusher and Apex Twin who their own patches to cut up their drum breaks. But they’re still using their own aesthetic choice of what they use. I’m not in the camp that if it’s not 100% played by a human on an instrument, then it’s not real music.
The problem with the word creativity is it has the word create in it. So I think a lot of the focus goes on the creation of materials, whereas a lot of creativity is about listening and the framing of what’s good. It’s not just about creating artefacts. The editorial part is an important part of creativity. Part of what someone like Miles Davis did is to hear the future.
An interview with Interaction Designer Arthur Carabott Part I
Dom Aversano
If ever evidence was needed of the power of DNA it was demonstrated to me just over a decade ago, when I walked into a room in a dingy industrial estate in Hoxton, East London, to attend one of the early Music Hackspace meet-ups, and much to my surprise saw my cousin, Arthur Carabott, twirling on an office chair, listening to a presentation on ultrasonic sound.
The population in London at that point was roughly 8 million, and there were fewer than 20 people in that room — the odds of us both being there were minuscule. Although we both grew up in an extended family that was very musical, we came to the Music Hackspace by entirely distinct routes, at a time when it was little more than a charming and eccentric fringe group.
Having known Arthur since childhood, it’s not surprising to me that he ended up working in a field that combines artistic and technical skills. He always approached technical problems with a rare tenacity and single-mindedness. Several times I saw Arthur receive a Mechano toy for a birthday or Christmas, only to sit quietly for hours on end working on it until it was finally built.
The Music Hackspace played a significant part in both our formations, so I was curious to know about his experience of this. What surprised me was how much I did not know about Arthur’s journey through music.
What follows is a transcript of that conversation — Part II will follow shortly.
What drew you to music?
There was always music playing in the house. In my family, there was the expectation that you’d play an instrument. I did violin lessons at 7, which I didn’t enjoy and then piano aged 10. I remember being 10 or 11 and there was a group, there were a bunch of us that liked Queen. They are an interesting band because they appeal to kids. They’re theatrical, and some of it is quite clown-like. Then I remember songs like Cotton Eye Joe and singers like Natalie Imbruglia, you know, quite corny music — I’ve definitely got a corny streak. But there was this significant moment one summer when I was probably 11 or so, and I discovered this CD with a symbol on it that was all reflective. It was OK Computer, by Radiohead. That summer made a big musical impact. It’s an album I still listen to.
How does music affect you?
I think music, food, and comedy are quite similar in that when it’s good, there’s no denying it. Of course, with all three, you can probably be a bit pretentious and be like, ‘Oh no, I am enjoying this’ when you’re not. But those are three of my favourite things in the world.
I heard a comedian talking about bombing recently. They said if a musician has an off night, and they get on stage, they don’t play well it’s still music. Whereas if a comedian goes up and they bomb, and no one laughs, it’s not comedy.
You became a very accomplished guitarist. Why did you not choose that as a career?
I went to guitar school and there was a point in my teens when my goal was to become the best guitarist in the world. I remember something Squarepusher had on his website once, where he wrote about being a teenager and giving up on the idea of trying to be like his classmate Guthrie Govan, who is now one of the world’s best guitarists. I resonated with that as there’s a point where you’re like, okay, I’m never gonna do that.
Part of my problem was being hypermobile and therefore prone to injuries, which stopped me from practising as much as I wanted to. Yet, there was still this idea that when I went to Sussex University and studied music informatics with Nick Collins I was going to go there, learn Supercollider, and discover the secrets that Squarepusher and Aphex Twin used. Someone told me they don’t even cut up their drum loops, they’ve got algorithms to do it!
I was actually signed up to do the standard music degree but my friend Alex Churchill said to change it to music informatics as it will change your life. That was a real turning point.
In what way?
What clicked was knowing I enjoyed programming and I wasn’t just going to use music software — I was going to make it.
The course was rooted in academia and experimental art practice rather than commercial things like building plugins. We were looking at interactive music systems and generative music from 2006 – 2009, way before this wave of hype had blown up. We doing some straight-up computer science stuff, and early bits of neural networks and genetic algorithms. Back then we were told, that no one’s really found practical uses for this yet.
We studied people like David Cope, who was an early pioneer who spent decades working on AI music. All these things helped me think outside conventional ways when it came to traditional music tech stuff, and the paradigms of plug-ins, DAWs, and so on.
What did you do with this training and how did it help you?
I had no idea what I was going to do afterwards. I was offered a position in the first year of the Queen Mary M.A .T. Media, Art and Technology PhD, but I was a bit burnt out on academia and wanted to do the straight music thing.
