16 March 02005

E-learning for music technology

Failures are interesting. It's often said that we would learn more if we spoke more about our failures. But no-one really wants to look bad in public so they just publish their 'little' failures. Like this one, which didn't crash and burn, it wasn't aborted, or even stillborn, because it never got 'fertilised' and never got off the ground.

In 02003 I attended two courses on the SuperCollider environment and programming language for real-time audio synthesis. Despite the best efforts of my teachers, Nick Collins, Fredrik Olofsson and Fabrice Mogini, I never got the hang of it. And I was in good company: other musicians on the courses — people I've paid to see perform — either dropped out or said months later that they were still struggling to get a purchase on how to apply SuperCollider in their work. In my case, the lack of both programming background and any grasp of music/synthesis theory, combined with poor discipline in working on SuperCollider outside of the course sessions, did for me pretty comprehensively.

But that's not the failure I was going to write about.

Given the nature of SuperCollider (for example, it is text-based, so exchanging SuperCollider files, or 'patches', doesn't involve large amounts of bandwidth) and the need I perceived for guidance and mentoring over an extended period to gain mastery in its use, I thought it would make an ideal candidate for e-learning. In particular I had in mind an approach that had plenty of tutor support but that could be fairly low-tech (as is my wont). I wrote a proposal (reproduced below) and discussed it with John Eacott, but failed to convince him sufficiently for him to dedicate any of his own time to it (which was absolutely understandable, given John's PhD deadline and impending promotion at the time). I tried it half-heartedly on some other friends and contacts, but my entrepreneurial efforts again fell flat.

There were two factors that led to this failure. The more obvious one was that I needed a partner who understood how to work the accreditation process and the support from funding councils that accreditation can unlock. Without that, I was at a dead end.

More subtle is the somewhat geeky culture that SuperCollider inhabits, with its strong ethos of self-taught mastery, using informal contacts in a community of practice to solve problems, rather than formal learning. The courses I did were the exception to this rule, and the teachers were so imbued with programming and audio synthesis that I'm sure they had never needed to do a course themselves. People who want to learn SuperCollider generally are not the kind of people who care about accreditation and all the bureaucratic disciplines that go with it. Against that backdrop, who needed another course, e-learning or not?

A few months after I wrote this proposal, I read Lucy Green's book How Popular Musicians Learn. They too have many informal learning techniques, but one day someone should do a comparative study of how 'laptop' musicians' practices differ.

Notwithstanding the reasons for failure with this proposal in 02003, I still feel there is the germ of something in it that will happen one day, sooner or later. The proposal has no value sitting on my hard disk, so here it is (minus the costings) in case someone else can learn from it or apply it. Anyone is welcome to copy part or all of this text, though please bear in mind the terms of the creative commons licence under which is published.



What is SuperCollider?
What might be the market for a SuperCollider e-learning course?
Why is SuperCollider well-suited to e-learning?


E-learning programme design




What is SuperCollider?

SuperCollider is a powerful, but potentially quite complex, environment and programming language for real-time audio synthesis. You can use it to write programs to generate or process sound in real-time or non-real-time. SuperCollider (SC) can be controlled by a wide variety of means, and can be used alone or in conjunction with other music software.

SuperCollider has been used in compositions such as Longplayer, the thousand-year musical composition, and by many performers at electronic music events such as the Placard festival

The basic SuperCollider program will run on any Macintosh bought or upgraded in the current millennium, and does not require Midi or other plug-ins.

What might be the market for a SuperCollider e-learning course?

For the last three years the University of Westminster has run one- or two-week Interactive Dance Music (IDM) summer school training courses, focusing on SuperCollider. The first week is at beginner level; the second is intermediate. Costs are around £150 per week or £100 concessionary rate.

These courses have consistently been well subscribed, and have attracted learners based in Scandinavia and Australia. Clearly there is quite a small proportion of people who can commit to a full week's course, pay the fees, and in some cases travel and accommodation expenses as well.

