The music at my gym sucks

My gym Fitness4Less is great, but the music there sucks.

I generally listen to audiobooks while working out. Currently I’m on the Iain M Banks culture series. However I wondered today if I could transform the crap music they play there into something more interesting..

So I made a quick RjDj scene which transforms the incoming audio in realtime into something more interesting. Here is a demo recording of it in action with some standard aggressively produced pop feeding into it :)

It features a crossover around 120Hz, below that the signal is passed through lowpass filtered and amplified with some tanh type soft clipping. Above 120Hz all of the signal is fed through a graindelay I made which has a long buffer like 20 seconds and randomly plays grains from that buffer in a stereo spread. The output of this graindelay is also fed through a platereverb which feeds back into the graindelay. The output of that delay / reverb is then sweep bandpassed with randomised frequency centers. The scene uses many of the RjLib Pure data abstractions :

Next time I’m at the gym I’m going to test it out. Wish me luck!


Turning Colour into Music

This was a short experiment to try to control a piece of music by R G B values from a webcam.

In this case, as in a few of my other Junto projects, I wasn’t really concentrating on making a good piece of music, rather I was interested in a short exploration of an idea.

It uses Pd extended, with Gem to extract the mean RGB value of the camera image using the pix_mean_color object. I did some live adjustment of the image based on these values using pix_refraction and pix_motionblur. It also uses some abstractions from the RjLib by RjDj.

The values also control the volume and tremolo speed of 4 sets of 3 oscillators, all derived from multiples of the frequencies 600, 540 and 450. These numbers were from the original brief of the project. I actually tried not to think about them much and used the most obvious translation of meaning – freq. Although they are interesting frequencies.

A higher red value makes the 600hz based frequencies louder , higher green value makes the 540hz based frequencies louder, higher blue value makes the 450hz based frequencies louder. The higher each value, the faster the tremolo on each. This creates some pretty interesting phasing patterns.

I also ( but not in this video ) tried the tremolo in the ring modulation register, and also right up into FM synthesis registers, which was pretty interesting too.

At one point here I tried hooking it up to some rudimentary more conventional musical structures just to see what kind of stuff could be possible. Beat structures and basslines with filters etc. Its ok, but for that kind of work a Kinect or physical controllers is much more practical.

This is part of the Disquiet Junto series of projects curated by Marc Weidenbaum.
More details on the Disquiet Junto at:

disquiet junto 0009 cross species collaboration

I managed to get some time to take part in the junto series of projects again. This time it was about creating music from a set of samples.

This one was really fun to do. I made it in Pure data again. This mp3 is a recording I made “performing” the patch live with my apple keyboard.

This is a brief summary of how it works :

1. The patch plays the original freesound files of the bird and the guitar into analysis objects ( fiddle~ )
2. This determines when the bird or the guitarist plays a phrase
3. Each time this happens the phrase is recorded into a set of 6 phrases. When they are full, they record over each other again.
4. An algorithmic pattern generator decides how the bird phrases and the guitar phrases will be played back in time with a musical tempo.
5. The drums are formed from two samples of the guitar scrapes / slides and messed with. They are played back by an algorithm, I just turned them on and off.
6. The ‘bass’ is formed from 3 samples of the guitar played at the wrong samplerate and messed with various filters and tremolo. I just turned the bass on and off.
7. The whole thing is also going through a set of secondary effects – repeaters, bitcrushers, reverbs, reversers, etc.

Hope you like it! Listen to that bird go!