This semester our main project was studio based, my group had the pleasure of working with Boomclap Studios on a low poly VR experience based on a TED talk on the creation of the world.
My group consisted of James, Lydia, Mark, and Shioban.
As I previously mentioned the virtual reality short was based on a TED talk on the creation of the world. The studio had written a 5-minute poem to be used as narration for the 3D visuals.
The experience begins in an interactive bedroom where the user can engage with and move several objects – For example, a fireable BB Gun, a football, toys etc. The room also has certain trigger points that engage other sections of the story. For example, when the scrabble board type pieces are plugged on to the periodic table it triggers a voice line/narration and focuses the lights on it. The same way as when the user looks up through the velux window on the roof the poem and visuals are triggered.
Unity – Scene assembling, FX, running it in VR.
Maya – Modelling, Rigging, Animation.
Photoshop – Concepting.
I personally mainly used Maya as I was asked to focus primarily on low poly animal designs and rigs, unfortunately not really getting much use out of Unity.
Communication Method – Slack
One of the weirdest parts of this project for me was getting used to modeling in tri’s for Unity.
Within this project, I got the chance to model and rig many low poly assets for VR, the majority of them being sea creatures and fish. This low poly style is new to me and I initially didn’t like it an awful lot. However, it did begin to grow on me throughout the project.
To get me into that fishy under the sea vibe I initially started out with a bit of concept art. Here is an Anglerfish that I drew.
Overall, I didn’t get much of a chance or opportunity to do concepts or 2D artwork for this project. This was mainly because everything was to be made low poly so the assets wouldn’t slow the frame rate for VR (and to fit in with the overall style of course).
Note – I was asked to leave my models untextured as they were wanting to use the “skinner effect” in Unity on them which just takes their active vertices and creates FX that trail when the model is animated.
I began with finding a fish reference for my model, it had to be a clear silhouette as the fish is just going to quickly swim past the viewer. So I began modeling a fish with plenty of fins. I based it on this illustration of a Bigmouth Buffalo (Ictiobus cyprinellus).
Below is the fish from various different angles as if it were a technical drawing. I began with the typical cube and modeled it in quads initially.
I then fired on a basic joint chain for each of the fins and the rough skeleton ready for animation. I made controls as well. However, they weren’t quite working out as well as I hoped they would, this was mainly down to the lack of joints in the spine.
I added a few more controls, joints and overall developed the rig a little more. However, it still wasn’t working how I wanted it to. At this stage the studio also asked for me to exagerate the proportions a little more as they weren’t going for realistic. So I stretched it out a little more.
Anyways, I thought that there bound to be a better way to do this rather than having to manually animate each control. I found a few tutorials that were useful in the rigging process. However, they didn’t quite cover everything that I was looking for, so I had to add a few things here and there. (Below is the tutorial that was most useful).
I redid the rig and used a sine deformer to create the main swim controller allowing me to adjust wave length/amplitude/speed etc. It made animating it substantially easier.
Here is the first iteration with the sine deformer method.
Below is the rig above in motion…
This rig actually has NCloth attached to the top fin which is why is it looks a little weird. The studio informed me that it is difficult to use NCloth in Unity, therefore, I redid the rig a bit. They also wanted a few more fish options… So I began modelling more.
However, at this point I was working on my whale models for the same scene. So I will switch to those for now…
Whale No. 1
I initially modelled a whale based on the references they posted in slack. I used this image as my primary reference.
I can see the appeal in this low poly art style. However, it really isn’t for me, but I made one in the same sort of style anyways to give the studio the option.
As you can see I made a few changes to it but I tried to keep it relatively similar to the one that they preferred.
Whale No. 2
I personally prefer my second iteration of the whale (below), I based it on a Dishonored creature. However, the proportions and style are far more intriguing for me (even if it is a couple of hundred polys more).
Here is the Dishonored whale that I based my design on.
After modeling the whales, I decided to model a pre-historic hammerhead type shark creature for fun. I enjoyed making this and modeled it slightly differently in comparison to the previous models. I made it in multiple different parts – The head/fins/body to try and capture the different areas correctly.
Back to Fish…
As I said before, the studio asked me for some more fish variants. So I used this image as a reference to make a few different fish variants.
