BA1b Hand In - Year 1 Final Lecture

BA1b submission details:

Primitive modelling
Lamp
Barrel
Phone
Pillar
Time Machine

01_development_fse.pdf

Final renders of tasks 1-4 screenshots
Lamp 1
Lamp 2
Barrel
Phone
Pillar

(Personal work in miscellaneous folder)

02_assets_FSE.pdf
Final renders of task 5 screenshots
Time Machine

03_turntable_task_#_fse.mov
Turntable renders of task 4&5

QuickTime (h264) // 1920x1080 // 25fps // 10 seconds maximum each

04_journal_fse.pdf
Research posts
Weekly summaries
Critical evaluation
Include blog URL on page

*Talk about issues and how you overcame them, mention research used to assist (YouTube tutorials, Google searches, forum posts). Limitations and stuff.

*Remember to do concise weekly summaries, ask coursemates for advice on this.

Marmoset_fse (folder)
Contains toolbag scenes (.tbscene)
Texture atlases for task 4&5
UE4 scenes *optional* be submitted here
Task4_Normal
Task4_Albedo
Task4_Specular
Task4_AO
Etc.
(Put obj/fbx of task 4&5 in folders just in case)

05_presentationSlides_fse.pdf
Submit your slides from the group presentation

06_gameStudies_fse.pdf
Submit your essay

07_selfevaluation_fse.pdf
Submit a completed 'BA1b self-evaluation form' - available on the VLE

*Submission at 14:30 in Media Lab (MO21) - get everything finished for Wednesday!

LO5 - demonstrated through blog
LO6 -
LO7 -
LO8 -
LO9 -
LO10 -

Damaged Pillar: Doric Column UE4 Screenshots

Time Machine: Sci-fi Circle Map Rendering

Pylon final maps

My final normal map, though simple, I nailed the look I was going for when I set out.

Ambient Occlusion map from Xnormal. This didn't really contain enough information to correctly define the model.

Using Quixel I converted my normal map to an AO map and then used the AO map from Xnormal as an overlay, the result on the left.

Diffuse map, I used this as a guide for my albedo map, which was to be quite simple in design.

Using a brush that I had designed in a digital painting workshop I applied large red stripes. I then worked into these stripes using the eraser tool, edited the layers overlay settings and opacity change.

This is my finished albedo map, wherein I also applied black paint to some sections of the model.

My Metalness and Roughness maps. With this model anything that wan't metallic was rough, which was a boon when designing these maps as I needed only to invert the metalness map to create the roughness map.

An unused Emissive map. I was trying to apply something similar to LED strip lighting, this didn't look quite like it fitted on my model, though.

Specular map, created by converting normal map into specular in Quixel. This worked really well and helped my model look as good as it does.

We do not speak of the height map.

Object space map (unused)

Time Machine: Sci-fi Circle Normal Map

Successes and failures whilst normal mapping my pylon.

I'm unsure what happened to the map on the left, but there's heavy distortion all over the mesh and none of the information is near correct.

This normal map had my indents going in the wrong direction and looked strange on my model, this is what happened when I imported the normal map that I had baked in Xnormal to Quixel and converted it from a PNG to a PSD.

Though I liked this normal map, she indentations were a little too extreme, I fixed this in quixel.

This is the map used on my final model, which worked really well.

Time Machine: Sci-fi Circle

Inspired by my previous assignment I decided to create a time machine composed of a circle of pylons. I took a more science fiction approach using large flat planes to define the shape.

I tested this idea before I went any further by duplicating my mesh into a circle in Maya. I was really pleased with this preliminary check and decided to continue.

Unwrapping the UVs was simple due to the objects extremely low complexity.

The completed UV map to the right.

This is a screenshot from the turntable of my final pylon design with Albedo, Normal, Specular, Roughness and Metalness maps applied. My 

Here is the pillar among friends, a circle of. This Circle of pylons would send players throughout time, to a random destination.

Damaged Pillar: Doric Column Map Renders

Albedo Map
Using two photographs of sandstone

Normal Map
Obtained from Xnormal after hours of work

Ambient Occlusion Map
Made using an AO map from Xnormal and Quixel merged together

Xnormal AO

Quixel AO

Specular map
Converted from Normal Map in Quixel

Height Map
Absolutely no, this did not work

Displacement Map
Also no, this was an absolute failure

Damaged Pillar: Doric Column Normal Mapping

1. Too low resolution

2. Not dense enough

3. What

4. This is better

















5. Why are you doing this to me

6. Finally

7. This is wrong, damn you Quixel

8. Strange glitches on square face

Object space map, just to see if I could.

Damaged Pillar: Doric Column

After struggling for a while to model my first pillar design, I moved onto something a little more achievable. I began to design the low poly mesh as a Doric Column.

After placing all of the largest shapes in the correct areas, extruding and combining the bottom and top half; I was ready to export the mesh from Maya as an OBJ file and import it to Zbrush.

Whilst in Zbrush I attacked my column with all of the wear and tear I wanted to see on the final render.

I realized I had missed a step in decimating my mesh, as the damaged sections were far too high in tri count in comparison to the rest of my mesh.

I decimated my mesh in Zbrush then reimported it. From left to right are the; Decimated, Low Poly, Damage and High Poly meshes. The damage now had to be bridged onto the low poly.

Bridging 1

Bridging 2

This process is not exactly difficult, but it is one of the most time consuming tasks I have ever encountered. It can become quite dull after a while, though the finished result is worth the effort.

UV mapping

Baking

Something went horribly wrong with my bake and the normals were crushed, stretched and distorted. I spent a long time attempting to fix this as can be seen below.

My first UV map, which did work, but it wasn't densely packed enough to make use of the 2048x2048 resolution.

I encountered issues with this map and I am still unsure why. The low poly and cage were both the exact same topology and the high poly was in the exact same position.

I decided it would be easiest to map the whole model again as opposed to puzzling over it for hours on end.

Here can be seen my final UV map, which worked perfectly. I am extremely happy to have this finished.