Assassin's Creed: Reclamation – Revolver

A lecturer of mine who is an incredibly talented concept artist (http://davidpaget.co.uk/) had an idea for an Assassin's Creed game set in Victorian London, an idea which I think we can all agree would be pretty badass! Once of the concepts he came up with was a revolver...

Considering I'm taking a break from doing full environments a nice weapon would be a nice challenge, plus I've never modelled a revolver before... I don't know why considering they're pretty sweet to look at.

First thing I did was break down the concept art to try and figure out how best to build it...

I decided to build it in a variety of parts. I set myself a triangle count of 4k. I was rather surprised at how complex revolvers were I ended up trawling google images for blueprints and schematics on how they were put together which ended up being a huge help. I found from my initial concept breakdown was to simplified, revolvers are made up from a lot more parts!

Once I had the modelled built I set up the smoothing groups and unwrapped it. I then exploded out the model and detached the parts that had high volume of detail on them into individual parts and prepared them for sculpting in zBrush.

Modelling the detail in zbrush was a fun challenge I ended up looking at a load of filigree reference to try and get the correct look. I ended up drawing the pattern on with a mask then simply flipping the mask and using the inflate deformation tool to get the detail. It took a few tries but finally I came out with a result I was happy with.

Next came the baking of the normal map. The normal map turned out loads better than my Blood Bowl helmet. I think I'll be exploding all my models for baking from now on.

Once I had my normal map created a colour map and baked out an AO map. I then plugged them into the Quixel Suite and off I went texturing. I adore this tool!

It took a lot of trail and error to finally get a texture I was fairly happy with. The beauty of the Quixel Suite is that I could make changes suer quickly even totally scrapping the texture and starting again was an incredibly fast process. This was my first iteration which I felt look a little too... Evil?

Once I had a good enough base I then made some tweaks of my own. Even with the Quixel Suite it seems you can't beat the control of a good ol' wacom tablet. I overlayed some pattern on the barrel of the gun which turned out to be my favourite part of the model, I'm a huge fan of subtle details.

In the end I'm a lot happier with this model than the last. But looking at the concept art now and my final model It seems I may have lost sight of the original concept a little. I could have stayed more true to the colour pallet of the concept, I got a little carried away with the Quixel Suite it seems. Restraint, Christopher... Learn some!

Peace and love x

Blood Bowl Helmet!

I've decided to drop the doing the monthly noob challenge this month. Not because I wasn't enjoying it or anything but I think I'm trying to run before I can walk, so to speak. Rather than trying to do full completed scenes I should be trying to get individual assets looking go on their own first.

I've opted to do a few single assets and trying to do them fairly quickly. Any troubles I hit on each asset I'm going to make the focus on the next one. This time round I'm going to attempt a Blood Bowl Helmet.

First thing I did was gather my reference. I knew I wanted to put my own spin on this version of the helmet but I wanted to remove the seal and damage it up a little more with a tri count target of 2k tris...

Original concept 

Reference #1

Reference #1

I began by analyzing my reference, I noticed that general shape of the helmet was not far off a barbute helmet but with the face section open rather than a general “T” shape. I decided I would push that shape a little further in my version of the helmet.

I started by creating a general low poly, I then duplicated the low poly and added any supporting geometry to the helmet and turbo smoothed it out, I made sure the mesh stayed all quads. I then took this version into zBrush and sculpted any detail I needed onto the model. After completing that I exported it back out of zBrush and into 3dsMax. I first optimized the low poly mesh to get it as close to my set tri count as possible, I then unwrapped it and started baking and that's where things went wrong!

Initial models - low poly and zBrush ready (before symmetry)

High Poly Sculpt

High Poly Sculpt

High Poly Sculpt

Red Helmet is the model that was taken into zBrush before sculpting.

I had a lot of intersecting geometry on the model. This caused the bake for the normal and AO maps to come out incredible dirty with a lot of artifacts. At this point I should have fixed that, I should have separated the low poly and high poly and baked out there parts individually, but I pressed on with what I had hoping once it had some albedo and gloss on it it would save it (spoilers... It didn't!)

I then took it into the Quixel Suite and began texturing. I have to point out how much I love texturing with dDo. I love the process and idea behind using masks and layering different materials on top of one another. It's incredibly fast, and it produces lovely results and provides a solid flexible base for the artist to work from.

Unfortunately a lot of the dDo loveliness was ruined by the shoddy bake...

Final Result

looking at the final result I'm unhappy with the normal map applied the blocky silhouette. I feel I could fix both of these by better approaching the model. I should have better optimized my model to keep as much of the high poly silhouette as possible. I should have baked each part of the model individually and combined the normal maps into one after the bakes where complete. I could have also maybe retopologizsed the high poly to get my low poly rather than starting with one, this was I could have had more control over my topology and maybe allocated some around the bigger cracks on the model and they could have added to the models silhouette.

I think my next model will be all about improving my baking!Hopefully it'll turn out better.

Peace and love x 

Polycount Monthly Noob Environment Challenge (July 2014) - Blog Update #3

 Today I created 3 of the textures I planned to create, tillable ground and wall textures along with “Primary_Texture_1” which I plan to have the majority of the assets in the scene reference. This is my first time working with modular textures and using the Quixel Suite.

