Some machine learning
Thinking about what to do / learn / experience next… and in following up my sound scape making desires, as described in the sound comparison post, I deliberated to continue with Pure-data now. Ed Kelly showed in his interesting masterclass during the low-residency what Pd is capable off and I know that I have to dig deeper in this one day. But looking into some reviews and comparisons between Pd and Max/MSP, I decided to jump to Kadenze.com and see what was on offer there… as they make learning (just like Lynda) fun and easy.
There I came across a great course of Rebecca Fiebrink, ‘Machine learning for musicians and artists‘. She is lecturer at Goldsmiths University and the inventor of the Wekinator. This is a fine piece of machine learning software that works with OpenSoundControl, which runs OSC-messages between inputs, patches and programs in real-time. By delving in the course in the past days, suddenly I saw a lot of threads and new potential coming together.
© pictures from Kadenze.com
Also the guest lecture of Laetitia Sonami gave fine insight in the potential of mapping gestures as a way of sound making. Together they showed that training accuracy can be a good strategy for me as an artist on how to improve an ‘instrument’. Also I did a bit of reflection on models that change over time, as this is really my ‘cup of tea’. Dynamic Time Warping and the Hidden Markov Model came up. The probabilities of Markov were already mentioned by Ed when he saw my installation ‘The Sound of Indifference‘ during the summer exhibition. Now I also understand why he made this connection. Things fall into place for me.
The advantages of machine learning are now clear to me. In this way you are able to make quick tests and many prototypes in a much more organic way than by making the mapping in code. I am excited and will soon experiment furthermore in this way as it sure supports the live and embodied parts of my analogue work.
Next awaits the Kadenze course about Programming Max: Structuring Interactive Software for Digital Arts. So, Merry Christmas and a Happy New Year!