Week 6- The end of it all and last minute matinees

So here we are, at the end of the beginning of year 2, it’s been a long 6 weeks, but at least it’s been a fun one.

A couple weeks ago we started the film room project, where we need to replicate a scene from a film as faithfully as we can.

We spent a good two days working out what film we wanted to do and we ended up going with a scene from Guardians of the Galaxy where the main characters are all having a pep talk on a spaceship. Looking at the scene it looks quite complicated with all the wires and beams and just general stuff spread out everywhere but we ended up going for this scene because of the amount of assets we could play with and learn from and because of how many times we could iterate on them.

We also loved the color; Deep Reds with bright blue accents from the hologram screens, we thought this would be really fun to pull off with some emissive textures then we looked at the size, its not too small a room so that we had trouble finding stuff for people to do, and its not too big so that we wouldn't be able to cover everything, it was just the right size for a team of 5 and looking at the scene we could easily find places to use what Mike Pickton had been teaching us during the week.

But we also looked at the Starship enterprise bridge from Star Trek as an idea.

Which was a very serious consideration because of fun the models would be to make, but at the end of the day when it came to a toss up between Guardians of the Galaxy and Star Trek, Star Trek had some very fake looking props (which is just Star treks rather campy aesthetic) and we thought they would have looked even more fake in 3D thus came to the conclusion that Guardian's Rusty Grunge aesthetic would be alot more fun and it would be easier to make the scene more immersive.

We also Looked at Oblivion.

We loved the colours and the props we could have modelled in this scene, then we realised its mostly an exterior shot, I mean we could have modelled just the interior, but the inteiror has transparent glass walls, meaning we'd have to do both an exterior and an interior and if we had more time we would have done it, but given our time frame and the amount of assets we'd need to pull off we thought it would be too time consuming to do well.

But back to our chosen still from Guardians of the Galaxy.

Looking at the scene just by myself I really do like the colours, they’re my favorite part of the scene and when we get down to actually making it the colour and textures are what will make or break the project. I was also quite worried with the scene perhaps not having enough reference (since we couldnt see alot of the room in this scene), but again upon further analysis and watching more of the film we gathered sufficient enough reference to try and make the scene as faithfully as we could, eventually creating two master pdfs filled with every scene containing this room. Wanting to focus on character art I’d have more fun trying to model groot or Starlord in this scene, but you make do with what you have, so I volunteered to do the center light piece and wires,I figured the pressure of trying to model one of the centerpiece assets I’d really be able to push myself to improve my skills. The wires would have ended up being the most complicated part of the scene just due to the sheer amount of them, it would have destroyed our Tri count if we hadnt found another way of modelling them in the scene.

So when it came to jumping into modelling my first reaction was to do some concepts as at this point some of the others were sorting out documents in the way of spreadsheets and checklists

for us to use to help our admin along. I only actually ended up finishing one Ortho of the chair before we had everyone agreed on what we were doing, the fact I made an Ortho of the chair and not the light is evident of this. Still helped Luke model his chair in the end so it wasnt any wasted time, he made a damn good chair.

But since I was given the Centre light, the wires and a small door console to do for my assets I did some quick drawings of the light and jumped straight into modelling, first making a cylinder and extruding, chamfering and insetting from there, cleaning up Ngons as I went, I still find it scary how much more efficiently I can model compared to last year. Having said that though, this project clarified my understanding of 3ds max, especially concepts I found dubious last year, like resetting the pivot and setting my unit measurements to the correct one, small things but ones that can have a big impact later.

UVs and texturing is still one of my bigger issues, its getting better but I cant help but feel they can always be better. However Unwrapping has been made much clearer doing this project, one of my group members Beth helped me out and helped me figure out a few tricks to help it along, and as a result I’m very proud of my UV layout, with the center light being effectively a 4-1 texture and I understand unwrapping alot more now.

I've been mostly able to get textures just from taking pictures but with the one select texture from CG textures just to provide a base for the center light.

Otherwise a good majority, like I said were from my pictures of pots and pans from around the house, which I can always use at a later date.

The main problem I have with texturing is just making it look realistic and not so flat, I can problably blame that on my inability to paint digitally, something which I am now working on with my personal projects. It wasn't a particularly fun texture to do, namely because its very clean and grey metal, however the emissives were extremely fun and probably my favorite part of the process, learning how to do emissives also paved the way for modifying metalness and roughness in engine which is painfully easy once you know what your effective “ingredients” are.

Engine work was actually very compelling, it wasnt easy but very compelling because you could see your results right away and change them on the fly, I also got a deep satisfaction when things started coming together. Sean and I also worked on making a blueprint to add the wires more efficiently, thus we compiled a blueprint to procedurally generate meshes on spline points throughout the scene, effectively making ourselves a lego kit from which we dressed the scene with. However, we just looked up a video telling us how to do it, which we are still trying to figure out, to sum up what we did; we baked a cake and its great, but we dont quite know what eggs are, so an objective for us to understand more in depth on what we actually did in the blueprint, since down the line this could be very helpful. Sean has been working on most of the lighting in the scene, Sean had the smaller assets, many of which are not in the main still, so he chose to be our engine monkey and I personally think he did a great job, at the time of writing this the scene is a little dark but I believe we can fix this with some post processing work.

