Week 1: First Week Back!

Our first week back had us get back into the whole Game Art world and the courses new structure, which now combines the 3D and 2D sides much more effectively. The critical studies section of the courses has also been updated, one part of it is to write up small weekly updates and this is exactly what this post is!

The biggest change we all had to face was the loss of both Diffuse and Spectral maps with the new PBR (Physically Based Rendering) system. The Diffuse map has been replaced with an Albedo map and the Spectral map has been replaced with a Roughness map, and honestly the new maps are not much different from the old ones. The thing I need to get use is the sensitivity of the Roughness map, as if it's pushed too far objects can look insanely shiny.

[1] Example of a Roughness map usage

The new map to come with PBR is a Metalness map, which is a simple black and white map. Black is non metal, white is metal but I personally think adding some grit into the metal could help portray a more realistic result, unless it is like a brand new car for example.

Our "Welcome back!" project was a group project where we were thrown into random groups. In these groups we had to work through a pipeline creating a small Viking dinning table scene, where we’d all eventually work on each others work. The Tri and texture budget was quite low making this a great first project. Working with random people was at first unnerving but by the end of the week things felt like they were fitting together within the group.

I modeled a hog's head, with a 600 Tri limit 

In the group, I had to texture this barrel, while someone else would texture the hog's head I created

At the end of the week we were also set our first full project, again a group task, but this time groups could be chosen by us which I think is a mistake, because within the first few hours all most everyone had grouped up back into their friendship groups. I personally think having random groups is a much better way to prepare us for the industry, as it’ll be very rare if we ever get to chose who we work with during our Junior years.

I’ll go into more detail about the project next week when we've actually started it and hopefully I will have updated the design of the blog too, I think it's time for a change!

Notes & References
[1]http://www.marmoset.co/toolbag/learn/pbr-practice

UDK: First Level Complete!

Yay! My first level in Unreal has been finished as part of a Summer project, I have been doing the level for a few hours a night over roughly 2 months. The most time consuming part has been learning about the UDK engine, figuring out how certain things are done like skydomes and fog while also trying to build the entire level out of modular pieces.

As I said in my previous post, the project was given to us at the end of our first year to do over the Summer, I have no idea if it's mandatory. I chosen to create a playable level of an abandoned subway, with a character (I haven't started the character yet!) within the given resource limits. We had 15k tris and 20mb worth of textures to use, at the beginning I had no idea how this related to a small playable level.

I started with a sketchbook to throw down my ideas, while also searching the internet for inspiration and reference. I knew I wanted to have this abandoned subway lived in, for a while at least, maybe an abandoned subway turned into a bomb shelter?

I also worked out my texture budget in my sketchbook

 Once I had a good idea about the level I tried to draw out the entire thing. I did this before I spent a vast amount of time going through the How to Draw [1] book, therefore some of the constructions are off. I definitely want to return to this drawing to see if I can improve it.

I now had the difficult task of figuring out how to build this level in UDK. My knowledge at this point with UDK was non-existent so it was going to be painful, I started to create the main pieces of the level in Max but had to restart them a few times because I wasn't thinking in small enough sections.

Eventually i got on a roll with this stuff and even managed to get my collision meshes working without much problem. Unfortunately I had no idea what a lightmap was, never even heard of one, and UDK kept spitting out these errors mentioning them. After some research I realized that I would need go back and do a lightmap for everything I had built, similar to doing the UV Maps only more tedious. Doing 20 lightmaps in a row was not a fun experience.

One of my level props, an old drinks machine

As I progressed with building the level plenty more little things kept popping up similar to the lightmaps, small things I had no idea about that my level would need to be even semi-competent. Creating the skydome was one of these, so I took this opportunity to create my own. The texture for the dome broke my texture limit, but in my opinion it was more important to learn than stick within the budget. 

