Week 4: Film Room Project Post Mortem

So after finishing the film room project last week I've had few days to let my brain calm down so I can actually begin to pick it apart.

The room we chose was an ambitious project for the time constrains we had to adhere to, which is not necessarily a bad thing, I just feel that as a whole we underestimated how long it would take to create the assets and room to a standard we were happy with.

Delicatessen Bathroom Scene Reference

We didn't do any concepting before modelling as a lot of the assets were common household items we could find reference for on the internet, however I feel it would have been good to do some concepting as it would have helped us achieve as more accurate feel to the room overall.

There was an issue with our planning as we had got confused with how long we had to do the project, meaning we had to adjust timeline which added pressure to complete it a week earlier.

Edited Time Plan

The modeling aspect I thought was a breeze until I had feed back from the group showing me where I could improve. In one case I had used unnecessary geometry and in another I could have used just a bit more to make a better overall model. I've learn a lot about the efficiency of modelling and where to implement strict use of a tri budget and where to add geometry where its really needed.

With texturing I learned that I need to keep my layers separate so that when it comes to adjudicating levels for the roughness/height maps it can be a breeze rather than a struggle. I also made a porcelain texture that didn't end up getting used as Josh's texture was better, this taught me to not be so precious about work that took time to make as sometimes it just isn't good enough and you have to try again.

Engine work was what we struggled with the most. I personally had a huge issue with my glass texture. In unreal you can't make a translucent alpha texture reflect any light, so I had a massive panic about how it stuck out like a sore thumb and how I felt like I had let the group down.

I then calmed myself and asked other groups how they had achieved a useable glass texture. I was directed to an online tutorial, this was perfect for clear glass, however I was trying to create a colored glass. I played around with a few options in unreal and was a bit closer but it was far too transparent so I edited the refraction options and it looked a bit better but it was far from what I wanted. There was simply not enough time left to fix what I wanted.

I really feel that we underestimated how long you needed to edit stuff in engine, and how its not just an add on at the end kind of thing. When sorting out the lighting, I was given my albedo textures back and told to fix them as the color correction for the green overlay had edited the appearance of half the objects so they didn't look right in the scene.

Final Scene

Our final scene was not 100% perfect but I think we did really well, we worked really well as a team which greatly benefited our productivity and efficiency in what we were producing. If we could do this project again, I would definitely plan more time for engine work and integrate it along my timeline of tasks rather than smack it on at the end.

Week 3: Film Scene Stress

This week I finished modelling my bathroom assets, I was very happy with them and thought I was right on track with our timescale.

Cabinet Model

Shelf and Bottles Model

 

Wall Light Model

However when we presented our current work to the tutors they informed us we had one week less than we thought we did. Great.

My calm collected feeling of this project running smoothly was soon upturned. So my group got together and we edited our timeline to fit the new time constraints.

Edited Timeline

This meant that I had to rear in gear and crack on with the unwrap[ping and texturing. My porcelain texture wasn't very good so we ended up using Josh's texture for all the porcelain to have a uniform feel to the room.

Bathroom Assets Unwrapped

After texturing everything else to a decent standard we started to put our assets into Unreal Engine 4.

I was really struggling with the glass texture, it was stressing me out so I gave up, I'm going to research how to make glass over the weekend and come back Monday fresh and ready to tackle the problem.

Week 2: Another Group Project - Film Scene

The first 2 days of this week were used doing the asset swap which I covered in my last blog post so I'll just dive in with the last half.

I learned that our next project was again a group project. The brief was to recreate an interesting film scene as accurate (minus the actors) as possible in 3D and put it into Unreal Engine 4 and include lighting with a 20,000 tri budget.

I joined a group with Josh Sam and Momo, last week they had their first idea of a scene from Amelie shot down, so it was back to the drawing board.

This is my moodboard that i presented to the group, I picked colourful scenes that had interesting geometry and that would still hold their own if the actors were not present. The group favorite was the top right, the engineering section of the Enterprise from 'Star Trek 2009'.

However we ended up picking the bathroom scene from 'The Delicatessen'

This scene was interesting enough when the figure taken out, and there was a manageable amount of assets to make between us. The tutors liked this scene too so went for it.

 We then delegated the assets between us equally using coloured dots to indicate how many assets and features we each had to make. Using this we drew up tri budget guide to try and stick to when modelling.

So far out of my list I have made my Bath and soap shelf, which together only came to 1572 tris, nearly 1000 under budget! I also made the cups which totalled at 220 tris each, almost double the budget but it was in well placed geometry.

Bath Model

Soap Shelf Model

Cup Model

Including catching up from last week and the work for this project I feel like I have achieved a lot this week. The feedback from my peers has helped me where I've been stuck and I hope I can continue to be this productive throughout the year.

Week 1: Viking Tavern Asset Swap and PBR

Well it's the start of my second year of university, first week back and I've missed everything. 

I missed this week due to outstanding circumstances out of my control, so I couldn't attend any lectures. However I caught up the following week in the first 2 days so I could go back and make this blog post.

Our first warm up project was a group project and the brief was to make the assets for a viking tavern scene; a table, a barrel, a bench, a shield, some cutlery and a hog. It sounds pretty simple, however each assets had to be passed along to another team member at each stage of development. So each team member never fully developed an asset by themselves.

However, since I missed this week I had to fully develop an asset myself, luckily one group was missing a shield, so I took on the project and produced the following work in about a day and a half.

Sheild Mood Board

At each stage I had Freddy Canton from my group give me feedback on what ideas to ditch and which to push forward in place of having the asset being passed on. It was really valuable to have another person's input on each stage of development.

Shield Concepts

Final Concept

Shield Model

Shield Unwrap

I found everything up until this point fairly easy but the course had a big shakeup in the summer and I was thrown in the deep end and had a mini breakdown when I learned that there was a new way of texturing. Physically based rendering (PBR). It definitely took a while to get my head around but once I did things began to flow smoothly again.

Shield Texture Maps

Above are my final renders for my shield. I'm really proud of what I achieved in the little time I had to catch up, sure there are faults that I want to improve but for how I've done in the time I had, I'm proud.

I've learned two main things. The first is that feedback from your peers is invaluable, everyone wants to see each other improve and they can sometimes provide the insight you need because you've been looking at your work too long to notice whats wrong, what can be improved etc. 

Secondly, Unreal Engine is horrible.