© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Allegorithmic, Epic Games, game, game development, gamedev, Games, games industry, indiedev, Maya, Substance, Substance Designer, Substance Painter, UE4, Unreal Engine 4, xNormal, ZBrush

Feed enhanced by Better Feed from Ozh

Clinton Crumpler from The Coalition (Gears of War 4) talked about the production of his amazing King Wash Laundromat scene. It’s an exciting environment experiment with beautiful lighting, hyper-realistic look. Clinton was kind enough to go into the smallest details and talked extensively about the pre-production stage, as well as about his search for colors, the creation of materials with Quixel and the general production in UE4.

Introduction

My name is Clinton Crumpler, and I am a Senior Environment Artist at The Coalition in Vancouver, BC, Canada. We have an amazing team and are almost done with Gears of War 4. My role consists of work on the look, development, and art creation of environmental aspects of “The Swarm,” the newest threat in the Gears of War universe.

Previously, I was an artist at Bethesda’s Battlecry Studios, KIXEYE, Army Game Studio, and various other independent studios and contract projects. My primary focus is environment art, shader development, and art direction. I have also produced a number of video tutorials with Digital Tutors on game art pipelines and production, and I am about to release a collaborative textbook with Sam’s Publishing focusing on game art development using the Unreal Engine.

As for my personal time, I enjoy designing scenes from specific times and places in history. It can be fun taking a seemingly mundane image or idea of a place and giving it character within a 3D environment. Crafting mood or telling a story through visuals is my favorite part of the environment artist’s production process. Working on personal projects also gives me the ability to test my artistic abilities across multiple disciplines of game development such as lighting, composition, and art direction. Below are a few samples of my past personal works:

Past works in Unreal 4

Additional samples of my work can be found on my website or my Artstation.

King Wash Laundromat

In the following topics I discuss the process that I use to create the King Wash Laundromat. I step through the general process I use for production and hit on a few key notes of technical or organizational aspects of my process.

Concept

When initially determining a location and concept, I stumbled across the image of “Laundromat at Night” This inspired me to delve deeper into the intricacies of a laundromat set in the late 80s to early 90s.

Lori Nix, Laundromat at Night, 2008, from the series The City .

This was the second time I have had the opportunity to work on this project. The first time was in 2011, while I was still learning about environment art production. I really wanted to create an interesting composition out of a seemingly mundane location, while adding a touch of color contrast to enhance the mood. I created this environment using UDK (Unreal Engine 3) and took a couple months in my spare time to create. While I was pleased with the result at the time, upon revisiting this work, I was happy to see how my talents have grown to produce the newest version of the scene. Below are some samples of the original 2011 project.

UDK King Wash Laundromat from Clinton Crumpler on Vimeo.

Original King Wash Laundromat Screenshots from UDK (Unreal 3) created in 2011

Mood and Visual Storytelling

Mood and storytelling is critical when creating a realistic, everyday scene. It can be very easy to simply emulate a real life location such as a bank, car wash, or a laundromat. But because the typical viewer sees these locations on a regular basis, the real challenge as an environment artist is telling a story, evoking a response, that is captivating and inspiring the viewer to look closer at the piece, and investigate the insinuated backstory to the piece. In order to achieve this, it is important to establish the aesthetics of your project early on and understand the mood, tone, and setting you are trying to establish.

While reworking the 2011 project, some of my original ideas changed in terms of what I wanted to achieve in the visuals, as well as its influences. I wanted to create a similar nighttime atmosphere, but add a bit of an eerie and lonely feeling. I wanted to instill that this laundromat was in a more isolated dark area instead of a city. I also still wanted to retain some of the varied colored lights from the original scene. In looking for references, I started to draw more on nighttime photography of isolated nighttime buildings and varying color palettes of nighttime scenes. I really liked the visuals of the softer peach and pink against the moody bluish green cool colors set in the background.

Below are some of the photos and films I found influential in determining the style and look development of the scene:

Meet Me in the Morning: Photos by Henrik Knudsen

The Ghost Writer, directed by Roman Polanski.

Fight Club

Photography by Paul Nevison

Belmont Eco Laundry

Referencing these images, I could better understand the tone of the scene and my general lighting scenario that makes each image visually compelling. I could also begin to visualize how to lay out various cameras, lights, and points of interest.

Additionally, I often find visual influence from the games I play. Examining density of scene assets and visual construction of a scene in today’s games provides context for what other developers and artists are producing. With this knowledge, I can judge what level of visual capacity my scenes need in order to meet the latest standards of game art direction.

Some of the most influential games while working on this project in particular were The Division, Batman: Arkham Knight, Dying Light, Rage, and obviously the Gears of War Franchise. All of these games have a good sense of color vibrance and saturation, and all generally have a good contrast range. They all use filmic color grading and visual techniques to make more a more movie-esque visual style. The easiest way to understand and emulate a painting, movie, or game is to take a single frame or still and break it down to its simplest colors. It’s easiest to think of this as a kind of squint test. If I squint or blur my vision and look at the image, I can learn a lot about the palette by asking a few key questions.:

• What colors are easiest seen?
• Where is the focus of the image and how do the colors draw your eye?
• What colors are in proximity to each other?
• Where are the resting places or darker dull colors within the color palette?

