While theatre is a social art form, music doesn’t have to be, and so over that semester (as well as before), I was creating music any chance I got. One of those products, a rendition of “Rainy Days and Mondays” by the Carpenters in the style of Animal Crossing’s KK Slider, has become the subject for my experiment. With it, I wanted to explore the visceral experience of a song who’s tempo and arrangement borders on melancholy as it morphs into ambient desolation, and then returns seemingly unaware of its own metamorphosis

One of my favorite audio manipulating tools in the world is Audacity’s “Change Speed”, in which both tempo and pitch are affected by a multiplier ranging from 0.010 to 5. The ramifications of that range are immense: ten seconds of audio multiplied by .1 can become a minute and forty seconds of continuous sound, while also illuminating overtones and frequencies that the human ear would have taken for granted from the sample.

With this piece, I worked backwards, taking ten seconds from the end of the track and decrementing the speed by .1. I continued to do this for every ten seconds preceding it. Once reaching 0.1 itself, I reversed the process and incremented by .1 up to 1.2, returning back down so that the track begins at the initial pitch it was created in and would end up at. The result is a piece nearing 9 minutes in length, which devolves into a disorienting ambient wasteland during an acoustic guitar solo in the mid-section. There’s a decent number of glitching-out sounds during the switches in tempo: this is due to the fact that the 10-second chunks were all mapped out by hand, and so in human error, there are overlaps of the speed being devolved. I rather like it, though, as I feel like it only intensifies the personal connection.

I used a photo I took of a dog. For each iteration, I increased the saturation by 25. I continued to do this until there was no noticeable difference when I continued to increase the saturation, which occurred around iteration 50.

This project exploits the lo-fi sampling capability of the 1987 toy keyboard, the Casio SK-5. It maximizes the SK-5’s processing power by using all three sample modifications available, sample reversal, pseudo-reverb, and sample looping. Beginning with an arbitrary voice sample, we record the dry, unaffected sample onto a computer, then apply the three mentioned modifications, then record on a computer once more. 

This modified sample is then played from the computer into the SK-5’s microphone input to resample as the next recursive level. Each subsequent level of resampling thus reverses a previously reversed sample, applies another layer of reverb, and loops the sample at a slightly different time, creating both rhythmic and harmonic artifacts.

the chain, more simply:

recording of unknown voice with sk-5’s built-in mic  ->

{ record onto computer without modifications            ->

internal SK-5 processing: reverse,  ‘reverb,’  loop       ->

record onto computer with above effects                 ->

 resample affected sound from computer onto SK-5}            

repeat  {}’s 20 times

When I was sitting down to do this project, I really wanted to find a way to combine both pre-recorded and live sound. The technology I used was my phone, using voice memos, and my laptop, using audacity, to record the whole process. I first recorded my voice as almost an intro to this recording. I then found the audio of the beep boop on the official Carnegie Mellon University Instagram page as its memorial to the change of our favorite crosswalk sound. I used my phone’s voice memos application to record my process as I used my computer’s speaker and microphone to continuously record onto audacity using the acoustics of my bedroom in my apartment. I am really interested in the relationship between human voice and technology so being able to hear the difference between the two in dramatic distortion was really exciting for me. 

It took me a while to come upon this process as my knowledge of Audacity is very limited. I found it to be really cool though to use the technology I have accessible to me of my laptop, social media and my phone in order to create this project.

Recently, a friend of mine introduced me to an app called Huji Cam – a retro disposable camera photo app that adds red-hue filters to your photos as soon as you take a picture. These filters are completely random and you cannot choose or change them.

As an example to demonstrate what Huji does, I took the following 4 photos of my apartment keys. For comparison, I used my normal phone camera for the first one (left) and I took the other 3 using the Huji app. You can see that each of the photos have a different filter:

As I was playing around on the app and taking random photos, I got the idea to layer pictures together to see if I could come up with some kind of abstract art. I walked into the nearest hallway, and, standing at one end, I took a photo using Huji, took a step forward, took another photo, and repeated the process until I reached the end of the hallway.

I then tested a few free photo layering apps that I found on the app store and came across Photoblend.

This was the final result:

For your reference, here are some iterations showing the evolution of the photo: