Oil Lamp Study for Skills Development

July 2016

I intended to study multiple photos so as to improve my modeling, lighting and compositing skills. I used substance painter, Arnold and Nuke for practice. Trying to match the outcome from Substance Painter, which is pbr based, to the render from Arnold was a bit tricky. This could also have been due to lighting. Matching the perspective between Maya and the photograph also turned out to be more challenging than expected.

These were some of the tutorials that I looked at for my oil lamp study.

Substance Painter Tutorial – Model Preparation 01: UV Mapping and Texel Density


Substance Painter Tutorial – Fundamentals 05: Baking textures


Substance Painter Tutorial – Model Preparation 05: Mesh export options


Forum Topic: best way of baking a curvature map for painter


Forum Topic: best way of baking a curvature map for painter


Forum Topic: Substance plug-in to Arnold


Using Xnormal:


This video shows a great technique for unwrapping curvy objects such as the wire on my lamp model. Instead of spending crazy amounts of time trying to select faces and edgeloops, this method uses the unitize function and then makes a selection of every UV, except along the seam, and sows this selection together in one neat shell which can be unfolded.


Blast Wave Compositing Practice

Following on from the Video Copilot blast wave tutorial, I attempted to create an explosion in my driveway in Donegal. The footage is kind of terrible as I don’t have access to a decent camera for filming but it still provided a good challenge for matching the assets to a similar quality, colour and contrast.

Driveway Explosion

The results are quite similar to what I created in the tutorial so now I’m curious about what different types of explosions could look like. I used the Action Essentials 2 pack (from video copilot) to create this so I’m also curious about how I could go about creating my own assets if the story so needed. Also, there’s no camera motion in this so it’s more apparent that the debris is disappearing.


Iglu Media and Football Euros Animation

It has been an informative 2 weeks working at Iglu Media. Andrew Coyle, Kerry McCormick,  Jonny Shields and I worked on two quick turnaround videos for the football Euros. The first video was an explainer style animation that explained the difference between Irish and Northern Irish football. That one was very successful with views as it very quickly got over 2,000,000 views on Facebook alone, then many extra on YouTube.

Irish and Northern Irish Football Explained

We then started the next one on Friday evening (June 10th) and managed to have it complete by Wednesday (June 15th). That one was more complicated as it involved character animation and lip syncing for the first time! This one is currently on around 200,000 views on Facebook. It’s interesting to see the difference in views between YouTube and Facebook.

Paris Bun music video:


Some reflection on what I’ve learned over the past two weeks:

  • There’s a great advantage to everyone working together in an office space and keeping the same office hours. Everyone was in early in the morning, had a reasonable lunch and worked until 5 or 6.There was simply no time to be distracted during our set work hours. In the past I’ve felt guilty about having this expectancy of everyone in the team. I understand that some people like to stay up extremely late at night and then sleep as needed. However, after seeing how well a team can work when we keep regular office hours and be together to discuss ideas or troubleshoot problems (or just as happily work at our own desks), I would really love to see this happening in our final year team.
  • When it comes to animating, having pre-recorded sound for timing is so priceless!
  • Having a tight deadline really forced us to be efficient as individuals and as a team. Planning was greatly helped by having a locked down script or song to work from. We could immediately jump into the storyboard stage. We were forced to quickly decide on a simple style that the whole team could adapt to, for drawing and animating. Jonny Shields was a great organised force to work with and set up the initial assets list that let the team plough ahead.
  •  The presence of a ‘boss’ like figure also helped. There was that extra subconscious pressure not to be fluffing around or having unneeded tea breaks.
  •  I’ve discovered gaps in my knowledge, particularly around graphic design and typography. It would also have been useful to know how to use Illustrator for creating vector based graphics to animate. I ended up drawing with the pen tool in Photoshop for the first time but I’ve seen previews of Illustrator tools that would have made my life easier. Jonny Kane pointed out that this is a recurring area (Illustrator) that people passing through have said would be useful but lacked knowledge in.
  •  Jonny Shields pointed out and helped me fix poor typography layouts. Having a knowledgeable person as part of your team to learn from is a great plus! Andrew Coyle also pointed the way to expressions that I had now known about in After Effects. Having everyone around for immediate feedback, when I asked, was a great plus also.
  • Usually I’m not afraid of failure but I found that this fear became a greater obstacle at the thought of other people having real dependencies on the outcome of my work. I came at this project with the thoughts that I knew absolutely nothing about football and I’d never created motion graphics like this in After Effects before. On top of that there was a crazy deadline. Looking back, it was a risk worth taking but I didn’t know that at the time. It was a relief when Andrew joined the project and then Kerry soon after. I couldn’t have asked for more reliable team mates. I still would have tried to tackle this by myself, even if that might not have been wise.