Examining the palette can help establish what kind of color style and palette early on in a project.

Below is a screenshot from Batman: Arkham Knight

Batman: Arkham Knight

Below is a blurred perspective of the same image. Notice the color palette becomes more visible. Isolating the color palette to core colors (between 4-7) makes for a strong composition.

Batman: Arkham Knight screenshot blurred to primary colors clearly

Batman: Arkham Knight screenshot color palette 

Rage

Rage screenshot color palette 

The Division

The Division screenshot color palette

After examining these games and looking at some of my influential images I landed on this color range for my final outside palette:

King Wash Laundromat Color Palette

I find that while choosing the visual style and language of the scene, it is excessively important to gather as much reference of color palette, composition, style, and asset components. Not only is online reference gathering important, but I love to go out on foot and find real life references. Taking influence from where I live and work every day makes for easier image gathering and the opportunity to get out and actually measure the real world equivalent to the assets I’m making. Establishing the correct scale and real world use of an asset can really help to make or break a scene and help settle all of the individual assets into a scene together. Sometimes the internet just doesn’t cut it and I find myself questioning, “Is this asset the right size?” It may seem strange at first, but understanding real world spatial distances and correlations by taking a ruler out and measuring a door, a window, or a shelf helps my projects and gives me a keener eye for 3D scene creation. Below are a few images I gathered around town to get real world reference for the laundromat.

Gathering Real World Reference Photos

Gathering as much reference online as possible to fill in the missing gaps allows me to get more creative and combine ideas found in different images. Taking the idea of how the window decals look in one scene and adding it to how the door looks in another helps me culminate my favorite aspects of each image and reference into my project. Below are some random images that helped me establish the final look of the laundromat.

Gathering Online Reference Photos

For collecting even more ideas, I find that using Pinterest can be incredibly helpful. When I tell people this I mostly hear a lot of “Isn’t that for like wedding and cooking ideas?” While this is generally true, there are also an increasing number of images that can be used for reference from photography, film, games, and other inspiring visual media. I also use Pinterest to upload my own images and reference I find anywhere online to collectively bunch my thoughts into one place that I can access from anywhere at anytime online.

My Laundromat Pinterest Board

Goals of the Project

While beginning any project it is important to set goals for yourself.
I had a few goals in mind to improve my own artistic ability and process.

• Better comprehension of lighting and understanding of types of lighting (static, moveable, stationary)
• Deeper investigation into post process and Look Up Tables (LUTs)
• Balancing large and small props while keeping player or camera perspective in mind
• Creating mood and setting visual tone

Project Management

Time management and understanding how to pace yourself in a project is hugely important while working alone or in a small team. It takes a lot of self motivation and can help to break down each part of development into smaller, more easily digestible chunks of time and effort. Also as a game artist it is vital to not work on in a vacuum and to be observant of new techniques or construction pipelines used by other artists to achieve the best visual results. At the same time, I have to balance how much time to invest with each individual asset compared to looking at big picture. It can be alluring to spend hours perfectly tweaking a small asset or prop and later find that it is almost completely cast in shadow in the end scene. Using time-saving measures can help to ensure overall project success while helping maintain a steady and productive workflow.

Trello

Trello is a major part of my production process while working on solo or small group projects. It holds me accountable and is a cheap system for moving forward in a project and tracking progress. While starting out I can easily separate the project into individual tasks or parts. This helps me take production one step at a time. It is also a very helpful reminder tool, documenting bugs and keeping track of where I left off on certain tasks. I typically will use it to show the progress made on certain aspects of the visuals. I may make a list for models and meshes, then another for textures, and then move the “card” per asset from one list to another as I complete different parts of the process. Best of all, it’s persistent online so you can access it from anywhere at anytime.

Trello Board Setup

Production

Tools

My primary tools for this project were:

• Maya and Zbrush for modeling
• Quixel and Photoshop for texturing
• Mightbake for texture baking
• UE4 for game engine

Blockout

When first blocking out the scene, I tried to keep the pieces as modular as possible. At this point in the process it’s best to not force yourself into a corner; I prefer trying multiple placement scenarios of assets and proxy meshes to try and get the most satisfaction overall. Using a quick modular kit can keep the ideas flowing and easily tested. Spending time working out the kinks of the design at this stage can definitely save time for the remainder of the project.

A quick kit of modular pieces used to establish the initial blockout

Using a good scaled model of a person can be another strategy for beginning to block out and understand scale and dimensions. This can help to confirm proper scale measurements in the scene. I typically scale the man to my size but a general rule of thumb is to make your person reference around 5′ 11″ in cm or 180.34 cm.

Scaled reference person asset in scene

Assets

After creating a basic asset list using Trello, I started to work on the assets from biggest or most dramatic impact on the scene to the least. A common pitfall is getting to the end of the project and realizing that some of the props or assets you originally set out to make, or made proxies for, are not as important or are outright unnecessary. This is why it is important to create the key pieces first and work down to the less important aspects of the scene. Large assets are also good to check for scaling issues or visual sightlines in the scene so you can make the necessary changes early on in the project.

