Friday, 27 May 2016

FMP Post mortem.

   This project has had a huge impact on me as an artist. I’ve had a fair few problems along the way, but through all the problems has come a huge mass of knowledge.
   The project we intended to create was a duo environment piece with a large space ship vehicle in one side and a series of characters to suit the world. The aim of the project was to have two individual environment pieces separate from each other, yet still connected. We decided on a contrast of sci fi futuristic and dystopian. I did the dystopian environment, as I wanted to include a mix of some foliage, industrial textures and a variety of landscape materials.
   The idea behind our project is a town split by some kind of ideology, which resulted in the erection of a large wall and barrier between them. Much like the Berlin wall. Sharn’s side (the futuristic side) retained all the power and money, so their side build up and flourished. Whereas my side (the dystopian side) crumbles and became more decrepit and rustic.

   Personally for my side, I wanted my area to have a sort of home hub feel. So it would include things like a weapons stall, apparel shop, repair place and somewhere for food. Maybe even a house for the player. This idea sparked from the central hub in borderlands 2, which I took inspiration and research from. Looking at the layout from multiple angles, including the building placement. Thinking about how the player moves around it really helped me build the foundations on my environment, especially when it comes to the different layers you move up and down in when you progress through the level. From this research I gained valuable knowledge of composition and environment planning.

Borderlands sanctuary research.
   At the start of the project I set a plan out, that would aid me through the project, so I wouldn’t deviate and get side tracked. My plan included setting time aside for things like modelling, texturing, engine work, group organisation at the end etc. I also left myself a two week buffer period at the end, thanks to the suggestion of my mentor, Dan. Just in case anything went wrong. I stuck to my plan throughout the majority of the project, but started to become overwhelmed at the end as I gave myself a large amount of work to do. I ended up cutting a few smaller elements from the project, such as the particles. I did this because I’d never used particles before, and knew it would take a bit of time to get used to using them and learning the basics. Because this would have taken a fair bit of time, I didn’t need to cut anything else, and the rest of the project ran fairly smoothly.
   Throughout the project I feel I have progressed a lot as an artist. I can see a very noticeable improvement from the style matrix projects. I started using substance designer and got to grips pretty fast with the workflow and interface. I have a long way to go with using it, but I still feel the textures I produced using the program are well made. Not only did I start using substance designer, but I learn about how to import parameters and make changes to the materials in engine will ease. This process gave me a good amount of flexibility when working in engine. Having the ability to change texture colours/densities/ brick amounts etc with the use of a slider really sped up workflow and gain better results.
Substance designer bricks.
UE4 brick parameters.
    Another program I really enjoyed working within was marvellous designer. This program allows you to made detailed pieces of clothing. Its main use is for fashion students. But I found it very useful for making canopies, banners, tarpaulins etc. I found its workflow easy and efficient. The act of adding fabric and cloth to my environment really pushed it as a whole and broke up the hard surfaces that I had filled the level with. From this workflow it lead me have a large amount of canopies with the same texture on them. So I used a blueprint with a colour parameter setting on it, to alter each colour individually.  I also used other programs to help with my workflow, such as Zbrush for sculpting more organic assets. 3ds max plugin’s for the wire fence, which reduced my workflow time. Leaving me with more time to spend on other things too. I feel I learnt a great deal within unreal engine too during this project. Things like using the command prompt, a further understanding of materials, decals, the foliage brush and more importantly the landscape editor. This project has been the best so far for the sheer amount of things learnt.

Marvelous designer progress.
   All these things have really helped me progress as an artist. I’ve not only learnt a lot when it comes to the programs I’ve used. But I’ve also learnt a lot with team work and working on a larger scale project. With both physical size, and time spent on a project. That was one of my biggest downfalls overall though. I had it in my mind that a larger environment would be better. But in reality, it’s about quality over quantity. I scaled down my project twice during the project. Once at the beginning near the end of the block out and secondly about 14 weeks in when guest lecturer Ben Keeling came in. If I hadn’t scaled it down then, It would have been unfinished a failure of a project I think.
   If I was to do the project again I would have only done an environment the size of the market. This way I could have taken a lot longer on assets to polish them to a higher standard. Another thing that I would do differently, would be adding more geometry to the assets I created. Giving them smaller sections to each asset. I was told half way through the project, that it’s nice to see smaller sections to an asset that you can focus on. Such as things that extrude from the side or top of the object. This can’t work for every asset, especially when basing models off real life. But I feel I could have done that for a handful of them in this project.

