Module 4 - Sewer Shark Part 03

Notes for the third draft:

Gills are non-descript. Increase geo on the nose. Reduce polycount overall (dense geo in fins, tail, face). Collapse down some loops. Check loops all around. Finish the remaining 2 tasks from last week (side damage + drool). Add tertiary noise.

Sprint #7 'C' Level Delivery

This sprint I updated the neutral robot's arm, arms, and leg pieces (and evil arms and evil legs) to include the base sculpts I created before in addition to some new discernable features. I also created some quick materials to make everything easier to tell apart in the engine for now (the different pieces will be more realistic for their final iterations).

Module 4 - Sewer Shark Part 02

 Notes for Sculpt Second Draft:

- More attention to detail in eye area. ☑

- Harden up the armored plating. ☑

- Tighten up gills. ☑

- Sharpen tip of the nose. ☑

- Eye smaller (make it look more mad). ☑

- Corner mouth anatomy. ☑

- Add some kind of details to the stripey side (scars, garbage?) ☐

- (Check anatomy overall). ☑

- Teeth are very straight. Curve them out. ☑

        - Add acid drool. ☐

Plans for 3rd and Final Sculpt/Mesh Draft:

        - Finish the two remaining edits from list above (minor).

        - Procedural tertiary noise on skin areas (minor). 

        - Finish/polish teeth topology and UV's (didn't get to finish this week)(minor).

        - Possibly add tongue?

Module 4 - Sewer Shark Part 01

Week 1 - High Resolution Sculpt

Week 2 - Game Res Retopo, UV's

Week 3 - Materials and Final Display

Sprint #6 Delivery

My goals for this week were to get both the evil robot's head and arms into the experience. Unfortunately, we figured out that it isn't just a simple switch of the meshes to replace the current placeholder head - the new meshes need to be rigged and weighted to actually work in the level and we didn't have the resources for that this sprint.

However, I was able to decimate the head and test it in the scene + figure out a retopology strategy. I have most of the mechanical details separated by groups/subtools so I think I just need to bring down each individual piece to get the highest resolution game-res version vs. the current copy which is a derivative of the high res dynameshed together. I'll definitely try to combine clusters of pieces so that the final game-res won't be a collection of 50 different pieces; more like 3-5 or so.

I also started the sculpt of the evil arms - I blocked out the robotic hand piece in Maya so I have the low res of it to help when I retopo this (it will come in handy, you could say). I don't anticipate this taking too much more time, so I think by the next sprint it's fair to say I should have all of the evil pieces ready to go in the engine and functioning in the experience (head, arms, torso, legs).

Module 4 - Creature Project Proposal

Module 4 Project Ideas:

1. Bipedal Possum for Capstone Project

Additional References:

2. Creature Concept by Jordan Nelson

ArtStation Link

3. Shark Concepts by Olivier Désirée

ArtStation Link

Dragon Project - Final Materials/Textures and Final Display (Module 3 Week 11)

Environment scene and render above by Piper Seng!

Final dragon turn-in! I've spent most of this project fighting with the topology but I think I have finally got it under control enough that now it just needs a bit more polish to even out the last of the little bake problems. At this time last week, the bake was in bad shape - many scales I added into the sculpt added complex height around the mesh that I wrongly anticipated being able to simply bake down onto my original smooth, un-detailed mesh (great theory, not so great execution). I spent a lot of time trying to modify that mesh last week but I eventually cut my losses and ended up redoing the entire game res this week through manual quad-drawing.

If I could go back in time, I definitely would have tried to work out a way to bake each scale down to one flat plane each. I would still create the same set of IMM scales but I would have maybe tried to use dynamic subdivision levels for each so they could all be simply brought back down to the plane level at the end. Not entirely sure this would work better, but I'd be more likely to try a method like that for similar high-detail features in the future. *Edit: This would not necessarily work in all cases.