I See You: Composition for a short video

I had a really great chance to compose a short piece of music to accompany a short fashion concept video. Here’s a quick write-up on how I went about putting it together to a deadline!

Due to the amount of coursework I’ve had on the go (more on that in the next few posts), I had to crack this out quicker than I would have liked, and I had to streamline my approach to the composition based on what worked in the past. Once I worked out I’ve got a certain method to composing and mixing things (and embraced it), these projects got much more efficient and streamlined.


This project was started from talking to a friend about their upcoming fashion exhibition, and them showing their ideas at that stage. Her ideas were heavily centred on the ideas in the George Orwell book 1984, and the idea of a futuristic surveillance state shaping the clothing of that kind of word. As a fan of the book, and dystopian sci-fi in general my curiosity was piqued, and after a lengthy chat the idea of me contributing music to a promo video came up.

I saw her designs as very reminiscent of the vault jump suits in the Fallout game universe, and the kind of thing that is practical for labour tasks. My mind also went to the world created in the Philip K Dick novel “The Penultimate Truth,” which shares the idea of underground societies, but in this case with the idea of people being kept in “ant tanks.” With this in mind, the music felt like it needed to deal with this sensation of being monitored constantly, and also with this feeling of being lied to as to the reason why.

Spotting and Planning Sessions

The next step from this was to sit with the video and watch it extensively, making notes of key hit-points in the video, and things I wanted to emphasise in the visuals. I then had a Skype session with Natasha to compare notes, and find out what she was expecting from the music, and if my impression of the video was the same as what she intended.

I made a few changes to the plan here and there in line with what we discussed. I initially wanted to use a lot more glitch artefacts in the music, and make it a lot more music concréte, using the extra sound effects as a rhythmic feature. Natasha didn’t want this to be such an obvious element of the music, so I cut that idea at this stage.

I can’t emphasise how much time it saved to just sit and plan the thing out with a pen and paper before committing any ideas to sound. I hadn’t opened any software, and yet had a full skeleton of the piece, combined with feedback from the person whose opinion mattered the most with this project. It’s so easy to get wrapped up in an idea, present a draft, and then be at a loss when the other person requests major changes, as you’ve committed so much to an idea.

It also helps with short videos, as you can’t always fit in all of the great ideas you have into 1 minute, and it’s much easier to have done the trimming before you bloat the video, or worse have no real direction to begin with. 

Making the move to a DAW

Moving these ideas and sketches into my preferred software (Ableton Live in this case) was the next step. This meant importing the video in, and just spending some time setting each marker into the timeline. This is incredibly similar to the method I adopted when working with the Bowhead videos, and again I added in the tracks I thought I’d need, and some automation on the track volumes to jog my memory on what the piece’s contour would be during parts of the video.

Sound effects to set the structure of the piece

At this stage, I also added in a couple of semi-diegetic sounds that I felt would add to the ideas behind the clothes. These were the camera shutter, phone vibrating, and keyboard typing sounds, which I sourced from freesound.org using a Creative Commons 0 licence. With this project, I felt that implying these sounds were in the video would add drama, as they are recognisable, and without a clear source give the videos that sense that the model is being monitored. The phone vibrating is an exception, although I took the liberty of adding it for effect, and it starts before the phone is on screen to try and knit this idea that it’s been ringing for a while, and is a consistent thing for the model’s life.

 Now it’s composition time!

Once this was in place, it made the actual composing work so much easier, as I knew what I wanted before I played it, and everything felt much more deliberate. Sometimes having a blank canvas can be tough, so just having the markers, tracks, and real-world sound in made such a difference, and I was thinking “I need a big crescendo here at 2:48 using a synthetic sound” rather than “what on earth am I doing for this whole piece?”

 I had the luxury of having the Absynth library available, so I used that for most of the synth performance. I used a mixture of “self built” and tweaked presets to get the sounds I wanted, although having a synth patch named after Philip K Dick’s novel “A Scanner Darkly” was just too coincidental to not use!

It was tough to start with getting the bass right, as I knew I wanted a descending set of notes to give this repetitive but unsettling effect, but I also didn’t want to lull the audience to sleep. It started making sense when I added a modelled acoustic bass guitar underneath the envelope filtered bass, and took out a lot of notes in the initially clustered bass line.

There was going to be a drum part throughout the final section, but sounded much to obvious and kind of didn’t allow the other instruments to have their full impact. I settled for using them to create a glitch sound by triggering a drum hit a lot of times in a row, but there wasn’t much need for them beyond that.

I used a few tricks to make the sound jump out a bit more, such as using a high-pass filter on the master channel to make a sweeping effect, which then creates a drop-type sound when the sound is back to normal. I also made sure to manage the audio effects and volume for each channel so that they were changing during the piece, and also the tails of reverbs and sounds weren’t ever suddenly cut off unless I wanted them to be.

Refining and presenting the music

After a quick mix, I was happy enough with what I’d put together, and I sent it to Natasha for some feedback. To my surprise, I got the all-clear for version 1.0 being the only version needed! I had admittedly broken my own rule, and spent quite a while on this draft, but that was largely because Natasha had been so involved with the planning that I had practically written the music. I’m pretty satisfied that for about 6 hours of work, it was fairly well executed as an idea, Natasha was happy with it, and I’ll chalk this one up as a project that went well!


I hope this write-up has an idea or two for getting started on a music project, and I would love to hear anyone else’s approach to video composition if anyone would like to leave a comment! Thanks for reading. 🙂

Leave a Reply