I have created a music video for my song “TRU LUV”. I wanted to create an audiovisual experience that had a present in the past vibe. It's very much a collage of experimental techniques in the world of digital media. I intended the video to be an extension of the song, with every aspect of the video synced with the rhythm of the music.
Success would be to challenge the idea that all media must have meaning. The goal would be to remove meaning and be left with raw experience and “feeling”. A lot of what determined the success of a house track, and the house genre was if everyone in the room could feel it.
Achieving success would also be the application of the various techniques I have researched. I wanted to test the limits of the hardware and software I had access to. I did some research on topics like Databending, Datamoshing, the Eclipse of Titan optical illusion, and the 3d software, Blender.
My goals were to create an audiovisual entrancing experience. I wanted to experiment with what kind of visual experience I could create with the tools I had. I had access to both windows and mac, so a large variety of software was available. I used Final Cut the mac for editing the video (since the Mac was far more powerful). I used windows for the smaller VFX parts: Avidemux for Datamoshing/Databending, audacity for Databending and Natron to create the illusion.
In the future, I would like to distribute the song and video. I will continue to promote it on my personal website and YouTube channel.
I had researched various experimental techniques to include in the video. Below I will go more in depth on each. I spent most of the time in preproduction learning the software that I would need to use for the effects. I had planned out the structure of the video in one of the diary entries, but I started drifting away from it when the deadline was coming closer.
The Eclipse of Titan illusion is a variation of the Eclipse of Mars illusion. It's a red circle with a white dot in the centre. After one to two minutes of staring at the white dot, a circle with the opposite light frequency is left imprinted on your vision. The circle is also brighter than what a screen can produce, so I also wanted to push past limits of what the hardware was capable of.
I had made it in a free and open source VFX program called Natron. The program uses nodes to generate shapes and add effects to the shapes. I had used ramp to make the gradient, the read node to import the radial lines and sparkles, the radial node to generate the circle, etc.
I had made it for the build-u, but I used it through-out the video. Since the build-up is slow, it would give the illusion enough time to work.
The illusion had worked in the video; but since the blue imprint was over a yellow sunset sky, the brightness of it dulled. If I were to remake this, I'd use a different version that showed a warm coloured circle since it was supposed to evoke the sun.
I wanted to create an ambigram logo since each word in “TRU LUV” was three letters long. Ambigrams usually work by turning it upside-down, but I wanted to make it to be a reflected ambigram, so it could be used in the reflection of the water.
I had made it in blender with the help of some YouTube tutorials. I had used most of the same colours as the 2d logo. It originally didn't have a border, but it had trouble contrasting against the background, so I added one. The 2d logo does also have a border, but it's very thin and not visible in the video itself. The 3d version also has some 4-pointed stars floating in front to add some extra depth.
I had used both variations in the video in roughly the same way. the top part sat on top of the water and then a reflection sat underneath with an underwater effect on it. The readability of the “LUV” part is questionable since the “R” upside-down resembles a “G”, I couldve tried to make another pass of the design but as is, it does spell out “TRU LUV”.
Most of the video is built from stock footage from Storyblocks. I had downloaded lots of landscape footage of sunset beaches for two reasons:
I also used some night footage off the sea to match the filtered off, more closed off parts of the song. Every clip was colour graded to look bright and summery. A lot of the clips are only shown for a split second.
Databending is an effect that alters the colours and edges in a video file. It is created by corrupting the data in some way. I had experimented with Databending the past in my own personal projects.
The method I used was with Avidemux. Avidemux is a video encoding software, used to convert the video into a raw data file. This is then imported into Audacity and then an effect is applied to the data. For this project, I used Amplify +0.5bd. The data is then exported into a raw, then converted back into a video using Avidemux.
Databending is used the video though it is very subtle. It's on the second iteration the illusion. You can see some it in the edges of the radial lines.
Datamoshing is a technique that creates a glitchy transition from one scene to another, like that of a broken television. Datamoshing only requires Avidemux. I converted the video to the Mpeg4 ASP (xvid4) encoder. In the video configure window, disabled all B frames and set the Gop size to maximum (500).
It took a couple of attempts to get a successful Datamosh since a lot of the tutorials and documentation were outdated. The I frame interval setting had since been renamed to Gop size in the latest version of Avidemux.
In the end, some of the Datamoshed segments seamed too clean and not as glitchy as Id hoped. If I were to do this again, I would add some fake camera shake to make the transition less linear.
I took my main feedback from Lee; he's been in the house scene since the 90s. He had positive things to say. He had a criticism of the transition between the sea at night to the whirlpool, saying it could've used a splashing effect. This criticism could extend to the entire project; some more time could've been spent on the transitions.
Half of the people who watched it had trouble experiencing the illusion. It only worked when looking at the white dot of the centre off the red circle. I could've made the white dot more contrasting or everything else around it less contrasting to make it more effective.
Overall, I'm very happy with how it turned out. I felt it was successful because I had managed to use all the techniques to create an audiovisual entrancing experience. Some parts were rushed, likely more rushed than they needed to be but the final product meats my goals.