I volunteered at The Vortex in London as a sound engineer. I had done paid work at university in Brighton but mostly for teenage punk bands. The Vortex being unpaid worked better because it meant that I only did it for gigs I wanted to see. I was already into Acoustic Ladyland, but there I discovered bands like Polar Bear and got to know people like Seb Rochford and Tom Skinner. I admired their music and also got to interact with and get to know them.
How did you come across Music Hackspace and how did it influence you?
I’d heard there was this thing on Hackney Road. I remember going on a bit of a tour because they would do open evenings and I went with a group of people. It felt like the underground. The best music tech minds in London. A bit of a fringe thing, slightly anarchist and non-mainstream. Music Hackspace was for me mostly about connecting to other people and a community.
What led you to more technical, installation-type work?
I remember seeing Thor Magnussen who had been doing his PhD at Sussex while I was in my undergrad and he taught one of our classes. He was talking about doing an installation and I remember thinking, I don’t really know what an installation is. How do I get one?
Then came the opportunity to work on the 2012 Olympics which came through my sister Juno, and her boyfriend at the time Tim Chave who introduced me to the architects Asif Khan and Pernilla Ohrstedt. I met them and showed them a bunch of like fun things that I’d made, like an app which took Lana Del Rey’s single Video Game and let you remix it in real time. You could type in every word contained in the song, hit enter, and she would sing it, remixed, in time with the beat.
They asked me various technical questions but after the meeting, I didn’t hear anything for a while. Then got a call in December 2011 from Asif. He asked, ‘Can you go to Switzerland next week?’ And I’m like, ‘Wait, am I doing this project? Have I got the job?’ He responded, ‘Look, can you go to Switzerland next week?’ So I said ‘Okay, yeah’.
So then it became official. It was six days a week for six months to get it done in time for the Olympics.
Part II of this interview will follow shortly.
You can find out more about Arthur Carabott on his website, Instagram, and X.
Dom Aversano is a British-American composer, percussionist, and writer. You can discover more of his work at Liner Notes.
A Q&A with AI regulator Ed Newton-Rex
Dom Aversano
In November last year, Ed Newton-Rex, the head of audio at Stability AI, left the company citing a small but significant difference in his philosophy towards training large language models (LLMs). Stability AI was one of several companies that responded to an invitation from the US Copyright Office for comments on generative AI and copyright, submitting an argument that training their models on copyrighted artistic works fell under the definition of fair use: a law which permits the use of copyrighted works for a limited number of purposes, one of which is education. This argument has been pushed by the AI industry more widely, who contest that much like a student who learns to compose music by studying renowned composers, their machine learning algorithms are conducting a similar learning process.
Newton-Rex did not buy the industry’s arguments, and while you can read his full arguments for resigning in his X/Twitter post, central to his argument was the following passage:
(…) since ‘fair use’ wasn’t designed with generative AI in mind — training generative AI models in this way is, to me, wrong. Companies worth billions of dollars are, without permission, training generative AI models on creators’ works, which are then being used to create new content that in many cases can compete with the original works. I don’t see how this can be acceptable in a society that has set up the economics of the creative arts such that creators rely on copyright.
It is important to make clear that Newton-Rex is not a critic of AI; he is an enthusiast who has worked in the machine learning field for more than a decade; his contention is narrowly focused on the ethics surrounding the training of AI models.
Newton-Rex’s response to this was to set up a non-profit called Fairly Trained, which awards certificates to AI companies whose training data they consider ethical.
Their mission statement contains the following passage:
There is a divide emerging between two types of generative AI companies: those who get the consent of training data providers, and those who don’t, claiming they have no legal obligation to do so.
In an attempt to gain a better understanding of Newton-Rex’s thinking on this subject, I conducted a Q&A by email. Perhaps the most revealing admission is that Newton-Rex desires to eliminate his company. What follows is the unedited text.
Do you think generative artificial intelligence is an accurate description of the technology Fairly Trained certifies?
Yes!
Having worked inside Stability AI and the machine learning community, can you provide a sense of the culture and the degree to which the companies consider artists’ concerns?
I certainly think generative AI companies are aware of and consider artists’ concerns. But I think we need to measure companies by their actions. In my view, if a company trains generative AI models on artists’ work without permission, in order to create a product that can compete with those artists, it doesn’t matter whether or not they’re considering artists’ concerns – through their actions, they’re exploiting artists.
Many LLM companies present a fair use argument that compares machine learning to a student learning. Could you describe why you disagree with this?
I think the fair use argument and the student learning arguments are different.
