IxD and Assistive Technology at UNSW Art & Design

Digital Video 2 – Week 08

Digital Video 2 – Week 08

Process

The ‘rostrum camera’ project was the most testing for me in Digital Video 2. If you’re not familiar with the concept of a rostrum camera, it is used often in documentaries or other films where there may not be access to moving image – a documentary about the First Fleet landing in Australia in 1788 would be a good example. To add some dynamism to the images, the camera will pan or zoom across the surface of the image. One of the best known examples of this – and one which was offered to us in class – is Chris Marker’s excellent La Jetée from 1962. This film was the inspiration for Terry Gilliam’s (also brilliant) 12 Monkeys

%CODE1%

For the most part, I really enjoyed working with After Effects this semester, but as I’m a novice with the moving image, there are more than a few caveats which I stumbled across in my enthusiasm. Being the trouble child in class, I didn’t want to go down the route of using only 2-dimensional images. After Effects actually has some excellent tools for working in 2.5D (the difference between 2.5D and 3D is generally fairly academic here, but if you want to know more, Wikipedia is your friend), so I wanted to put them to use, whilst still exploiting the visual effect of flat surfaces.

Lately, there have been a few commercials (none of which I can find online) that have used 2-dimensional images within a 2.5D space. Much like a theatre set, these elements shift around the space and really play with the idea of flat surfaces working together to make a something greater than their parts. Aside from being an interesting process, this idea appealed to me, because I would be able to take images from different locations to invent my 1964 setting, rather than needing to find the perfect location.

Much like the audio samples I appropriated from Freesound for my accompanying soundtrack, I took quite a few images from talented users at stock.xchng:

These images were mixed with others I photographed myself, as well as combined with simple artwork in Illustrator, to make the final composition.

The list of things that went wrong with this project in the final couple of days is what I’ll actually take away as being the most helpful part. Whilst I did get extremely frustrated with the project, in retrospect I learnt some excellent lessons which are worth passing on.

Resize images appropriate to output. Who knows how I got it in my head to keep everything at full resolution and just drop the scale for each. Doing this with one image is bad enough, but my method was to use around a dozen copies of each element to create depth in my ‘set pieces’. When you add ten different elements together for a scene, suddenly there are 200 layers – each of which are a larger file size than necessary. Add some animation to the mix and even using After Effects for the simplest tasks is a nightmare.

Lower viewing resolution. What you see isn’t what you get in the world of motion graphics – it’s simply a reference. Not being used to this, I didn’t even think to drop the preview resolution from Full to Half, as I do now, even for basic projects. Your CPU will love you for it.

Changing After Effects' preview resolution to Half will help save your sanity

ProRes 422 is a great intermediate format. I needed to break down my project into ‘scenes’, just so it became more manageable. I made the mistake of exporting each scene at the highest resolution I could muster, before pasting them all together to export as a complete video. Bigger file = better quality, right? Perhaps, but as with my oversized images, there comes a point where it’s not really making a visible difference for a relatively low-end project like this one. Since finishing this piece, I’ve found that ProRes 422 seems to be a good trade off between keeping pretty good picture quality, whilst not going absolutely overboard with file size. It also plays nicely with most software, like After Effects and Final Cut.

The render time almost killed this project for me in the end. When I first began the render the night before, each 10 seconds of footage was taking around 30 minutes to output. Being a 3 minute pieces, this was never going to work. I had to scramble to cut down file size, just so I could get something ready. It cost me an entire night of sleep and ultimately, I slept through my presentation. It was the combination of my above mistakes which led to my less than impressive delivery.

Outcome

Clearly, I’m not totally satisfied with this project. However, I did learn some important things from my mistakes and in keeping with the spirit of my Uni Journal, I thought it best to present it, warts and all. There are moments that work quite well with the soundtrack, but I’d certainly never approach another project like this in the same way again.

%CODE2%



Leave a Reply