   We encountered a handful of problems out of our control too. Such as the foliage not being able to cast shadows due to the limitations of the engine. Which I confirmed by importing my grass in UE4.9 and testing. Decals didn’t shade properly either, again due to the engines limitations and not being able to render them in shadows. This problem is a reoccurring one and as far as I know it still isn’t fixed in the latest version. So this was unavoidable too.
Broken decals.
   What went right though, was my workflow, minus the reduction of the level. I feel I planned out the project pretty well, and the things that I changed, have only thought me what not to do in the future. I’m also very happy with the overall layout. The modularity of it all, and how I can make changes very easily. Building up sections and swapping things out is a great feature to have with a playable level this big. I also feel my textures are really coming along and growing by the project.
   Teamwork was a huge part of this project. We wanted to retain the teamwork essence. But still have individual things to show for it at the end. I definitely think we achieved that. Jake has his characters. Sharn has her side of the wall. I have my side and Emily has her ship. Despite us not having very entwined sections of the project, we still needed to work together towards a style, colour pallet and feel. Which again I think we achieved nicely. Jakes characters fit well in each environment. Emily’s ship looks like it’s a part of my environment and finally Sharn’s and my environments clash and have distinct contrast, as we planned. Group work does have it’s problems when you’re sharing an engine file. But we couldn’t find a way to all have access to the file at once. This became slightly more of a problem towards the end of the project. But it wasn’t anything we couldn’t handle.
   In conclusion, I deem this project a big success in my opinion. I have defiantly learnt a lot about various software, teamwork, workflow and other general things for when I start my next project. I had fun doing it. At no point did I feel overworked or too stressed. My routine was a steady one with a lot of hours put in and I can appreciate what I’ve achieved in this time. I think this will make a great addition to my portfolio and I can’t wait to move onto making improvements to it before degree show and then moving onto something else.

FMP Week 20 - The final Week of FMP and Uni.

   Well. Not only is it the last week of FMP, but its the last week of uni technically. After this we're on our own! Its been a very fast 3 years to say the least. Now its just degree show and graduation and then we're out into the world. But before all that, we still have this hectic week ahead.

   We were making some last minute changes and adding a few texture sheets. The idea was to have the project finished 2 weeks ago, but thanks to the suggestion of my mentor Dan, I gave myself a 2 week buffer at the end. And thank god I did. The texturing took a little longer than originally intended and we encountered a few other problems like decals, grass, crashes, loss of files etc.

   Jakes final versions of the characters were imported and placed after he tested them in a previous engine file a week or 2 ago. The animations ran smoothly aside from one mishap which was easily fixed. One character sits in the centre of the market, a nice start to your journey through the environment. The second is at the end of a dark tunnel on your way to the ship zone. I particularly like this one as you pass through the dark tunnel to only be illuminated by Jakes character at the end with his emissive backpack sign.

Jakes female character.
Jakes teenage merchant character.
Casto gave me some good crit today this week too. Dispite it being the end of the project. Either way it'll give me some things to do before degree show. He told me to include more grime on the walls, as more of a gradient from the floor. The plaster looks too much like primary colours, so maybe tweak those. The statue in the middle of the market could do with something more to make it stand out, more artistic vibrant graffiti etc. Which was my original plan, but I didnt want to rush the graffiti on it as its a centre peice to my level. So I opted for more of a subbtle graffiti and stickers. The final thing he suggested was a particular concrete tilable I have as one of the trim pieces on a building in the market. Not a big thing, but something to do before degree show.

   This week we wanted to get everything done by Monday to packages, and build lighting. But due to more tweaks we were pushed back to Wednesday. We had a test build on production lighting overnight. Which all was good and the engine was built again on Thursday and packaged that day too.