I used this as a plane and created a few of the ones that I thought were more unique/interesting.
As you can see from the image above, some took more iterations than others. But I went with 4 different kinds of fish from this image – the Bass, the Dolphin Fish-Mahi, the Marlin and the Salmon.
To begin with, I modeled the basic shape in quads to start off with, then later using the reduce tool I lowered the poly count then I triangulated the model. I then manually merge more vertexes to bring out the low poly style and sometimes reduce it again.
I also made another one basing it off the references on the studio slack channel and looks quite heavily stylized.
I feel it turned out quite cool for a low poly stylized fish. (The reference image is below).
In the end, the studio liked the Salmon the best as it had the strongest silhouette. So I remade my rig again to help me with the process for the other rigs later on.
Rigging the Salmon
I followed the same process as before making use of the sine deformer within the rig. I tidied up the outliner in this one and made better custom controls.
This rig has been through a number of variations but eventually, I am happy with it and provides the model with a nice smooth movement that I feel accurately represents a fish movement.
Since this rig is done, I decided to apply the rig to the other fish, to give the studio the option again. They still preferred this salmon just for simplicity purposes and for its easily read silhouette (as it is intended to be in a shoal of fish).
Here is the rig applied to the other fish (below), I made them all swim at different speeds.
The next thing was figuring out how t make them swim along a path…
Fish Swimming along a Curve
Alec suggested looking at this tutorial below, I looked at it and tested it. It was good, but I wanted to see if there was another way that would work better with my rig (since I had spent the time making a decent rig).
After a good bit of searching, I found this one which seemed much better and made use of both aspects I was looking for – a curve to swim along, the sine deformer.
This did need a few adjustments to make it work with the rig as it was just showing the basis, which made it a little bit of a hassle. However, it turned out quite well in the end I think.
The only issue is that at the encounters a few problems at the end of the line it reverts to the sine deformer. Therefore becoming a bit of a problem later when I was trying to move a shoal of them passed the camera where the end of the path is visible.
They decided to remove this shoal section from the story anyways. They just want to have the animals passing the camera one by one now, for that reason I moved on.
I was trying to think of other animals that would be fun to model and rig, the manta ray was one of the main ones that came to mind. Therefore, I started looking at reference images of real mantas and fired away with it. (I also looked at mantas swimming at this point as well).
(References on the images in the slideshow – Please have a look at this video – Diving with Manta Rays, there is something peaceful about it.)
My first version was way too high poly for what they wanted, I knew that before even showing them it. Therefore, I started making iterations of it before showing it to them, each lower poly than the last.
It still is low poly, but it could be optimized better, I didn’t need the gills actually connected to the interior of the mouth (as shown in the images/video above). Therefore, I lowered the poly count and filled in the holes.
Next I triangulated it and reduced the poly count further.
Below is the final iteration of the Manta Ray with reduced poly count and slightly better topology…
I then rigged this so it would be able to be animated. It is made using IK spline handles mainly with a various cluster deformers along the wings and tail.
The Eagle was definitely the most complicated model out of all my models for this project. I began with looking up references, here are a few of the ones I was looking at…
Here is my concept.
As you can see my model changed substantially from the original concept sketch. I personally feel that the model looks substantially better, but the concept helped me get there like. I used a combination of the reference images of actual eagles and the concept when modelling this eagle.
I don’t think I have any progress shots. However, I started by making the main wing sections and the body blocked in. All as separate pieces, I then began making the individual exterior feathers on a new object. I positioned them all and then started merging vertices. I then created the interior feathers and merged the wing together. I then made the head and joined it to the body and the same with the rear feathers and feet/talons.
Here is a rough turntable below.
The rig was also the most complex in relation to my other rigs for this project.
I used an IK Spline with Cluster deformers along the curve allowing me to create the wing flaps.
I made an overall wing controller to control the individual feathers as well as groups of feathers.
As you can see it is a very similar rig to my Manta Ray, the Manta Ray used elements from this including some of the NURBs controllers and joint setups. However, they did have to be adjusted.