One of the biggest things I picked up was how versatile the UDK material editor can be, by following this a set of tutorials [2] I managed to set up an animated skydome with a working sun, all using the material editor. I began using the material editor even more, creating working lights, controlling the specularity and other things.

It became complicated

I made sure to get the rest of my textures within the 20mb budget... just. I again thought it was more important to learn about making objects light up using the emission section of the material editor instead of staying within the limit. My final texture budget (without the skydome diffuse) was 20,120kb.

Staying within the given 15,000 triangle limit was slightly easier, but I was shocked at how quick that can be filled up with level geometry. I wanted to have many smaller props to fill the level, making it feel real and alive but I soon had to scale my vision down as the tri-count quickly piled up. If I restarted the level I'd scale my idea down to simply an abandoned subway, without the "lived-in shelter" thing going on, because a 15k tri-limit is tiny unless you're going to make a low-poly game.

The final product might come short of my original concept, but when I started I wasn't entirely sure how far 15,000 tris would go. I now know the tri-limit lends it self to maybe an isometric level where the low-polycount wouldn't be directly in front of the player like an FPS. Sadly I tried to create my level for an FPS, leaving the level feeling empty. Below are some screenshots of the level!

Despite me missing the mark on my original concept I think I have created some nice looking screenshots, especially with the added post-processing and fog effects, even if I found controlling them a little confusing. I wanted to add in dust particles and some sinister ground fog but this would have further increased my texture budget.

The candle light is animated! Woooaah!

I have now a much better grasp of the UDK engine thanks to this summer project, which will hopefully place me in a better position for Year 2 as we begin to work in-engine more. For the rest of the summer I plan to continue drawing and also begin the character portion of this project, while eagerly waiting for Destiny.

Notes & References
[1] How to Draw written by Scott Robertson & Thomas Bertling
[2] Tutorials created by Julio Juarex

Summer Break Update.

I'm almost half way through the summer break now and I've played a few games, done some drawing and messed around in UDK enough so that it doesn't scare me. I have also made some progress in the project that we we're given at the start of the summer, so I can finally show off some of that too! First up I'll go over what I've been playing and share my thoughts on the games.

Dark Souls II

I finally got around to playing Dark Souls II! The first game completely took me by surprise and I fell in love with it, so the sequel had a lot to live up to for me. I personally think the game just misses the mark of the first one but as a game by itself it is still overall fantastic, with the same great combat and exploration.

Wearing a dressing gown I stormed the castle... [1]

I think it's a number of minor details that made Dark Souls II  fall short. While the game can look breath taking (like the above screenshot) there are a few areas that look unsightly. The story this time around was even slower, taking around 30 hours for anything to really pick up, and with some of the areas and bosses feeling like a filler certain moments of the game became a real struggle to push through, not because of difficulty but because I was bored. Despite this, Dark Souls II is still more Dark Souls and that is never a bad thing!

Spec Ops: The Line

I decided to pick Spec Ops up in the Steam Summer Sale after hearing many people praise the story and they weren't wrong! A game that could often look like your average military shooter the story is great, with a superb ending to wrap it all up. Unfortunately the sections in between the story telling are your generic shooter.

To The Moon

Another game I grabbed during the sale due to the internet talk about the game's story. The most interaction you receive with game is exploring the environments in search for items that help you advance the story, so the entire game is basically leaning on the story aspect. Luckily the story was engaging and heart warming with some likable characters, but i can only recommend this game if you can look past the lack of actual gameplay.

Metal Gear Rising: Revengeance

Ever since Bayonetta I have been a huge fan of Platinum Games, therefore picking up Metal Gear Rising on the cheap was a smart move for me. I am only a few hours in but every boss fight has had me smiling and at times out-right laughing due to how ridiculous they are.  The combat is quick and brutal, but at times can feel button mashey, this doesn't bother me though, the game plays well and is an utter laugh.

