Granular Patch

This patch was created for future exploration of mapping techniques for controlling granular synthesis. Currently, a combination of play~ and poly~ objects are used to generate the grains and the 2d.wave~ object was used to allow interpolation between different amplitude envelopes. Parameters are also provided for introducing randomness to the grain length and duration between grains. Initially the curve~ object was also used to allow for manipulation of the pitch envelope for the grains. 

The patch could be developed further by improving the enveloping, reducing clicks and providing a better interpolation between envelope shapes. In addition to this, control over the density of grains could also be added using the mute functions of the thispoly~ object.

The scaling for randomness could also be improved by creating a stronger relationship with the incoming audio and parameter values. For example, markov chains derived from incoming audio could be used to provide values for grain length duration between grains. It could also be interesting to create a feedback loop using the granular output as the source for the markov chains. Other ideas to explore include the triggering of grain clouds from an amplitude threshold of an audio signal and controlling pitch distribution using spectral analysis.

3D Printing

Two simple designs were created in an attempt to make the potentiometers on the cello body more accessible. The longer one was found to be very effective, especially when placed parallel to the strings, as it compliments the vertical flow of movements relating to the strings. However, the body of the cello and other potentiometers restrict the maximum range.

The flatter, disc shaped design was created to allow adjustments to be made using the bow. This worked well but the orientation of the potentiometer made it awkward to access and needs to be reconsidered. Additionally, the finite range of the potentiometer was also problematic and could be replaced with an endless encoder. This would allow for the maximum range of the encoder to be adjusted to suit the bow length. 

Future developments will include experimenting with various different sizes shapes and placements.

Modular Experiments

This recording was created to document the exploration of the Max/MSP modules in their current state. During this process I found that a ramp generator was needed to drive simple linear processes. 

Cello Update

After playing the modified cello and acquiring a sense of the types of parameters needed, a more streamlined iteration was created. The areas around the neck were chosen in order to compliment the more traditional playing techniques of the cello. Furthermore, placing the sensors around the neck allowed me to visualise the potential for reducing the overall size of the cello whilst still remaining functional.

Additionally, The current mapping have been organised around the bimanual techniques of cello/guitar playing (left side of body = pitch/timbre, right side = duration/volume). This has been found to be particularly effective as it facilitates the incorporation and development of previously obtained techniques.

Aside from these modifications, the overall shape of the cello still feels slightly unbalanced and uncomfortable, making dynamic and spontaneous navigation of the instrument particularly difficult. However, it is still unclear if this is a problem with the shape of the instrument or my lack of technique and practice, or more realistically, both.

Contact Microphones and Convolution

This recording was produced whilst experimenting with contact microphones and convolution using Ableton Live. The signal from two contact microphones was fed into two separate convolution processes with differing impulse responses. In this way, the experiment attempts to find a more expressive method for manipulating recorded sounds

Initially the contact microphones were used to generate impulses but a wide variety of sounds can also be made using differing excitation methods.

In particular, the coarse surface of the kitchen sponge and the use of a guitar pick across surface were found to be effective.

This technique could be further improved by 

crossfading between the two input signals as both convolution processes are always heard regardless of the position of the initial impulse.

In addition to this,  a more dynamic timbral response from the convolution could be obtained using analysis of the input signal to manipulate parameters of the convolution. As the spectral content of the output sound is determined by the correlation between the initial impulses and impulse responses, experimenting with different surfaces and their combinations could also provide a more dynamic and intuitive control of the spectrum.