Stargazing Concept in After Effects

Christian suggested adding parallax and a vignette to the 2D stargazing scene to make it stronger for the pitch video. I finished the 2D concept a little more and animated the layers with a camera in After Effects. Unfortunately we didn’t get to complete our 3D version of this as we didn’t plan enough time to add the characters. I tried to use Sorcha’s design of Julie in this  (as it most match the Moebius-like style that we had discussed as a team) and then tried to match William to Julie. The lighting hid the line a little though and the colours changed after I colour corrected to fit within the scene.


Paint Effects City

Week 11

After the grass, rocks and trees, the next challenge to tackle was the city in the distance. To create the impressionistic style of the city from my concept I figured that simple building shapes that could catch the light would be enough. With not a lot of time left, Paint Effects seemed to be the quickest solution, plus I really enjoy how organic the process of distribution feels with Paint Effects. Unfortunately, I haven’t yet been able to figure out a method of distributing my own geometry with PFX so I resorted to using one of the default building brushes from the visor database.

I found that PFX will paint shapes perpendicular to the surface which is made paintable. This was a problem for my city designed on a steep hill and I found that adjusting the elevation produced too random of results (as one stroke had more than one house). I had houses pointing in all directions. I tried stacking planes which intersected the hill and painting on these instead. This method wasn’t perfect as there were some floating houses once I hid the planes but I got away with it seeing as the city was so far away and the style was more abstract than realistic.

Even though I tried to paint only on areas that would be seen, I still ended up with quite a heavy load of building geometry and some excess. After some frustration I remembered file referencing which I realised I haven’t been using enough of in the past. Even with using file referencing though, it still took a long time to load materials and switch between render layers with materials applied. Rendering was extremely quick though. Also, I built this approximately to real life meters which I’m not sure how necessary that was. I thought it might aid the parallax effect if we wanted to test the camera in Maya. Does the size of a polygon influence memory usage or is it just the number of polygons? Probably just the number right?

For shading, I used a ramp shader and directional light set-up similar to what I used for the rocks and played with the hue/saturation/brightness of the incandescence and shadow colour. For the water I used a reflective shader on a polygon plane and for the land I used a second plane placed over this with a transparency map made in Mudbox to indicate the coastal line. The land in particular needs developing. I was planning to make it seem at least more mountainous/hilly and perhaps indicate more forestation somehow.

Some render tests/progress:

Molly painted a sky which I tried adding to the scene also. The area light which I used to light the sky influenced the colours quite bit. I also haven’t added the clouds yet.

I rendered selectable areas for the sky, sea and land for colour correction in After Effects. This is the scene so far. We have yet to play with the characters in the scene and the environment still needs developing.

Stargazing: Rocks

Week 11

Getting the form, lighting and texture in 3D to fit the style which we wanted took a lot longer than anticipated. This is the aim which I wrote out for myself when I started:

“Make the rocks so matched to the painting that it would be impossible to tell that they were 3D if you composited them side by side into the painted concept. But, don’t let the concept dictate all, feel free to develop and improve when the opportunity arises.”

I started by taking the blocked proportions of the dolmen into Mudbox for sculpting. Although rock seems like a simple thing to sculpt it took me a while to get suitable forms. My first attempt was too rough and digital-clay like and looked even worse  when I tried to render the detail with a displacement map in Maya’s software renderer.

I looked up how to improve maya software displacement map settings but discovered that Mental Ray is a wiser choice for displacement maps.