Laundromat top load washer asset

A few of the assets used in the scene

Typically when creating assets, I work with quick proxy models first to populate the scene and then replace them over the course of the project with the fully developed asset. I typically will apply a brightly colored or simple color fill on the proxy meshes. These placeholder meshes can keep the focus on what assets are missing and serve as reminders for the overall scale of the project still to be completed.

Emergency Light Prop

Lastly, remembering the scale and player perspective is very important for creating assets. If the camera and player are always in 3rd person like in a game like Gears of War or The Division it is important to get a good overall visual read from materials, wear, and silhouettes on assets from a mid level view. Meaning some aspects how you create the asset will need to be exaggerated or slightly enlarged to get a good overall visual for the player. While a first person game such as Dying Light or Call of Duty allows the camera and player to get much closer to the environment and therefore development of the assets needs to be a bit more on the micro level. This is not to say the quality of either is adjusted overall, but instead understanding the distribution of texture density to best suite the type of gameplay perspective for the most optimized game visuals and experience. For the laundromat, I took a third person visual route and placed emphasis on some of the wear and tear and roughness values throughout the scene.

 The Division – Third Person Perspective

Dying Light – First Person Perspective

Materials

Most of my materials in the scene use one of two master materials. The two primary materials are a generic prop master material and a generic tiling floor, wall, ceiling type material.

Prop Master Material

Prop Master Material Set Up

The prop master material had a few unique functions to make the process of importing new assets and setup much easier. To name of few of the material’s functions and what parts they played:

• RMA Texture packaging – Using the individual channel of a single texture to place my Roughness, Metalness, Ambient Occlusion, and sometimes using the Alpha channel for emissive or subsurface. This allowed me to use less texture memory with only one texture import versus three. I can also easily manage all my textures in one Photoshop file. When using this method, I make sure to change my compression settings appropriately for the texture in UE4.

Channel Packed RMA Texture

• Tuneable roughness and tint separated by whether the area of the mesh is labeled as metallic or nonmetallic. I did this using an If node in the material editor. I controlled the roughness or tint pending whether the material’s metalness was above or below .5 (since most all nonmetals are well below .5 and most metals are above .5).
• Ability to input specific emissive texture, use the alpha of the diffuse as emissive, and control the emissive color injected into the light propagation volume separate of the final emissive glow output. This is done through the lightmass replace node in UE4. I can set the amount of emissive influence injected into the light propagation volume separate the emissive shown in editor. This is really useful if you have a bright light that you want to overdrive the emissive for to get a good bloom, but you don’t want to overpower the lighting within the scene.

Lightmass Replacement Material Setup

Tiling Master Material

For this project I ended up creating and using about 9-10 tiling materials using mostly the same master material. Each would sometimes have two texture sets for a damaged and a non-damaged version. These two versions would be blended in the scene using vertex paint to give the final result. Some other switchable features were a tiling dirt overlay and a water/puddle material that were also both paintable by vertex color.

Material Vertex Transitions

Basic Tiling Materials

Quixel Suite

Working inside of Quixel Suite on the vending machine

Breakdown of Textures used to compose final television asset

While creating assets for the scene, my primary texturing tool was the Quixel suite. Using ddo allowed me to work through texturing the assets very quickly, while maintaining good quality. Often I can find myself noodling too much with each little part of texturing an asset, especially larger or more complex ones. Before I textured each asset, I used a material mask that I rendered out of Mightybake that labeled each type of material on the asset with a color. Then using the mask, I found that I was quickly able to take on each section of the mesh one at a time and keep a clear vision of moving on to each part while texturing. At the end, I would go back and add smaller detail touches and overlays to get the final material result.

Laundromat Washing Machine rendered in Marmoset

Blueprints

As I use the UE4 engine more and more, there are more cases where a blueprint will make life much easier and will quicken production speed. I know very little about coding or programming, but knowing some basics can help make simple, yet very useful blueprints. While developing a scene, it’s vital to keep in mind that the composition, placement of assets, or lighting may change during its evolution. The better assets are being managed, the easier time you will have as an artist. Even just grouping asset with blueprints can be a big help.

Some of the first assets I remade from the former scene were the laundromat machines (washers and dryers), each of which had doors attached. While placing these assets in the scene I found it quickly frustrating to continue to place a door on each mesh and rotate it to a degree that looked best. In order to combat this I created a blueprint to set a random rotation on the door every time the blueprint is placed in the scene. Then after I tested a few versions of how far it opened, I clamped the number to make sure the door would never open enough to clip into the object next to or in front of it. I also set the random angle that the door was open to favor being closed or nearly closed. This allowed me to quickly move the asset in the scene until I was happy where it was placed and then not have to worry about the accompanying door attachment.

Dryer Door Blueprint Setup

Blueprint in action

Another blueprint that I set up for this scene was the overhead lights. I knew I would be using these lights throughout the scene and didn’t want to have to set up the lighting scenario each time. Also if I decide to change the lighting I can simply do it through the blueprint rather than a per light scenario, saving me lots of time and allowing for quicker iteration and testing.