I don’t think generative AI training falls under the fair use copyright exception because one of the factors that is taken into account when assessing whether a copy is a fair use is the effect of the copy on the potential market for, and value of, the work that is copied. Generative AI involves copying during the training stage, and it’s clear that many generative AI models can and do compete with the work they’re trained on.
I don’t think we should treat machine learning the same as human learning for two reasons. First, AI scales in a way no human can: if you train an AI model on all the production music in the world, that model will be able to replace the demand for pretty much all of that music. No human can do this. Second, humans create within an implicit social contract – they know that people will learn from their work. This is priced in, and has been for hundreds of years. We don’t create work with the understanding that billion-dollar corporations will use it to build products that compete with us. This sits outside of the long-established social contract.
Do you think that legislators around the world are moving quickly enough to protect the rights of artists?
No. We need legislators to move faster. On current timetables, there is a serious risk that any solutions – such as enforcing existing copyright law, requiring companies to reveal their training data, etc. – will be too late, and these tools will be so widespread that it will be very hard to roll them back.
At Fairly Trained you provide a certification that signifies that a company trains their models on ‘data provided with the consent of its creators’. How do you acquire an accurate and transparent knowledge of the data each company is using?
They share their data with us confidentially.
For Fairly Trained to be successful it must earn people’s trust. What makes your organisation trustworthy?
We are a non-profit, and we have no financial backing from anyone on either side of this debate (or anyone at all, in fact). We have no hidden motives and no vested interests. I hope that makes us trustworthy.
If your ideal legislation existed, would a company like Fairly Trained be necessary?
No, Fairly Trained would not be necessary. I very much hope to be able to close it down one day!
To learn more about what you have read in this article you can visit the Fairly Trained website or Ed Newton-Rex’s website.
Dom Aversano is a British-American composer, percussionist, and writer. You can discover more of his work at the Liner Notes.
Music in the browser or app?
Dom Aversano
As The Bard famously put it, ‘The app, or the browser, that is the question.’
At some point, your inspirational idea for digital music will have to travel from the platonic realm of your thoughts, into either an app or browser. Unless you can luxuriate in doing both, this represents a stark choice. The most appropriate choice depends on weighing up the advantages and disadvantages of both. The graphic above is designed to help categorise what you are creating, thereby providing a better sense of its ideal home.
The most traditional category is recorded music, as it predates the proliferation and miniaturisation of personal computing. In the 20th Century, radio transformed music, and then television transformed it again. In this regard, Spotify and YouTube are quite traditional, as the former imitates radio while the latter mimics TV. This might help explain why Spotify is almost entirely an app, sitting in the background like a radio, and YouTube is most commonly used in the browser, fixing your gaze as if it were a TV. Whether a person is likely to be tethered to a computer or walking around with a phone, may help in deciding between browsers and apps.
Turning to generative music, a successful example of this in the browser is Generative FM, created by Alex Bainter, which hosts more than 50 generative music compositions that you can easily dip into. It is funded by donations, as well as an online course on designing generative systems. The compositions are interesting, varied, and engaging, but as a platform it’s easy to tune out of it. This might be because we are not in the habit of listening to music in the browser without a visual component. The sustainability of this method is also questionable since, despite there still being a good number of daily listeners, the project appears to have been somewhat abandoned, with the last composition having been uploaded in 2021.
Perhaps Generative FM was more suited to an app form, and there are many examples of projects that have chosen this medium. Artists such as Bjork, Brian Eno, and Jean-Michel Jarre have released music as apps. There are obvious benefits to this, such as the fact that an app feels more like a thing than a web page, as well as the commitment that comes from installing an app, especially one you have paid for — in the case of Brian Eno’s generative Reflection app, it comes at the not inconsiderable costs £29.99.
Yet, more than a decade since Bjork released her app Biophilia, the medium is still exceedingly niche and struggling to become established. Bjork has not released any apps since Biophilia, which would have been time-consuming and expensive to create. Despite Bjork’s app not having beckoned in a new digital era for music, this may be a case of a false start rather than a nonstarter. As app building gets easier and more people learn to program, there may be a breakthrough artist who creates a new form of digital music that captures people’s imaginations.
To turn the attention to music-making, and music programming in particular, there is a much clearer migratory pattern. Javascript has allowed programming language to work seamlessly in the browser. In graphical languages, this has led to P5JS superseding Processing. In music programming languages Strudel looks likely to supersede TidalCycles. Of the many ways in which having a programming language in the browser is helpful, one of the greatest is that it allows group workshops to run much more smoothly, removing the tedium and delays caused by faulty software. If you have not yet tried Strudel, it’s worth having a go, as you can get started with music-making in minutes by running and editing some of its patches.