 Looking at my textures I can see some things I'm not 100% pleased with, they seem too bland to me. I'm unsure whether this is a positive thing or not, considering the context I plan to use them in. On one hand, they cant be overly detailed and unique because they wouldn't work across the scene as a whole as you would end up seeing specific repeating details, hence why I actively avoid adding too much unique detail to them. On the other hand I don't want the scene to look bland across the board! I'm going to carry on using these textures to see how they hold up. They may not look as bad in practice as they they do in my head, on the other hand they may look terrible.

 

“Primary_Texture_1” consist of three main elements light stone, stone similar too the ground texture and brass. Most of the assets in my scene will be referencing this texture. The grooves on the ground-esqe texture will be used on the stairs to break up the steps. Having not worked with modular textures before I'm unsure if this is a good way to go about things. Looking at it I can see problems with general blandness and not enough variation across assets. I'm hoping that I can use the UV's of the assets to full effect in a bid to create variation.

I'm in two mind about the lighting in the scene. Because of the colour intensity I'm loosing a lot of the albedo colour information in my textures but I rather like the bold, hot lighting in the scene, hopefully I'll get a better idea of what to do with it when I some textured assets into the scene.

As of not I have my worries about this environment but I treated this as a playground. For me it's trail and error to get used to using modular textures and practice with the Quixel Suite. Even if the scene comes out looking rubbish I'll be able to learn from it and do better on the next scene I have a shot at!

Peace and love x

Polycount Monthly Noob Environment Challenge (July 2014) - Blog Update #2

I started my a BSP blockout in UE4, using the 3^rd person preset provided in UE4 for scale reference. I knew I was going to use BSP for the walls and floors. Any other BSP in the scene is purely placeholder for static meshes, there function is purely to ascertain dimensions for when it comes to making the static meshes in 3dsMax.

For example the holes in the ground of the upper platforms are there so I know how big the grates will have to be to plug them. Although in the concept the gaps in the grate are painted in black. I imagine from a functional stand point these grates act as a way for water to drip back down to the lower levels in order to prevent the upper floor from becoming slippery. So when it comes to making the grates the player will be able to see down trough them from the upper floors and vice-versa.

I was fairly happy with the “eagle door” section (although after finding out it's not even an eagle above the door but rather falcon, or more specifically Ra! NOTE: read up on Egyptian culture!) I also felt the stairs may have looked a bit narrow but not so much as to break the scene.

The trim was my next challenge. The trim is made up of four main parts, a long, medium and short piece, with a block that is used both for decoration and to hide any seems in troublesome areas where z-fighting or lighting errors may occur (hopefully they wont mind!)

I'm not 100% pleased with the lighting as I feel it is too in your face with only one colour going on. This may be due to the colour of the lights being too saturated or I need more variation in the colour of the lights.

Next I'm going to create some tiling textures for the walls and floors. It'll be my first time working with PBR so I'm curious to see the results, I hear good things!

Peace and love x 

Polycount Monthly Noob Environment Challenge (July 2014) - Blog Update #1

I've decided to do this months Polycount noob challenge in a bid to improve my environment art. My end goal is to become an environment artist so this fits rather nicely. I've done environments in the past but looking back I've not been entirely pleased with them.

My goals for this challenge are to...

• Get better at planning!
• Work with PBR textures.
• Work with modular textures.

The first thing I did was break down the concept art provided into it's various elements to figure out what I would actually need to build to construct this scene. 

Normally from here I would have went straight into an in engine block out. But as I'm trying to work more on my planning I took this opportunity to create an asset list focusing on how long I though each asset would take to create and what textures I would have to create. As I also want to work with modular textures this time around I want to try and have as many assets as I can sharing the same texture, I tried to reference this in my asset list.

I've intentionally allocated extra time to the elements I feel need more time and attention them (i.e. the “Eagle_Wall_Door_SM” NOTE: I'm very aware of my that horrific name!) and any elements I'm not to sure about or will need some extra research into how to do properly, in this case it's the particle effects (as I'm not very experienced in working with particles) and the wall trims (I'm not sure how they'll fit together until I try in the blockout, they may need a different approach to be taken, we'll see).

Using the estimated times (which are totally subject to change, going from past experiences things always crop up that eat up more time!) I created a Gantt chart to plan my schedule for the rest of the month. If all goes to plan I should end up with around a day or two buffer time which I can use to fix errors, polish any areas or to try and squeeze out more efficiency from the scene.

This is what I have so far, apologies for the dryness and lack of actual art! But I'm hoping this will be a strong basis for the rest of the project. I guess time will tell.

Peace and love x

Hello!

Hello!

My name is Chris Reid and I'm currently a student at Teesside University in the North of England. I'm studying computer games art.

This blog is essentially a place for me to blog about any projects I'm working on. My goal for this blog is to have a place to store not only my work but allow me to look about and track my progress as I learn more and take down my thought process, basically it's my work diary. 

Hopefully I'll be able to keep this updated fairly frequently, lets see how I fare! 

Peace and love x