The game looks great when caught mid-fight [1]

Drawing

I have spent a large chunk of my summer so far practicing my 2D skills. I've tried to keep a good balance between traditional and digital, studying and creative. Simply drawing for creative purposes is something still quite new to me, which probably sounds wierd.

An area I have been trying to improve upon is my draftsmanship and to help with that I bought a copy of Scott Robertson's How to Draw which spends a huge amount of time simply explaining how to draw in perspective correctly. The drawing below is probably the most interesting thing I've done from the book so far, but I already feel more confident in perspective.

Drawn with help from the book "How to Draw" [2]

Below are 2 drawings from imagination, using reference for the armor on the knights. Playing Dark Souls has had a clear influence on me, as I've been drawing plenty of knights and weird monsters.

The end results after a few hours of practicing drawing armor

An idea popped into my head, maybe a Dark Souls boss?

Along side all the traditional stuff I wanted to seriously increase my digital painting abilities, so I've made sure to spend a few hours a day painting in Photoshop and I think it has worked so far. Progress is slow but I can see small improvement with each painting I do. Below are a selection of paintings I've done.

Virtual plein air

Virtual plein air 

Virtual plein air

William Bougueraeu Master Study [3]

3D

The main summer project had us choose from a list of themes, and the themes were split into 2 categories. One category wanted us to create a character or vehicle set within a small scene, the other was to make an actual playable level with a character. Both of these had the usual resource restrictions.

The list of themes covered everything, from ancient ruins to desert canyons. I picked the abandoned subway theme because as soon as I read it I imagined a Metro 2033/The Last of Us style subway. Looking back I wish I chosen another theme as I already feel like I've done the survival thing to death, but I am really happy with everything I've done with it.

I will go into detail about the entire process once it's all finished but for now I'll just show a few WIP screenshots.

The above shots are very dark but that is because the lighting isn't final yet and there are many props to add to really bring the level to life but hopefully it gives the idea I am aiming for. Below I've added a screenshot with more light to help with viewing. Learning UDK has been really fun because it appears the engine has limitless options when it comes to adding things to make the level stand out, trying to think about how to build the level out of smaller modular pieces has also been interesting.

While I'd love to become a conceptual artist, working as a level builder inside the engine is something else that's interests me, it is like playing with virtual Legos and this is why I've been having fun with figuring out UDK. I without a doubt have a long way to go with both UDK and my 2D work but the longer I spend doing each the more confidence I'm gaining, when the level is complete I will do a more in-depth post about it.

Reference & Notes
[1] Both screenshots taken in-game

[2] How to Draw written by Scott Robertson & Thomas Bertling
[3] William Bouguereau - Moissoneuse Original

End of Year 1.

I have now finished the first year of the Game Art course, but before I disappear for the summer I have a few pieces of work I'd like to show of what I did in the final term. The first batch of images are some general traditional drawings from various tasks that I'm pretty certain I haven't shown yet.

Traditional Drawings

We went to Bradgate Park again and this time I wanted to do a final in pen, I don't know why but I prefer using a decent pen than a pencil, especially for quick sketching.

The final Visual Design project of the year was to return to the New Walk Museum and produce a 2 hour final. I decided to use pencil just in case I made a mistake, I didn't want to ruin a 2 hour drawing with no undo's! I did warm-up drawings before the final too. The final piece I feel is a lot more tonally interesting than my other drawings, I just wish I finished it.

Roughly 2 hour final

1920's Chicago

The last big project was to design either a character, vehicle or building from 1920's Chicago. This was a 2 week project so we had a good amount of time to try things out, which is exactly what I did. I wasn't sure what subject matter I was going to at first, so I researched a bit of everything until eventually I wanted to draw some 1920's people.

After doing the assortment of above characters I decided to make my finals about the moonshiners, an area of the 1920's that I think is actually quite interesting. The first painting I tried to paint straight into colour, which eventually led to me scrapping the image, I definitely need more practice using colour. I scaled back my next attempt at the digital painting and kept it values only.

