Final Assessment video script 18/6/23
Figure 1: Camperdown Country Park entrance: https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.parkrun.org.uk%2Fcamperdown%2F&psig=AOvVaw30wW958qI8u76Df-37ZE25&ust=1687193597388000&source=images&cd=vfe&ved=0CBAQjRxqFwoTCOjptYqkzf8CFQAAAAAdAAAAABAE
My name is George Salisbury and this is my final assessment video for my Animation/VFX degree.
Without any other external projects going on, aside from the commencing job hunt, the main priority for this semester has been to finish the final project and ensure that it's in the best possible state by the final submission deadline. This has allowed me the opportunity to refine the compositing of each and everyone of the shots, and also to refine the flickering effect on the sketch filter sequences.
The issue that came up with the previous versions of the shots were that the effect on the whole frame resulted in eye strain and a confused message conveyance, especially when viewed on a big screen. I found a remedy for this problem by utilizing the rotobrush feature, and alpha matte functions. This was accomplished by making a duplicate of the original footage without the flickering or the pencil sketch filter footage from Photoshop, and applying the rotobrush over the figure in each of the shots. The pencil sketch footage with the flickering effect was then sandwiched between the two duplicates and the track matte was set to alpha creating the effect where the flickering pencil footage was only visible on the main protagonist. The main downfall of this is that the rotobrush can be temperamental and result in areas where the figure isn't fully covered, and results in a lot of trial and error. The other caveat with this feature was the increased rendering time: the flickering footage and normal footage combinations have never been easy to render due to their file size, but often times flickering shots could take over an hour to render. The normal animation composition shots only took about two minutes to render in comparison. The main changes I applied to the shots for continuity sake including a gradually dimming effect on the flickering figure, indicating his improving connection as he travels further through the narrative, and a temperature colour correction to further indicate his connection. It will start off as a cold blue temperature hue and then progress into a warmer, more welcoming orange. If I had to choose an inspiration for these sequences, it would primarily be the glitching effect from Spider-Man: Into the Spider-Verse, where you can immediately tell that the character is not in their original universe and their body is reacting violently as a result. Whilst my effect is not as extreme as the effect in the films, it does draw some influence from the indicated environmental feedback. Initially, the flickering effect was accomplished using a wiggle expression on the opacity, but I switched to a keyframe effect that could be more easily controlled. If I could back on try these effect shots again I would see if I could try the luma matte track matte option on the shots to see how the effect would look and what it would convey.
I have to give credit to Blue Lightning TV for the main pencil sketch tutorial that helped me produce the filtered footage. It was an extremely easy and clear tutorial to follow and helped me alter the footage very quickly. Whilst the videos were very hard to render in Photoshop, his clarity and structure definitely eased this part of post-production.
It was during this term that I first tried out the DJI Osmo Pocket, having been introduced to it earlier in the year. I have to say from this experience, I have fallen in love with the camera. The image quality is great for the camera, the device is very easy to connect to your phone with the included app, and the shots that I've got out of it are smooth and organic looking, and work great with the 3D camera tracker inside After Effects. Whilst I doesn't have the most depth in its shots, it is the camera I will to use more outside of college.
The shot in the film that I believe I've spent the most time on of all the shots is the deer shot at Camperdown. All told I've been to this location at least seven times in the pursuit of a shot that was spot on. The first attempts at this shot involved the camera sitting on a tripod with no movement and the deer moving across masks on top of the footage. The shots were fairly effective but the lack of camera movement on them made them boring to look at and the lack of trees to project the animation on also hindered the shots. I tried filming again with more trees in frame, as well as using a CC cylinder effect on the animation in addition to the masks to make the deer appear as if they're walking around the trees instead of just on top of them. The issue of the static shot still persisted, so I then decided to try a panning shot, which resulted in the animation flailing around and not synchronizing with the footage. Eventually I tried using the DJI Osmo Pocket for the shot which had brought me great success with the refined boar sequence filmed at the Tay riverside. The shot produced a smooth panning sequence with a slight hint of shake from me holding the device, and whilst it wasn't as perfect as dolly track, it was the best I could ask for for quick, efficient filming. The effect I aimed for was a smooth panning effect through the trees with the deer passing by as I moved forward, as well as having a set of double discs visible at one point. The attachment of the deer animation to the trees was achieved by using the 3D camera tracker which created a null object in the 3D space of the shot that allowed the animations to be placed and moved accurately. The effect was well received even though the lack of sufficient trees was a bit of a hinderance. If I were to try this sequence again, I would have to find a location with the most trees in alignment so that I could have the effect appear more often. I also had to change the frequency of the animation to create a herd-like effect that mimicked how deer migrate.
Another shot that took a few attempts to master was the Cox's Stack sequence with the animation of the jute growing up the brickwork. The initial attempts at this sequence were static and unengaging, with nothing else happening in real-life other than a brief glimpse of a plane flying by. The second attempt at the shot involved a "timelapse" shot with footage of the tower filmed over one hour and sped up 10000% in Premiere Pro. Whilst this shot did have more engaging features with the cloud movement, it was still a static shot with no real engagement. I went back again to film another timelapse shot but then decided to try a range of panning shots around the tower, and settled on a shot where I walked towards the brickwork, which gave a more accurate sense of how tall it really was, in addition to a double discs Easter Egg. This was the shot I was the most happy with, despite a comment that the brickwork could be removed but this was nothing major.
If I were to narrow the film's appeal down to a specific target audience, I would choose those with discerning tastes and film students aged between 18 and 35. I could definitely picture the film playing in the experimental film selection at a short film festival, such as Aesthetica or at Oberhausen. I hope to submit the film to Aesthetica once the degree is over.
The process of editing the film has been a lengthy procedure. I will have produced at least seven cuts of the film before I submit the final draft. The very first draft that I put together in February without visual effects or enough animation was a complete fiasco and I'd rather not discuss it in detail. As I moved to producing further versions, I managed to cut down the footage to what was really necessary and I have gradually improved the narrative structure. The first cut was approximately 12 minutes in length, and the current version has been reduced by 25% to a more compact length under 9 minutes. I had to get rid of the glasses sequence as it really contributed nothing to the narrative, and it was suggested that I change the order of the sequences so that it started out more intimidating but eventually progressed to something more tranquil. The downside is that this would involve changing the colour correction and the narrative wouldn't make geographical sense to the more educated viewers. The temperature colour correction process was tricky, in that the overall colours of the flowing shots had to somehow match but the structure of the compositions meant that the footage had to corrected in After Effects which meant more time reserved for rendering. It was a tough process but I made it through in the end. The final film is something I can definitely be more proud of than the final film for my BA degree.
These last ten months since starting the degree have definitely been a transformative experience, and I am very surprised I've managed to complete the work as much as I have given my learning previously. Whilst there have inevitably been some down times, namely at the start of the degree when I was trying to settle in, I have managed to pull through these rough periods and produce some of my best work. My final film for this degree is a much better connotation of my aspirations and my perspectives and I look forward to continuing to learn more about the animation practice and what it can provide creatively. My journey after this will involve me moving to Edinburgh in the pursuit of work in editing/animation, whatever that might involve. Whatever job I get, I will commit to and make the most out of.
So with that said, mar sin leat University of Dundee. I hope I did you proud.
