As always compression from online video hosting has destroyed the effects of the neural network in the video, but this is more-or-less the final version of the thesis film I have been working on this whole semester. And there are two main goals that this film is meant to get across.
The first, and most important, being that this film was made to show that a neural network can be used to make any and all art. There is almost no instances online of neural networks being used to make anything longer than a GIF. So WOMPANOAG was made with this in mind and the neural network is the main driving force behind the visuals of the film.
The second major goal was to create a representation of a specifically impactful and “bad” LSD trip of a close friend. The distortions created by the neural network resemble the visual distortions seen by this person. The audio becomes more rhythmic as the film progresses becoming more of a score for the film than just the sounds being made by the environment and actor.
I would also go out and say that there is a third goal for this piece and this is something I try to get across with all of my major personal projects; a reaction from the audience. Even if the reactions had are not ones that I had intended the fact that there is reaction and response, whether it be positive, negative, or even neutral, is what is important to me when making art. The feelings and conversations that are brought on by my work are the main driving factors behind me.
Items that are crossed out have been completed
First Check in
Rough draft edit of dream/trip sequence
Base rough edit render
Update calendar based on responses
Finalize and cut down dream/trip sequence edit
Start rendering dream/trip
Acquire needed props
Find positions, apply transitions, and start renders
- Check renders
- Have dream/trip sequenced finished
- Continue editing the rendered sequences
- Finish Audio Recording
- Main sequence re shot and edited
- Sequences are all edited together
- Audio is added
- Film is essentially finalized
- Touch ups and edits
- Film is finalized
There is a pretty major change to the overall idea of this film. Instead of the original dream/tripping sequence being played from the record, where it displayed altered/filtered video sequences that were of similarly structured in the frame with similar camera movements, the video sequences shown in this half will now be of an instance of a separate, but specific, experience of a drug altered trip. This experience is from the same person as the original idea and was explained in what could be remembered and felt during the trip. The video below is a rough, filterless, edit of this new experience for the dream/trip sequence. It’s representative of what the person described their environment to feel like in the moment of the trip. The change was also made because the program used for distortion was not producing the desired effects with 2 very similar images composited together, so this change will also assist the distortions.
This week I worked on some camera tests with the GoPro series of cameras. As I am doing a first person view I’m trying to make sure to set the camera to an appropriate eye-level for the main shot. This first video is set at a level where the camera is on the actor’s forehead.
This angle however is set too high and there would need to be compensation made to put things at a “normal” level. The next video has the camera around the actor’s neck.
The angle on this is set at a more believable level but too close to the actor’s chin as scruffing can be heard. The camera is also too loose. I need to make a new rig to hold the camera at a level more comfortable position.
This is an 18 hour render of one of the distorted segments of the second half of the film. But unfortunately due to web compression this video clip cannot be shown here in it’s highest (watchable) quality. For better results set the quality to the highest that the player allows for.