First attempt

I also had time to do another painting of a moonshiner which was a bit more extra practice.

Master Study

As a quick final project for the year, we had to study and reproduce a master painting. I found doing the master study extremely helpful, in places it really showed how poor my colour perception is but this is something I can hopefully improve. I'm without a doubt going to continue to do master studies during the summer break.

[1]

I was planning on focusing entirely on my 2D skills during the break but I have just found out we've been given a few summer tasks, one is to read Mortal Engines [2] and another is create a small playable area with the usual constraints. Thankfully the brief is fairly open so I can have some fun with it and have already started to plan out the level and ideas around it, I will still be doing plenty of 2D practice though, both studying and just for fun. The drawing for fun is something I have only just got back into, but so far it's been refreshing with some satisfying results! I will post a few up eventually

Notes & References

[1] Black Bashi-Bazouk 1869-1869. Original painting by Jean Leon Gerome.

[2] Mortal Engines (2001) from the Mortal Engines Quartet series. Written by Philip Reeve.

Gladiators Ready!

I finally have finally got around to updating the blog with some 3D work. The latest project we had was to create a Roman Gladiator, this has been complete for a while now but I never had the files on hand to throw them into a post. I'm now back in Leicester and also have a spare afternoon so I can do a quick run down on the Gladiator!

The overall brief was fairly simple, to create a Roman Gladiator. The design of him was completely up to us as well, so this was our first real creative 3D project. As usual the project had resource limitations:

Construction

• Character Tri-limit: 2500
• Additional Props Tri-limit: 300
• Rigging bones limit: 28

Texturing

• 1x 1024 (Diffuse, Spec, Bump etc.) for the Gladiator
• 1x 512 (Diffuse, Spec, Bump etc.) for the Props

I done some research into Gladiators and quickly made up my mind on what sort of Character I wanted to create. I wanted my character to be almost like a slave gladiator, with very basic equipment. The image below from Gladiator shows the sort of thing I was aiming for.

A small and dusty Gladiatorial arena. [1] 

With an idea in mind I began researching equipment and armor types, it wasn't long until a made a quick sketch of my final character. Luckily we were given the required reference for the actual man, this meant I could begin building him straight away in 3DS Max, the armor on the other hand would need me to get creative with when it came to texturing, but I'll get into that more later.

Only half of the model was actually created, the other half was simply done with a symmetry modifier which helped speed up this stage. The tri-limit was constantly making me judge what details I could keep in or leave out, especially with some of the muscle definition but once I had a rough body constructed I started to model the head. I knew I wanted the majority of my resources focused towards the head, in most games the camera is at eye-level or above, therefore the head will be very prominent.

I now know I made a few bad choices when it came to tri placement. Once I had everything built and I knew I had few spare tris I placed them in stupid places, like under the pectorals, instead of important areas like around the main joints. This leads me onto something else that I wish I knew about during this project and that is topology, I had no idea that this was an entire other skill, like the best way for polygons to be positioned when joints are concerned. Topology is something I need to research more for my future 3D work!

Texturing the armor and props for the Gladiator meant I had to get creative with the different textures I needed, even with a local Roman museum I still had to think on my toes. One example is for the different leather, I used my friends leather jacket and with some Photoshop tweaking you can not really tell it's the same patch of leather repeated. The creases in the jacket also made great looking scares on the Gladiators skin. 

Many of the decorative studs I referenced from the local museums

The base of the shield was textured with a simple barrel top

I brought over many of the things I learned from previous projects and also tried to improve in areas that were lacking, while also picking up a decent number of new things that I need to improve on but as this was my first 3D character I am really happy with the results. Seeing the gladiator animate also put a huge smile on face, it was like bringing the model to life for a few seconds! Below are a few renders and a the 3D model viewer if you're interested in the final outcome.

Roman Gladiator by Braden May on Sketchfab

Reference & Notes

[1] Screengrab from Gladiator (2000), directed by Ridley Scott