Lighting

Lighting has always been a challenge to me, as it is for many other environment artists. It is really about trying to understand the process and how each light behaves and how it plays with the surrounding materials. Even with the perfect setting of materials, lighting them properly can make or break a scene.

In this scene I experimented with a balance between realistic and moody lighting. When I first placed more realistic light setups in the scene, it had little contrast and overall just felt too well-lit for a kind of eery dark nighttime scene. Then when testing with fewer lights, the scene became too dark, which can be an even worse scenario for an environment artist–to make a ton of assets and them place them all in a dark room. To get a better idea of my lighting setup, I turned the saturation within the post processing volume to 0 to get just a grayscale image. This helped to visualize the balance of lights and darks within the scene.

King Wash Laundromat – Lighting Only

After trying a few lighting scenarios I found a balanced middle ground by using a spot light with a slightly narrower cone with a soft falloff on the sides. Then to emphasis the feeling of concentrated light from the bulb, I added some tube-shaped point lights over each of the bulbs in the ceiling lights. Lastly, in order to add a bit of mood I added a cone-shaped mesh that used a sort of cloudy god ray look. To quickly speed up the process of testing all these lights together and how they would interact, I created a small side scene to test baking the lighting on simple geometry. I also ended up using this room to test materials under different lighting conditions.

Vanilla Lighting test scene

One last thing to keep in mind with lighting is that sometimes hand placing lights can make a huge difference in creating a more visually appealing finished product. Sometimes a single overhead light won’t cut it and can make the scene a bit flat and dull, and by placing a few point or spot lights subjectively around the scene I can help to frame specific assets and points of interest in the final scene. I like taking a few artistic freedoms when it comes with lighting as I find it leads to a more compelling end visual.

Hand placed smaller lights can bring more interest to specific parts of your scene and highlight some assets

Post Processing

Post processing is one of the most rewarding parts of working in the Unreal engine. It really has the potential to turn a good scene into a fantastic scene with simple tweaks within the editor. Post processing in UE4 has a lot of special functions and the ability to add some extra tweaks on your own through blendables. I ended up using a lot of the film and bloom options to achieve the final, more cinematic result. As an additional blendable, I added a simple function to slightly sharpen the final image as well.

Unreal’s many editable post process options 

King Wash Laundromat’s post film settings

One of the most influential parts of post processing while working on the laundromat was the use of color grading or Look Up Tables (LUTs). LUTs are used to take the real time rendered image and change how each pixel of color is represented in the final rendered game. Using a lookup table, I can skew or alter the brightness, saturation, contrast, or specific color of any one particular pixel of color in the final scene.

While choosing the LUT of the laundromat scene I referenced some of the games I discussed earlier, as well as look at some influential films to see how they were using color to achieve the movie-esque quality they all shared. Also by focusing on the color palette I discussed earlier, I was able to narrow down my overall look development. With a few small tweaks you can see how dramatically the scene can shift in overall visual quality and contrast adding a nice finished touch to the final product.

King Wash Laundromat with no post processing – Notice the difference between the use of the LUT compared to the final scene

Results

Here is the final screens from the King Wash Laundromat Project:

 Closing

In closing I really enjoyed revisiting old assets and seeing how I have evolved as an artist–not only in my comprehension of visual language and style but also technically. Looking back and reflecting at old work can be a powerful method to see personal growth and understand areas that you should continue to improve upon.

For more of my work, visit my portfolio or my artist Facebook page for what I am working next. This project can be found on Artstation. I am also currently working on videos to provide more thorough breakdowns of the technical methods I used to create this scene. You can purchase this UE4 scene to take a look for yourself on my Gumroad and Cubebrush. It will also be available directly on the unreal marketplace in the next couple weeks.

Thanks for reading!

Clinton Crumpler, Senior Environment Artist

INTERESTING LINKS

LIGHTING COMPARISON IN UE4 AND CRYENGINE

STUDYING LIGHTING & MATERIALS IN UNREAL ENGINE 4

LIGHTING SCENES: REALISTIC LIGHTING FOR AUTODESK MAYA AND MODO

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Adobe Photoshop, batman arkham knight, division, Epic Games, fight club, game development, gamedev, Games, games industry, indiedev, Marmoset, Maya, Quixel, Quixel Suite, UE4, Unreal Engine 4, ZBrush

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: contest, E3, Epic Games, game development, gamedev, gamesbeat, indiedev, UE4, Unreal Engine, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

Anya Elvidge creates some amazing 3d dioramas for her projects. We’ve discussed the production of these pieces: modeling, details, materials, ligthing and post-production. Her style is absolutely amazing.

Introduction

I’m Anya Elvidge, or Anya the Artist as I tend to market myself. I’m a student, though I’m set to graduate in July when I will be finally let loose to find work in the game industry. I live in Leicester, UK, where I have just reached the end of my studies in Game Art Design at De Montfort University.