I worked on a single rock, modified in Mudbox, for a while to see if I could achieve the style. I tried making very hard edged light by adjusting the intensity of a directional light. To create the fine carvings, I first tried using the knife tool in Mudbox but this gave me very lumpy results even after using the smooth and scrape tools. I also tried using the freeze tool but couldn’t create the desired results. I found that making a displacement map in Photoshop with a hard edged brush gave a much cleaner result and also maintained the smaller detail of the brush’s edge. The depth of this could then be increased by adjusting the alpha gain of the image file so as to make deeper shadows. By adding a few edge loops I could make hard rock edges without too much resolution and I tried bump mapping a canvas texture to see what this detail could look like.

From talking to Clare I was directed to look at the ramp shader with shading set to light direction and interpolation set to linear. From going through the ramp shader attributes I found that I could make the shading even more graphic like the painting if I cast the scene in shadow with a box and used the shadow colour to control the light colour and the incandescence to control the shadow colour.

The depth of the bump map, made from painted/canvas texture, controlled the detail on the interface between shadow and light giving a painted/dry brush edge to this area.

To create the rock shapes, I imported my blocked versions from Maya into Mudbox as these had roughly the silhouette and volume that I needed. I made the mistake of spending time UV unwrapping my modified boxes as I thought I would be applying maps onto them. As simple as they were, they still took time to unwrap using planar mapping as there were a few of them. Instead of importing the Mudbox sculpted detail as maps I ended up importing the level 1 subdivision of my rocks as .obj files into Maya and re-unwrapping them. Simple silhouettes were all that I needed but it might have been faster to figure out how to combine my Mudbox map and Photoshop map for applying as one map onto my first boxes.



Paint Effects for Grass and Trees

Week 11

Using the stargazing scenario, we challenged ourselves to see if we could translate the style which we wanted into 3D. Sorcha and Aidan worked on the characters while Molly and I worked on the environment.

For the environment, I started by focusing on the foreground. I blocked out where the rocks would be and moved some vertices around on a plane until it resembled the shape of the ground in my painting. I figured that it would be good to try and match my painting as closely as possible as I had already made a lot of compositional decisions while painting it.

I tackled the grass next and discovered Paint Effects, as I discussed in my post Paint Effects for Stargazing Environment. I found that the direction of the grass followed the direction of the stroke so it was better to paint stokes towards the focal point for compositional flow. I experimented with making different shapes with the tube’s attributes until I made a suitable grass shape, length and elevation.

Grass painted from left to right with no planned flow:

Grass painted with flow towards the characters:

Simple colours with a directional light:

As I learned from Mastering Autodesk Maya 2016 I was unable to use light linking with paint effects. I could unlink the light from the ground though and use incandescence to match its colour to the shadow colour.


Maya software render

This seemed to work well enough for the rocks also until I needed to use displacement maps which I’ll talk about in a different post. I rendered the grass separately in Maya software renderer and composited a texture in After Effects.

After I had some form of usable grass I moved onto tackling the rocks and then the trees. For the trees, I similarly spent some time editing the shapes of Paint Effects tubes to make the appearance of evergreen trees.


Trial and Error

Once I had a tree shape which I liked I looked up how I could save these settings as my own custom brush. This is the closest way that I’ve found to make my own brushes but I would still like to be able to paint with my own geometry made from scratch.

To animate the trees I used the turbulence attribute similar to the grass. The main challenge here was trying to make the trees look like they were in the same wind as the grass but reacting differently/slower due to their weight.

Turbulence too fast:

Adjusting the turbulence attributes:

For adding texture in After Effects, I had thought that using a background shader to shade the rocks with no shadows would work to create an alpha channel but in the end I needed to convert the Paint Effects to polygons and use surface shaders in order to make selectable areas. Using what I had learned about in my blog post Alpha Channels, Masks and Mattes, I was able to use these white/rgb shaded renders of the grass and trees as mattes for a canvas texture which I had created in Photoshop.

This was the composited result:


I don’t like how flat the trees look so I would like to return and continue experimenting with making the light catch foliage in a suitable way.