This suggests that there is very likely to be a much larger market for learning about SuperCollider among people who cannot afford the time and money commitments involved in the IDM summer school. This market may well be attracted to a SuperCollider e-learning course, on the grounds of:

  • Potentially lower course fees if technical overheads are lower than for classroom premises (though cutting production costs is not the primary reason for offering an e-learning version of the course);
  • Lower indirect costs (travel etc), bringing the course within the reach of many out-of-London learners for the first time;
  • Ability to schedule learning time flexibly over a number of weeks, rather than having it concentrated in one week;
  • Improved effectiveness of the course for many learners — for reasons covered below.

Why is SuperCollider well-suited to e-learning?

SuperCollider has a number of features that mean that working online could be a very effective means of learning about it.

Table 1: Features of SuperCollider that lend themselves to online learning
Feature Implications for e-learning
SC 'patches' — the file that is 'played' to produce music — are 'plain text' code and therefore very small files. Apart from one-off download of SC itself, there is no need for long and costly downloads of large files.
The only graphics files needed would be screenshots (e.g. of SC Graphical User Interface examples), though others could be used for 'decoration' if this is felt to be useful.
Learners' work can easily be shared for review and critique by peers or tutors.
SC coding is 'cognitively intense' and different learners may take varying lengths of time to carry out similar tasks. The advantage of e-learning in this context is that learners can work in their own time (and not 'hold up' the rest of the class if taking longer than average). Learners can also take a break from mental strain when needed.
Using SC draws on a wide range of skills and knowledge. Learners can follow distinct paths to fill in the gaps in their own background.
Mastery takes time. One learner writes "I still feel like a beginner after 1 yr of SC2!" and another replies "Believe me, that's a common feeling." (From an exchange on the sc-wmin SuperCollider email list) There is scope for ongoing coaching and mentoring, which can be easy to co-ordinate online, though much more expensive by traditional means.

Back to Contents


  • To understand at a basic level the size of the market for a SuperCollider e-learning course.
  • To design and develop a SuperCollider e-learning course for delivery via the Internet.
  • To establish an accreditation and funding framework for the course.
  • To implement and run the course as a pilot, as a means of finding out how to improve the course.
  • To inform a decision about the viability of, and business case for, a full-scale course and associated marketing.

E-learning programme design


E-learning is a different experience from classroom courses, and the design of this course should not be driven solely by 'shovelling' existing materials online and filling in gaps between them. Detailed design of the e-learning course should consider the potential of different kinds of activities that might form part of the learning process.

Table 2: Proposed learning activities
Activity Potential in the learning process
Tutor input There is not much need for tutors to present material as though in a classroom: learners can read materials independently.
However, tutors are likely to be indispensable when supporting learners in individual exercises: coaching them and troubleshooting when their SC patches don't initially do what they want.
Individual exercises There may be a limited role for the simplistic multiple-choice Q&A activities that characterise much knowledge-focused e-learning.
However, by far the most useful exercises are likely to be those where learners are set the goal of producing an SC patch to put into practice what they have learnt.
Group exercises There is scope for many different kinds of group exercise, ranging from learners pairing off (or working in small sets) to review and compare each other's work, through to full group compositions.
These have value in varying the learning methods that learners experience, and thus keeping their interest, as well as by forcing them to engage with the alternative approaches to SC adopted by their peers.
Group exercises can be logistically challenging, because they require learners to be at roughly similar stages in the course and in their own development. The technicalities of group compositions may also be complex. Nevertheless, where feasible, they offer significant advantages.
Group discussion Group discussion is often an important 'glue' for holding an e-learning course together. It can range from 'coffee room' discussions (e.g. sharing interests in music created by other artists using SC, news of gigs etc) to what some educationalists call 'vicarious learning' — learning by observing the questions that others are asking and imputing where this might lead.
Because SC is such a flexible and powerful environment, people can use it in radically different ways. Some learners may risk 'tunnel vision' by getting into a habitual way of using SC early on, so there is great value in giving them the opportunity to see — and hear — work created by others who may have significantly different approaches.
Tutors can also post messages about what they observe as common misunderstandings in the group.