I’ve only been making game art for 3 years as part of my degree, but I quickly began to realise that my passion was with environment art, specifically creating little worlds in the form of dioramas in UE4. The first diorama I created was at the end of my second year, just under a year ago, and it really kindled my passion for environment art. Since then I’ve been focusing on improving my portfolio so I can get involved in some awesome projects.

Most recently, I’ve been working on my Final Major Project for university.

It was a 20 week long project, and I chose to create a set of 3 hand painted stylised dioramas in UE4 based on the writing of fantasy author Robin Hobb, who was kind enough to give me permission to use her IP. Each diorama is playable, and has various sounds and animations to add to the atmosphere I’m going for. In fact, you can download and play my ‘game’ here if you want.

Dioramas

I actually just recently did a little lecture to my course mates about my dioramas, and this was one of the big topics I focused on. Dioramas have such a huge number of benefits, particularly to those just starting out in environment art, where perhaps a team isn’t available for you to work in, or you just don’t want to spend months working on one scene.

The biggest benefit to working on dioramas is how safe they are, and yet they are still very interesting to the viewer. You have very little to consider outside the assets you are creating and how you will present them the most beautifully. You don’t have to deal with much in the way of coding, or creating immense vistas, or composing an area from multiple angles as you move through and around it. Additionally, it’s very easy to plan your time for something so small, and for school projects this can greatly reduce the stress of planning a project and trying to meet a strict deadline than with full environments.

I also like dioramas because they’re just so concentrated. They’re small areas of a high quality and you are completely in control of every little aspect of how it will look at the end. You can get away with not using huge tiling sheets to create an area, using lots of unique models and assets instead to give a scene loads of character. And really shows off your skills. You can be much more creative with your colour choices and compositions, too, because you don’t have to deal with how it will affect a larger area. Certainly dioramas are much easier to compose than larger scenes and you can get really creative.

If you’re going to make a diorama, I definitely think you should go the whole hog and make it really striking and unusual.

Production

Every time I create a diorama, I start with 3D blockouts. I jump from 3dsMax to UE4 to see how a new potential blockout feels from distance and up-close. If the composition feels off, I’ll go back and forth making various sized changes until I’m happy. It takes a lot of iteration;

As I do this I’ll think about exactly what I want to be in my diorama, from assets to sound effects to what the atmosphere and colours will be. I make an extensive list of things, even silly little assets or sounds, and I can reconsider that list later and knock off things I don’t think will work. It’s easier to do this if you already have a strong idea in your head of what you want to achieve, or if you’re working from an existing description like I was, but I still think this is a good way of doing things. Pinterest is great for inspiration too, and I make a board for each of my dioramas.

Next I’ll start overpainting and photobashing my chosen blockout in Photoshop. Again, I had a strong idea in my head of what I wanted to do, but if you don’t then lots of quick iterative doodles and sketches can help an idea form.

Once I had a really simple room design, which had no furniture, I used a lighten layer with some artwork I liked to lay some abstract colour ideas onto my scene. It was a really creative and fun way of coming up with ideas I wouldn’t usually consider.

Next I revisited the list I had made earlier of asset ideas, and decided what would or wouldn’t work well in the scene and started to plan out where things would go and what they would look like. Then I just did some more photobashing and overpainting to try and get across a bit of atmosphere, using the colours from earlier and carefully considering the composition as I went.

For me, it’s important to be organised and know exactly what I’ll be dealing with in the future, especially when working to deadlines. I made an extensive plan of exactly what I would be making, such as if I would need a tilable texture for this or an animation for that, and input it into a spreadsheet.

At that point I’m ready to start modelling!

Sculpting

The vast majority of my work is made without Zbrush. I work in quite an old-school way of creating the albedo in Photoshop, and then using the albedo to also create the roughness and then a heightmap, which in turn is used to generate a normal map. Anything I sculpt in Zbrush is always started from a base mesh in 3dsMax, as I feel I have more control over elements there.

I’m not much of a sculptor- it’s something I really want to improve- however if an asset really calls for sculpting I have a really simple workflow involving as much 3dsMax as possible as that’s my comfort zone.

My brick wall tilable was sculpted in Zbrush, as was the floor, fireplace, and roof beams, but that’s it. I’ll always use Zbrush for things that need more uniqueness and depth beyond what I’m capable of hand painting and modelling in. Zbrush is my software of choice for more organic things such as stone and wood, however, I’ll always use a 3dsMax hard-surface workflow if I can get away with it. Oftentimes I’ll make the entire model in Max, and then just use Zbrush to beat it up round the edges and add unique details. I used this method for the lead on the windows, for instance.

Details

I think having such detail in scenes is important to bring it to life and really immerse the viewer. I love taking the time to lovingly add tiny details to my dioramas, though I still think there’s a level of detail I’m yet to reach in my work and am still striving for! Because the scenes are so small, and they’re going to be scrutinised so closely, I think it’s important to use that to your advantage and make the scene super interesting.

At the end of the day, showing an ability to pay attention to the details while still maintaining a nice overall feel to a scene is, at least to me, really important. I think it’ll impress people looking at my portfolio and hopefully get me an awesome job in the end!

