Prototype experimentation: Session 1
I managed to acquire some metal from Goldsmith’s university, choosing long thick pieces with the aim of having a lower and wider resonant frequency range. We started by grinding the sharp edges off the pieces to make them safe, then we drilled holes as a means to hang them. The heaviest piece was too thick to drill through, so we welded two hooks on.
I chose to use a larger transducer for more volume, the downside of which was the suction cup was too large for the beams, so I had to tape the speakers on. I don’t think this was the most effective method. I’ve found the struggle with this method of sound is balancing the strength of the connection (between the speaker and material) and the damping of the resonance that occurs through attachment. Im considering trying a g clamp next as they are metal and hopefully won’t dampen the sound.
My original plan was to have two beams working in stereo, but after having trouble with syncing the two transducers, and lack of volume, I decided to stick to one beam. I found the material influenced the sound a lot, and the input sound of an E-bow on a metal guitar string suited the timbre. In the interest of time I decided to play a recording I made a few weeks earlier with about 4 harmonies, luckily I had many different versions with some other instrument such as synthesisers, granulated vocal recordings and electric guitar. This gave me a better understanding of how to compose the arrangement, how certain instruments resonated with the material and what worked well when combined.
A drawback I discovered while using this method is that it’s very hard to achieve a decent volume, to solve this I decided to record the sound using a zoom Hn5 held very close to the beam. I found that the the sound drastically changed depending on where I placed the mic, I could change the amount of low end by moving the mic. Towards the end of the session I found this to be the best position, it gave me the most ‘material’ in the sound:
Through this method I could also experiment with swinging the beam to bring a more dynamic spatialisation to the recording.
I realised that the input files had too much reverb applied and the identity of the material was getting lost. Removing all of the reverb resulted in a very empty sound, and was not what I set out for so I gradually added small amounts of reverb until I found a middle ground.
Retrospectively I realised I misunderstood the order of these manipulations. My first mistake was having all 4 harmonies play at the same time through one beam, I feel this loses a lot of the detail I could have harvested from the material had it been layered. My second mistake was applying the reverb first. No reverb sounded empty to me because I was listening to the whole composition through the material, giving me the expectation of a finished piece. I believe the better method would be to record each harmony separately through the material, I can then combine these recordings in the DAW and recreate the digital landscape’s structure. These recordings can then have reverb added as a last measure to contextualise them in the mountainside.
Revised order of process:
- Record harmonies separately with no reverb
- Create the soundscape in the DAW
- Add these recordings in the space with Gain adjustment for distance
- Add a Convolution reverb to contextualise the recordings in a realistic space.
I spent the last portion of the session crudely putting together a soundscape comprised of a few woodland recordings found on BBC library, a chain recording I conducted in the performance lab quickly and a wind track. I then layered a few of the recordings from this experiment to see how they fit together. I believe it is a proof of concept, however the harmonic metal moans need to be separated and contextualised in the digital space.