Some individual exercises and, to a lesser extent, group exercises may be used to generate assessment evidence for accreditation purposes (see below).


This proposal is based on negotiating rights to use selections from the tutorial materials from the Interactive Dance Music summer school courses held at University of Westminster (these are available for download).

These materials will need to be adapted in a number of ways:

  • Review the overall structure of the course to check it still stands up for more 'flexible' delivery online;
  • Add narrative text to provide a context for what are currently very 'code-heavy' materials;
  • Add graphics — especially screenshots of SC Graphical User Interfaces (GUIs) — for elements that would normally be demonstrated in a classroom course;
  • Revise and restructure some of the existing exercises to ensure they work well online;
  • Add elements to lead learners through the additional parts of the process, as described above, as well as general 'Introduction to learning online' sections etc.
  • Some original material may have to be authored to ensure that the course hangs together as a coherent whole.


The Rising Tide Trust (Hackney) SC course has obtained accreditation from London Open College Network, so it should be possible to develop a framework for accrediting the e-learning course.

Assessment is likely to be based around a learner's portfolio of evidence: a set of SC patches produced by the learner, plus comments and reviews sent to the tutor and to group discussion forums.

Back to Contents


The main elements required to run a pilot e-learning course as described above are:

  • Learning materials (text and graphics) authored in HTML templates;
  • Relevant parts of SC code available for playing locally (either by copy-and-pasting or downloading as an SC file);
  • Group discussion platform (web conferencing and/or email discussion list and/or, to support group projects, multi-author blog);
  • Some form of learning management system (LMS) to control access to the above (some kind of password/login) and enable tutors to track progress of learners.

The implementation of these elements could take a number of forms. In one scenario, if, say, a university already has a bespoke Learning Management System for delivering e-learning, the pilot could be implemented using this system, existing templates and existing discussion systems.

At the other extreme, it would be possible to rent space on an independent server, devise a template from scratch and put the pages on the server in this template, with simple password access (using .htaccess or similar). Group discussion facilities can be purchased (or obtained free) from sites such as Yahoo Groups. Although not technically sophisticated, examples such as the Learning to Teach Online e-learning course, which has won a National Training Award, show that this approach need not detract from the effectiveness of the learning.

The roles required to implement this pilot will be:

  • SuperCollider specialist/educationalist — to adapt existing materials, originate new ones, and design exercises/discussion areas — 10-15 days
  • Technical author/coder — to convert the materials into web pages using templates etc — 5 days
  • Tutor — to support, coach, mentor learners as they do the course, and to assess their work for feedback and accreditation purposes — depends on scale of pilot course and on pedagogical approach, but on one model each learner might require 2 hours support for a 30 hour course, so 30 learners on a 30-hour course would require 8 days from a tutor
  • Co-ordinator for accreditation/moderation — to co-ordinate these activities in liaison with tutors and accreditation bodies — 2-3 days
  • Technical support — to ensure servers, logins, discussion facilities etc are configured correctly and keep working during the course — 5 days
  • Project manager/producer — to co-ordinate the implementation and the other roles, and fill in 'gaps' where necessary — 10-15 days

Some of these roles could be combined: for example, the same person could carry out the educationalist and the accreditation roles, or technical roles could be combined.

Please note all day allocations are first-pass 'guesstimates' only and subject to revision (experience suggests revisions often add 50% or so to first-pass, even if the first pass takes account of this experience, as I have tried to!).

The course should be designed to conform with recent standards in e-learning, such as BS 8426 A code of practice for e-support in electronic learning systems. However, it is probably unnecessary to spend time on technical details of interoperability and metadata for a small scale pilot: these can be addressed if and when the course is scaled up for wider use.

Marketing for the pilot should not need to be extensive, since the experience of the IDM course is that a reasonable number of learners can be found by low/no-cost methods, based on email and word-of-mouth publicity.

Originally written 19 September 02003

Posted by David Jennings in section(s) E-learning, Ideas and Essays on 16 March 02005 | TrackBack
Post a comment

Remember personal info?