Material Production Process

When I created materials for my final year project dioramas, I worked in a sort-of semi stylised way using a combination of hand painted textures and PBR techniques. I use the standard PBR maps; metalness, roughness, albedo, and normal, but the maps themselves are hand-painted or use a combination of painting and overlaying maps generated from bakes. I try to simplify surface properties of materials, so that they don’t appear too noisy and rough, for example with the wall tilable texture and even the curtains. It gives the overall scene the softness I like.

I experimented with having designs on the curtains and other soft furnishings, but to maintain such a soft appearance it seemed best to just use gradients and abstract designs. Sometimes over-simplification can go a long way in a more complex scene.

Softwares I tend to use are Photoshop for hand-painting the textures, and if I’ve used Zbrush for sculpting, I’ll bake my maps in Topogun and overlay them in Photoshop using various layer modes. I’ll create my roughness map and height in Photoshop too by modifying a greyscale version of my albedo. I use CrazyBump for generating a normal map if I haven’t sculpted.

Unlike standard PBR, where the albedo has no shadow information baked, I have shadow information in the albedo much like an old-style diffuse map.

I like the depth it gives my textures and it gives the scene a more unusual feel. Sometimes it’s difficult to keep the materials out of the realm of uncanny valley, but in most cases it works pretty successfully. The main problems I have is when I’m trying to work with metal. The changes in hue in the albedo can have some strange effects sometimes.

I’d mostly say that the success of my work doesn’t come from technical knowledge- I’m not very good at the fundamentals of game art such as material creation, modelling and sculpting. I can use my more traditional art skills and knowledge to make my work as appealing (at least, at a distance!) as possible without being the best at creating good models etc.

I am, however, looking to properly learn Substance Designer and Painter, and other softwares too, in future projects so I can bring my technical skills up to scratch.

Using UE4

In UE4 I always set up a very rigid file structure to follow. That’s the first thing I do;

When I create my UE4 scenes I always work from a blank scene with a sky sphere, atmospheric fog, a skylight, a directional light, and atmospheric fog. I’ll put a post processing volume in there too and make sure the auto exposure is off. This starts me with a good, standard base for setting up a diorama which I can play with later. It’s important to consider that the skylight, atmospheric fog, and skysphere have a massive effect on the overall colouration of your scene.

As you go through adding assets through your scene and you want to make changes to, say, the colour of the sky, the whole scene will change colour with it. To get round this I create a large sphere in 3dsMax to act as a second ‘sky’. I apply an emissive material of whatever colour I want, or I create a gradient in Photoshop, and I turn off ‘cast shadows’ on the mesh. That way I can change the sky colour totally independently of the lighting, which is especially useful in stylised art.

Even though I do have an atmospheric fog asset in my scene, I tend to ignore it and make my own fog using particle effects and meshes. For example the snow haze in my first diorama, the light beams in the Room, and the light beams and slight haze near the tree in the Dragon Garden. Again, it affords me more control and means I don’t have to deal with other things changing as a result of tweaking settings.

I barely touch the post processing until later in development, as I don’t want to become reliant on it. You can also completely confuse yourself with your lighting if you start playing about with things like global illumination and then forget you tweaked the settings. It’s good to occasionally turn off your post process to see if you’ve gone too far astray from the actual scene. Shadow tint, ambient occlusion, and global illumination are some of my favourite settings to tweak.

Lighting Setup

Lighting is super important, and in my most recent projects I definitely feel like I’ve learned a lot more through talking to guys in the industry. For each of my final year project dioramas, I start with a simple light setup of a skylight and a directional light, and build on that. Like I said, atmospheric fog and sky sphere colour also have a big effect on lighting, so they’re important to consider too.

The lighting setup for each of my dioramas is very different, so I’ll briefly run through each one.

The Fool’s Tent

Here, I have a very dim directional light (2) and quite a bright skylight (5), and then the rest of the lighting, such as for the tent, moonlight, and lanterns, is done with pointlights. I also have a setup where the Light_Ext lights turn off as you enter the tent, so that the dim lighting inside the tent is not interfered with by the bright exterior lighting, such as the light used to illuminate the tent.

I’ve learned not to be afraid of using point lights to aid my lighting, however you have to be careful because they are a big hit on framerates and optimisation. You couldn’t use as many lights as I have in a big environment.

The Fool’s Room

In this scene I have a very bright directional light (80) coming through the windows, and then a less bright directional light (25) coming in from the left to illuminate the front of the diorama. I have a skylight set to 1.5 make the scene feel softer, and to soften it even further a single very bright point light in the centre of the scene to illuminate the underside of the beams and brighten the walls. Bloom and light beam meshes are used in conjunction with the lighting to give the scene an airy feel.

The Dragon Garden

This diorama looks like chaos behind the scenes. Those floating trees you see are set to be hidden in game, but by ticking ‘cast hidden shadows’ they also cast shadows despite being invisible. I use this to cast additional tree shadows on my diorama to darken areas and draw attention to others. It’s a nice way of really controlling where the player looks in a scene.

