ANTHUS PC — Week 9— Term 3

R K Lewis
3 min readJun 11, 2020

This week primarily revolved around re-creating scene 3 so that it placed less pressure on the graphics card because the builds would only last 5–10s before it would crash. This was mainly because the L-system trees were each composed of several hundreds of individual parts. So while Žan was encouraging me to go back to the L-system code and tweak it until the trees look good without so may parts, as I had my heart set on keeping the ones I’d already made I simply set about de-bulking the existing tree prefabs and populating a brand new scene with far fewer (i.e. 5–10 trees, not 50–100!). This involved selecting each cube in turn, working out whether it was actually visible from the surface and if it was deep inside the tree, and thereby occluded from view anyway, I would delete it. When this process still didn’t bring the total number of parts lower than 100 pieces I would delete random cubes and make an aesthetic judgement as to whether the overall appearance was improved of worsened. I found this reductive process very enjoyable and found myself pruning back certain sections of the tree in order to create eyries at different heights up the length of the tree.

Another change I made in scene 3 was to make one tree right in the centre of the space considerably taller than all the rest in order that i could put some eagles right at the very top. The inspiration for this was a friend telling me that there is a very healthy population of peregrine falcons in the City of London that use the many skyscrapers as their eyries, diving down from a great height at great speeds to snatch pigeons out of the air before they knew what hit them. My eagles will be doing a similar thing when I come to start work on my Final Major Project, but for the time being they are perched there so that should anybody look up to the top of this tree, they will just about to be able to make out their tell-tale silhouettes.

Another big task for this week was getting the script right for auto-piloting birds around the scene. I found it relatively easy to get them moving from one place to another, but triggering the appropriate animations at the appropriate times for take off, flying, landing and idle again proved very tricky. I got there in the end. It seems that each bird needs their own animation controller because the parameters seemed to be changed in all animation controllers rather than just the one attached to the target bird. I now have four sparrows migrating from tree to tree and each tree has 3 possible perches they might use. Next up I need to do the same for the pigeons and the crows. I think I’m going to leave the gulls, ducks and owls to be introduced further down the line when I make the full VR version of this game for my Final Major Project. For the time being I just want people to have plenty to look at as they move around the space.

I also had some fun using blender to make a landscape on the basis of a black and white picture of the surface of the moon. It didn’t come out looking much like the pockmarked craters, but I did get the basic undulations I was looking for creating gentle peaks and valleys to divide up the scene in a naturalistic looking fashion.

Apart from that this week was all about getting the Augmented Reality parts of the experience working. I bought a webcam over the weekend for testing purposes and was able to get a 3D model of the sparrow in T-pose rotating like a ballerina in a music box with voice-over explanation of some bonus facts. I’ve made QR codes for each bird species and will scatter them throughout the terrain in scene 3 so that players can increase their knowledge of each bird species.

I’ve had some really nice feedback from a professional artist friend of mine who loved the vaporwave aesthetic and said that she found the whole experience very meditative :-)

--

--

R K Lewis

This is the journey of a Master of Art's in Virtual Reality