The final category of AI — or large language models — is the hardest to evaluate. Since there is massive investment in this technology, most of the major companies are building their software for both browsers and apps. Given the gold rush mentality, there is a strong incentive to get people to open up a browser and start using the software as quickly as possible. Suno is an example of this, where you can listen to music produced with it instantly. If you sign it only takes a couple of clicks and a prompt to generate a song. However, given the huge running costs of training LLMs, this culture of openness will likely reduce in the coming years, as the companies seek to recuperate their backers’ money.
The question of whether to build something for the browser or an app is not a simple one. As technology offers us increasingly large numbers of possibilities, it becomes more difficult to choose the ideal one. However, the benefit of this huge array of options is that we have the potential to invent new ways of creating and presenting music that may not yet have been imagined, whether that’s in an app or browser.
Please feel free to share your thoughts and insights on creating for the browser or apps in the comment section below!
Dom Aversano is a British-American composer, percussionist, and writer. You can discover more of his work at Liner Notes.
Book Review: Supercollider for the Creative Musician
Dom Aversano
Several years ago a professor of electronic music at a London University advised me not to learn Supercollider as it was ‘too much of a headache’ and it would be better just to learn Max. I nevertheless took a weekend course, but not long after my enthusiasm for the language petered out. I did not have the time to devote to learning and was put off by Supercollider’s patchy documentation. It felt like a programming language for experienced programmers more than an approachable tool for musicians and composers. So instead I learned Pure Data, working with that until I reached a point where my ideas diverged from anything that resembled patching cords, at which point, I knew I needed to give Supercollider a second chance.
A lot had changed in the ensuing years, and not least of all with the emergence of Eli Fieldsteel’s excellent YouTube tutorials. Eli did for SuperCollider what Daniel Shiffman did for Processing/P5JS by making the language accessible and approachable to someone with no previous programming experience. Just read the comments for Eli’s videos and you’ll find glowing praise for their clarity and organisation. This might not come as a complete surprise as he is an associate professor of composition at the University of Illinois. In addition to his teaching abilities, Eli’s sound design and composition skills are right up there. His tutorial example code involves usable sounds, rather than simply abstract archetypes of various synthesis and sampling techniques. When I heard Eli was publishing a book I was excited to experience his teaching practice through a new medium, and curious to know how he would approach this.
The title of the book ‘SuperCollider for the Creative Musician: A Practical Guide’ does not give a great deal away, and is somewhat tautological. The book is divided into three sections: Fundamentals, Creative Techniques, and Large-Scale Projects.
The Fundamentals section is the best-written introduction to the language yet. The language is broken down into its elements and explained with clarity and precision making it perfectly suited for a beginner, or as a refresher for people who might not have used the language in a while. In a sense, this section represents the concise manual Supercollider has always lacked. For programmers with more experience, it might clarify the basics but not represent any real challenge or introduce new ideas.
The second section, Creative Techniques, is more advanced. Familiar topics like synthesis, sampling, and sequencing, are covered, as well as more neglected topics such as GUI design. There are plenty of diagrams, code examples, and helpful tips that anyone would benefit from to improve their sound design and programming skills. The code is clear, readable, and well-crafted, in a manner that encourages a structured and interactive form of learning and makes for a good reference book. At this point, the book could have dissembled into vagueness and structural incoherence, but it holds together sharply.
The final section, Large-Scale Projects, is the most esoteric and advanced. Its focus is project designs that are event-based, state-based, or live-coded. Here Eli steps into a more philosophical and compositional terrain, showcasing the possibilities that coding environments offer, such as non-linear and generative composition. This short and dense section covers the topics well, providing insights into Eli’s idiosyncratic approach to coding and composition.
Overall, it is an excellent book that every Supercollider should own. It is clearer and more focused than The Supercollider Book, which with multiple authours is fascinating, but makes it less suitable for a beginner. Eli’s book makes the language feel friendlier and more approachable. The ideal would be to own both, but given a choice, I would recommend Eli’s as the best standard introduction.
My one criticism — if it is a criticism at all — is that I was hoping for something more personal to the authour’s style and composing practice, whereas this is perhaps closer to a learning guide or highly-sophisticated manual. Given the aforementioned lack of this in the Supercollider community Eli has done the right thing to opt to plug this hole. However, I hope that this represents the first book in a series in which he delves deeper into Supercollider and his unique approach to composition and sound design.
Click here to order a copy of Supercollider for the Creative Musician: A Practical Guide
Dom Aversano is a British-American composer, percussionist, and writer. You can discover more of his work at the Liner Notes.