In this diorama I have a bright directional light coming from the right, and then a second one coming from behind so that the tree canopy is more illuminated and to get some nice orange sub surface scattering showing on the dragon wings. It also lightens the overall scene a little. Point lights are used to brighten areas that are too dark, and to pick out focal points such as the dragon’s face.

Advice

Firstly I would absolutely recommend talking to the pros already in industry. Critique is so very important, and they can show you how to do things that might seem a little confusing or scary at first. If you see some work you really love, definitely don’t be afraid to ask how they went about it!

I have a few tips regarding creating successful dioramas… They’re not the end-all be-all of successful diorama creation, but they’re a good start.

• Plan everything. Like I showed earlier, it’s important to know exactly what you’re going to deal with and be ready for it.
• Really think about the placement of everything. Might seem obvious, but it takes a lot of thinking about even the tiniest of things to create a successful diorama. The placement of a single group of flowers might just slightly tip the balance of your diorama and be more of a distraction than help.
• Treat it like a character. Rather than thinking of a diorama as a platform for showing off some high-detail assets, create it as you would a character, where everything is of equal importance and detail. Everything contributes towards an overall final product, rather than fading into the background to show something more important off.
• The platform is as important as contents. Don’t just go down the ‘house on a disk’ route, because it’s boring. Make the platform be as much part of the scene as the assets in it. My favourite way to do this is making a ‘tearaway’, where it literally looks like you grabbed a chunk of a larger environment; bricks are falling away, things have smashed, tree roots are coming out of it, etc. You could get really creative though, I’m sure.
• Give it multiple levels. The most successful dioramas I see have multiple levels to them. This could be achieved with a stairway, a bridge, a treehouse. Anything, really.
• Don’t terminate visual elements too abruptly. Don’t just end the edges of your diorama with no thought. Have water falling off it, or sand, or cloth. Let the bricks of a wall break away gradually.
• Atmosphere is still really important. It’s easy to fall into a trap of not having any atmosphere on an area so small, but don’t be afraid to add fog, light beams, particles, and interesting lighting to your scene.
• Keep it balanced. Again, like you would with a character- if you drew a line down the centre, does it look like it’s going to fall over one way or another? If you have a tree hanging haphazardly off one side of your diorama, make sure to balance it out or it’ll be difficult to take nice renders that don’t feel off to the viewer.

I hope they help people make nice dioramas.

Anya Jo Elvidge, Student, Environment Artist, www.anyatheartist.com

INTERESTING LINKS

SIMPLE AND RICH LANDSCAPE BUILDING IN UE4

MOODY SCI-FI ENVIRONMENT CREATION

BE VISIBLE: HOW TO GET A JOB IN GAME INDUSTRY?

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3DS Max, Epic Games, game art, Game Design, game development, gamedev, indiedev, Photoshop, UE4, Unreal Engine 4, ZBrush

Feed enhanced by Better Feed from Ozh

As always, weekly news in our digest for your convenience! This week was full of surprises with E3 just around the corner.

4.12.2 Hotfix for UE4 Released

The 4.12.2 Hotfix has been released. This hotfix resolves a critical crash when refreshing blueprint dependencies on a blueprint with no class, as well as fixing 10 other important issues.

Please be aware that after updating to this hotfix, it will be necessary to recompile shaders on your projects. More details here.

DESTINATIONS WORKSHOP TOOLS – VR CONTENT CREATION PACKAGE

Destinations Workshop Tools is a VR content creation package that enables the creation and sharing of both real and imaginary worlds on the Steam Workshop. Destinations Workshop Tools lets you create, share, and explore both real and imaginary places in virtual reality. This is a beta release of a set of content creation tools that users can use to construct different worlds and share them on the Destinations Workshop. New software is available in early access. More here.

LUMBERYARD 1.3 BETA TO INCLUDE SUPPORT FOR VR GAME DEVELOPMENT AND NEW VR DEVICES

Lumberyard Beta 1.3, which Amazon is releasing in a few weeks, includes support for VR game development and new VR devices. VR implementation uses Lumberyard’s Gems, which are self-contained packages of assets and features that developers can drop into games. Lumberyard Beta 1.3 includes Gems for both the Oculus Rift and the HTC Vive. Moreover, these Gems serve as templates for users to build their own Gems and support any VR devices they want. Details here.

HORIZON ZERO DAWN: ALOY’S JOURNEY TRAILER IS AMAZING

Right before E3 2016 (probably to avoid leaks), Sony and Guerrilla Games launched a new trailer of their upcoming game Horizon Zero Dawn. This video is focused on the lead character Aloy. The game looks amazing!

2016 INTO THE PIXEL JURY PANEL ANNOUNCES VIDEO GAME ART COLLECTION

The Entertainment Software Association (ESA) and the Academy of Interactive Arts & Sciences (AIAS)  announced the official selections for the 2016 Into the Pixel (ITP, #IntoThePixel) collection. Established in 2004, the annual ITP art exhibit honors video game artists who continue to push the interactive entertainment art form forward. Co-produced by the ESA and AIAS, the 2015 ITP collection will be displayed during E3, the world’s premier trade show for computer, video, and mobile games, from June 14-16, 2016, at the Los Angeles Convention Center. All selections here.