We have slipped up slightly on engine though, when we imported our assets into engine we had scaling issues, it seems unreal and max have different ideas on what a centimeter is, but regardless we eventually got past it. On top of that lightmaps were giving us trouble, so we ended up having to make them manually, its easy depending on your asset, unfortunately for me it took about an hour to manually make my lightmap as I had so many overlapping UVs to save space, but we got through it quite well.

As a team I think we worked quite well, 4 of the 5 of us turned up consistently and helped each other when needed, and when a team member couldn't continue due to illness we were able to control the damage and pick up where he left off with little to no problems. For me this was sort of a blessing in disguise, Luke fell ill and couldn’t continue so I took up one of his assets textures, as a result I learnt a lot about making Alphas which is something I botched last year In my eyes. I think our biggest mistake was taking too long deciding who’s doing what, it took us about 3-4 days to figure out who was doing what during that time we could have finished modelling, it wasn't a train smash but I think we lost a bit of time there. We made an action plan and schedule early on in the project, but after the first week or so we stopped sticking to it and we ran into problems, again no train smash but we fell behind slightly with textures only being implemented during week 5-6 Fixing model problems once we imported into engine is also slowed us down a bit, we had a few scaling issues like I mentioned which we fixed over a week whilst optimizing our models tri count, not entirely sure how we could amend this since unreal and max seem to have different ideas on what a centimeter is, but I’m sure there is a way so it’s worth looking into. Overall I think we worked very well together, no glaring differences of opinion or personalities clashing, we were very professional in my eyes. Towards the end of the project however it we ran into a few roadblocks, namely in the way of Jack handing in his textures during week 6 with incorrect naming conventions and never really coming in so we could evaluate how things were looking. As a result his textures looked horrible and I needed to edit and repaint them so we actually had a floor texture in the end, it was just unnecessary stress but we did it and I’m glad because I’m pleased with the end result, even if I’m not totally 100% on it. We also ran into some issues creating the matinee and taking the final beauty shots, in the beauty shots we had a really strong bloom which blurred out the entire thing, so we needed to take down the emissive intensity (which is unfortunate as it looked great when playing the level). On top of that the matinee was hideously slow when we tried filming it, so we had to bump down the resolution so we could meet the deadline, unfortunate and avoidable, but again like every roadblock we got through it and learnt from it and even despite all the hiccups towards the end, I think we did really well, this Is the best looking project we’ve done so far and I think we’d be being way too hard on ourselves if we said we did badly because a few things didn’t look the way we completely wanted to, at the end of the day it’s a learning experiences and we can’t deny that we learnt a hell of a lot during the project, I can’t stress that enough.

Overall, If I had to look at our final image, I’m impressed as to what we pulled off; our lighting is great albeit a bit dark, something we could have fixed with post processing and our models look great with some small fixes to the tri count and texturing they would have been fantastic and those are small things which disappointingly we could have fixed if we managed our time better and were more disciplined sticking to our action plan. Our scene could have looked warmer in accordance with the still we chose, but again we could fix that in color grading. All of the problems I have with the project are easily fixed with some small tweaks here and there and its shame we didn’t implement them, I still think it looks great however and I’m looking forward to implementing what I’ve learnt here in to everything I have coming.

Team Manifesto: We chose Yondu’s ship from Guardians of the galaxy to recreate as 3D playable level, because we thought the scene stood out well without the characters, due to its well thought out composition and strong lighting. There are many of different materials which will make our scene much more fun to texture, and also make a captivating level for players to explore. The colour palette in our chosen still is quite warm and grungy but has contrasting cool tones placed throughout the scene. We felt that there was a good amount of assets involved in the scene, which will push us, but will also be achievable as a whole in the time we have been given and the amount of members in our group. We feel that our final still image, the ‘money shot’ will look impressive when everything is modelled, but even more so when the player will explore our 3D level. Overall we feel this project will challenge us, but still be an achievable amount of work to be finished within the time limit we have been given. We have also set a few stretch goals in place if we have enough time left over, but they are not vital assets to our scene.

At the start of our project, after we had all agreed on this film room and made sure everyone in the group was happy with the decision, we started to assign assets. We made a list of the objects and assets in the room and asked everyone what they would prefer to model. Once the group had individually expressed their interests, we divided the assets fairly whilst checking if everyone was comfortable with their workload. To ensure communication was strong within our team we exchanged phone numbers and made a Facebook group to keep everyone updated on the project progress. A Google Drive was set up so that everyone had access to each other’s work quickly and to have a group backup of all our work. The asset list was an editable excel document so everyone could update the file with their progress. We were able to keep track of our tri budget and see if anyone was falling behind. As a group we gave ourselves achievable deadlines for each part of the development pipeline that we felt were realistic, we suck to them for the first 4 weeks but toward the end of the project we fell a bit behind. During this time we helped each other if someone was struggling or needed help with certain aspects of the project, we all spoke freely about problems that we were each experiencing and got help from one another. As a whole our group was well organised and fair with distributing work within our project. We experienced a few problems here and there throughout the assignment but as a team we discussed the issues and worked together to resolve them.