   Prior to this the changes I made included tweaking a few light maps and light map channels. Moving certain assets to avoid people getting out of the map. Adding collisions to the wire fence, which was previously removed to help mobility when testing, along with the wall collision. One mesh didn't have any smoothing groups on it because I hadn't reimported it in ages. Along with the finalised housekeeping of putting meshes in their desired folders and naming conventions etc. All was fixed!

   With working in a group its a lot harder to get everything planned down to a T. You're always going to encounter problems and when it comes to fixing things at the end it's very time consuming. If we all have a few tweaks, and everyone wants the engine file at once. It feels like ages before its your turn. But in the end we all got things done and handed in on time. We may not have everything we want in the final version and they'll always be tweaks to improve your work. But in the end you've just got to call it and move on. I feel I've learnt a great deal from this project. Both about working as a team and as an environment artist in general. So I consider this project to be a huge success from that alone. Now onto degree show!


Saturday, 21 May 2016

FMP Week 19 - The ends drawing closer and we're running out of time!

   Only 1 weeks left now. The second too last week has passed, and we were meant to be finished by now. But as for most things, not everything goes the way you want. But that's ok, because we have 2 weeks of buffering to help with that!

   Well anyway, this week consisted of more texturing, as always. More decals, grass, even more texturing, optimization and I've also encountered a few problems. I'm not going to talk about the texturing I've done this week, because whoever's actually reading my blog will surely be fed up of hearing about how I've been doing it for the past few weeks. So lets skim over that shall we? Good!

   The problem I had last week where the grass wasn't causing shadows is still an issue. I had this problem before during the style matrix project, and after searching through the settings I found the hidden "cast shadows" tick box. BUT! this was already ticked in this project, and after a bit of research and help from Craig. I found that UE4.8 has no fix. they introduced the fix in 4.9. Which is what I used for the style matrix. The only reason we didn't use 4.9 for FMP was because it has a habit of glitching and freezing the lighting build. The only solution for which, is to rebuild the whole scene. Out of the question for something this big. 4.10 wasn't finished at the start of the project, so that wasn't an option either. We thought it'd be best to keep too a finished engine version with no harsh bugs. This has only gave us problems in terms of grass, decals, and other settings that aren't added to this version. But those problems have just given us a reason to search for alternatives, which can only improve our knowledge of the engine and workflow.

Over the previous weekend I made some paper decals to layout on the floor, litter papers etc. Originally I made two versions, but it was apparent they where a fairly square texture. So I made a third more spread out and sparce decal to add tot he mix. These together are enough to give a good amount of variation in my opinion. I feel I could add more in the future, possibly for degree show. But These are great for now. They really add that extra level of variation to the floor after the macro variation from last week.
Paper floor decals with 2 versions.
Paper floor decals with 3 versions.

   During this project I upgraded my substance painter to version 2. I knew I might have some issues, but I figured I could always go back if something goes wrong. It turns out when you open the version of your file in substance painter 2, you can't reopen it in a previous version. So I'm stuck working in 2. This has been fine apart from one error that happens occasionally with certain files. In the top right hand corner of my diffuse map. And only my diffuse map. I get a dark square overlaying my texture. This only occurs on a few of my texture sheets. I don't know anyway to remove it, aside form exporting again and again and hoping for the best. I tried exporting as TGA'S (Packaged and unpackaged) JPEG'S, PNG's etc. but nothing worked! I also put a comment on the forums about it. I'm still awaiting a response.

Top right dark patch error.
   The way I got round a few of them was to export the texture sheet with only half of it loaded in. As I assumed it was because of what was on the texture sheet. Maybe it was too complex for painter to handle. Exporting parts at a time worked for one of the sheets but I couldn't do it for another as all the layers were used on each bit of the sheet.

   I also ended up removing all the padding from my maps in order to save on memory. It turns out this gives you really bad seems along the actual seems of your assets. So I re-exported all of them again, back to the original version with padding.

   Anya gave a nice presentation about the process she went through when creating her dioramas for FMP. She talked about the sketches, what you might be more inclined to include, what not to include, levels of a diorama and she also touched on colour theory too. Really interesting stuff, and I'll definitely be taking her points into account if I end up making a diorama in the future.