In case it needed to do anything else I wanted to make sure it was pretty much fully rigged. Here is a quick action shot below…
The Plesiosaurus was fun to make, and the rig was very similar to the fish rig (making use of the sine deformer again). I based it on a Plesiosaurus/Elasmosaurus, which are very similar in shape.
Here is my model below with a basic animation using the start of this rig, I don’t think they wanted it so I haven’t finished the rig yet.
After the animals they wanted us to make other props and Islands
Islands and Stuff…
The first asset I made was a hand, I based it heavily on my own hand as they didn’t specify a particular style. Unfortunately, they were thinking more of an emoji hand. However, here is the one I made and rigged…
The rig looked something like this, it didn’t transfer overly well into tri’s initially and the skin weights were all messed up so it had to be re-attached several times.
I tested the rig by positioning the hand in a few different ways and with a few different movements. It is alright, it functions for what it needs to do, if I was to re-do it I would adjust the thumb a little more and create a better section for the knuckles.
As I said earlier, they wanted more of an emoji hand (which they still didn’t tell me at this point). They asked me to round off the base/wrist. Here it is below…
The didn’t end up using this hand model in the end, I think they cut the hand from the idea.
We were supposed to make islands based on different civilizations. This was left extremely loose and was supposed to be early civilizations, some others in my group had already covered tepee tents/early buildings. Therefore, I decided to go for a Jerusalem style building (like biblical times buildings).
Crashed Rocket Ship
This was a mixture of being an island and just a general asset. The crashed rocket ship obviously isn’t a civilization, but it was under a different category on the list and I was very keen to make it.
The wings were probably the most inconvenient bit to model since I redid the entire inner section where the engine would be.
I textured it and made an iteration that possibly looks more like a rocket ship.
I asked a few people and they said that the preferred the first iteration more, so I stuck with it instead of the second one…
I then made a low poly island like the previous one and created a crash trail along the ground of where the rocket would crash. I also added a bit of a scale animation on the smoke. If I were to do it again I would create actual wee smoke clouds getting smaller as they go up into the air.
We were asked specifically for some Sahara Islands, I wanted to put a creature into mine. Therefore, I considered the main animals that would have worked with it and I decided on making an elephant (more difficult than initially expected to make a low poly one).
However, before I started into the elephant I looked for references for a Sahara plain. I really liked the trees in Lion King as well (last image on the slideshow below).
I modeled a small Sahara Island with a lone tree like the images above and a small oasis type pool of water with some loose vegetation.
I modeled an adult and infant elephant both scaled slightly differently to suit their ages…
I then applied a basic (and unfinished) rig to the adult elephant just to get the positioning that I wanted it to have. Since the rig wasn’t complete and finished I just decided to soft select the infant elephant and move it around that way instead.
It was at this point that I also adjusted the general size/shape of the island to make it work a little better…
If I were to do this again, I would finish the try to finish the rig and then add more animation to it, as this is definitely my favorite island for this project.
The Studio asked if I could get a shoal of my fish swimming along and then get disrupted/broken by a whale swimming through them…
I started researching this, I asked if this is what they were sort of looking for…
I found these videos/tutorials very useful in regards to this.
Here was my first successful attempt at it…
Here was my attempt with it being rendered to look like it’s properly underwater…
I attempted to create the effect of the whale swimming through using passive colliders. The sphere is intended to be the whale in this situation…
This was more complicated than expected, as I found out that I was using a Legacy script, therefore, I had to change the .mel script for the particle curve emitter effect to make it function as a non-legacy particle to allow it to work with the current effects. Here’s a tutorial on how to do it, I tried figuring it out for quite a long time before coming across this tutorial on how to fix it.
Unfortunately, due to time restraints and other looming deadlines, I didn’t get this shoal finished. However, I found it very interesting to learn about and if other deadlines weren’t there I would have really liked to have finished it off.
In summary, I have really enjoyed working on this project. I have definitely learned from and gained experience through the creation of my assets and rigs. The Boomclap studio was very friendly and accommodating with meetings and it was great to work with them. However, I would’ve liked to do more or help more with Unity and learn more about it. I felt that we weren’t always kept in the loop, which I can understand why but it would have been nice to work closer with the studio. Overall, I am pleased with this projects outcome and the semester in general.