Tuesday, 6 April 2021

Fireplace PBR – Unreal, Materials, Lighting, Particles and Project Post-mortem

 Fireplace PBR – Unreal, Materials, Lighting, Particles and Project Post-mortem.

·         With the texturing complete, I can now move onto producing the final presentation of the project.

·         The process hit a major roadblock in one material which will take up most of this post in how it was problem solved, and then there will be a project post-mortem to finish the project off.


Tree Mesh Material:

·         Below is a collection of screenshots taken at various intervals of my process with the tree material, which was painful to get working in unreal:

1.     I made a basic set up for materials that had worked perfectly fine on my other materials – hooking up all my baked maps to the material output.

2.     This produced a non-transparent window that was white in colour, that somehow casted light patterns on the ceiling which I found strange.

3.     I added an opacity map through photoshop onto the textures which completely made the windowpanes invisible now. I needed to find a mid-point that would make the windows semi-transparent, but with an obvious orange glow.

4.     I then after talking with some fellow students thought to try making the material a subsurface shading modifier – which made the windows glow great but removed all colour from the rest of the tree.

5.     Here you can see the full extent of the lack of colour – the tree looks like it has been petrified.

6.     I got some help specifically with a node set up from Nina Klos in third year (https://www.artstation.com/ninaklos) who suggested this node set up where the areas not in the opacity map are excluded from the emissive glow instead of glowing black.

7.     I thought my opacity map might have been backwards, however after inverting the map it became obvious that the map was correct first time around.

8.     I did it the method that Nina suggested completely – however it had this strange effect of making the tree look like it had been inverted.

9.     I managed to fix the material through plugging in the RGB value to the subsurface colour result – which overall made the texture complete.

The final node set up for the tree can be seen on its own below the process.

Material set up process for tree

 

Final tree material set up


Making particle Butterflies:

·         When I originally concepted this piece, I knew I wanted to be inspired by both the colours and the panels in monarch butterfly wings. I thought I could nicely tie in that original idea to a set of particle systems, which could be used to add more life to the scene than just the fire material.

1.     This was the basic material node set up. The system was made through following this tutorial: https://www.youtube.com/watch?v=6f5Mrr_OT6s&ab_channel=CGHOW The core idea for the system is that you have a mesh particle for the butterfly, then have a gradient map set up with a value encompassing the centre body of the butterfly and then graduating to another value at the ends of the wings. You can then plug those values into parameters that control the height of aspects of the mesh at a given point, so you can animate the wings to flap as often as you want.

2.     This section shows the event timeline for the particle – where I just alternated between 1 and 0 values for the lifespan of the butterfly.

3.     The result was an overall success – and really helped make the scene more dynamic and alive.


Butterflies Process


Fire Particle:

·         The brief required me to make a fire particle set up, and I was given one of two methods; I could either draw out frames for an animation for the fire or make a map to be manipulated as a material. I chose the map material, simply because the other method was something I did before in my medieval shop for the smoke effect. I am also invested heavily into shaders and materials – so the idea of making a dynamic material from a simple map was more exciting to me.

·         I constructed a very similar map for my fire, and replicated the node set up for the fire almost exactly (give or take values and colours for my scene).

·         I kept this map 512x512 so that I could afford a 512x512 noise texture too, which helped with the panning of the fire effect.

·         I ran into an issue that I never really solved with my fire – for some reason, the material wanted to completely tile whenever I changed certain values, and really liked sticking to the top right corner when animated. I did my best to mask it – but it was something I never found a fix for.

·         To improve the atmosphere around the fire texture, I also followed the added tutorial on making little sparks come from the fire and used the same particle to make smoke coming from the flames.


Fire map

 

Fire material set up

 


Final scene with fire



Final Conclusions/post-mortem:

·         I am extremely pleased with the outcome of this project, and think it clearly shows how much time I spent on it. I was worried if I would be able to adapt to the new content in both this project and in my gladiator project, but I have to say I really enjoy the PBR workflow in both substance and in making high poly models. It eliminated the aspect I normally loathe in projects because it takes a long time to hand paint textures, and to get the textures to look like specific materials too.

·         I think I am getting more confident too overall and I have really been pushing myself to find something else to add in these projects, whether it be a particle system or a shader technique, and I can feel myself slowly taking in the potential of what you can and cannot do in unreal.

·         I also have been an avid user of my Trello again, and have taken extra steps this project to plan out the budgets I have set. The UV sheet planning was a life saver both for allowing me to understand how much I had to work with, and to let me edit the amount of content I was putting in this scene, because this project could have quickly gone wrong if I did not analyse composition at every step of the way.

·         I enjoyed the process of this project more than the gladiator project, but I love the results of both of my projects from this term.

 


No comments:

Post a Comment