DESTINY: RISE OF IRON – NEW EXPANSION ANNOUNCED

Destiny: Rise of Iron is the next expansion to the Destiny universe. It features an all new cinematic story campaign set within The Plaguelands, a brand new location on Earth along with new weapons, armor, gear along and a six-player cooperative raid.

WATCH DOGS 2: GOING TO SAN FRANCISCO

A couple of days before E3 2016 Ubisoft has announced the next part of the Watch Dogs series. It features a big detailed world based on San Francisco area. Click here for more swag.

Call Of Cthulhu – E3 Trailer

Focus Home Interactive presentedthe E3 trailer for Call of Cthulhu, the official video game adaptation of Chaosium’s pen & paper RPG. Developed for consoles and PC by Cyanide Studio in Paris, Call of Cthulhu will offer an RPG-Investigation title, packed with psychological horror and stealth mechanics. Cyanide Studios brings to life a disturbing world that players will have to explore and investigate while Cthulhu, the Great Dreamer, prepares its awakening.

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Destiny, Epic Games, game development, game news, gamedev, games industry, Horizon Zero Dawn, itp, lumberyard, Steam, Ubisoft, Unreal Engine 4, VR, Watch Dogs 2

Feed enhanced by Better Feed from Ozh

Warren Marshall is one of the few developers out there, who always tries to help beginners to find their way around gamedev. He publishes articles, uploads great talks on his YouTube channel (we encourage you to subscribe – there’s a ton of stuff for artists there) and does a lot of educational work. We’ve contacted Warren and recorded this little interview, where he was kind enough to talk about his career and talked about artistic work in games.

Introduction

I’m a 20 year veteran of the AAA gaming space.  I’ve worked on a lot of games and franchises including Wheel of Time, Unreal Tournament, Gears of War, Fortnite and We Happy Few.

I’ve been freelancing now for about a year because I felt if I didn’t give that a try, I’d regret it forever. It’s something I always wanted to do … and now I am!

You’ve been in the industry for quite a long time already, having contributed to some of the biggest games of Epic Game’s library. And you also worked on a number of different jobs: from programmer to level designer. How did you manage to do so much?

That’s really a testament to Epic Games allowing employees to pursue what they feel most passionate about.  My interests shifted over the years and they were always very accommodating, allowing me to move into different roles as schedules allowed for it.

This helped me to maintain focus and interest over the course of such a long career with them.

What are the most peculiar things of working in the game industry?

Probably the disconnect between what people THINK you do all day and what you ACTUALLY do all day.  People think you’re working on whatever you want all day and playing games. You’re not.  It’s a real job with real responsibilities and real deadlines – it’s not always a party.  But having said that, it beats getting a “real” job.

What are the biggest challenges of game development production?

The biggest challenge in developing large games is really just herding all the cats. Getting everything in line so that production doesn’t get stalled out.  Having level designers stuck because art isn’t ready, or art stuck because the concepts aren’t done, or scripting stuck because the level designers are still designing the level layout … that’s all bad and pushes everything back, increasing the chances of crunch.

But iteration is also key … and you can’t always schedule creativity.

And so it goes on.

Why do so many people burn out while building these huge projects?

I would say lack of self awareness. You have to know when you’re overworking yourself and ease off a little. Larger projects are multiple year projects and you have to remember that. Leaving it all on the field in the first 6 months isn’t going to be beneficial to you on the back end when the hours get longer and you’re trying to ship the game.

Then you’re piling overwork on top of already existing fatigue and it’s only natural that people will burn out under those sorts of pressures.

I’ve bee lucky in that I’ve avoided it other than a few years where I questioned what I was doing. But easing back a little allowed the fire to rekindle itself without going out entirely, so I’ve never experienced actual burn out myself.

How does the production of content usually work in game industry?

Well, if you’re lucky you get a piece of concept art. Often you’ll just get a verbal description. Sometimes it’s a few words on a spreadsheet.

From there you block the model out, get that approved, then move through the standard pipeline – high/low poly, baking, etc.

How do artists model and build all those incredible assets in such short periods of time?

Well, typically artists will become specialists in specific areas.  Characters, or vehicles, or hard surface props, or effects, etc.  Specialization allows you to create assets within that specialization with the most efficiency.

Beyond that, it’s just “being a professional” and getting the job done. The schedule is real and needs to be respected.

How do you optimise and cut the time during the production?

Usually by cutting features or deleting levels/content.

Seriously, that’s the best time savings ever. But that usually hurts the quality of the game so you try to find ways around that – whether it’s reworking a feature or putting in some extra hours to make sure the level/content reaches the quality bar required to ship.

What are the most peculiar things to keep in mind while building assets and environments for games?

The most important thing to keep in mind when developing assets for games and environments is HOW the asset is going to be used.  Know what you’re building.

Is this a background prop or something the player will be picking up and turning over in their hands?  Will this piece be used one time or hundreds of times in a scene?

Those kinds of decisions will dictate things to you like triangle counts, detail level, texture sizes, etc.

How is this type of 3d content different from things that artists usually produce for their own personal projects? What’s the difference here? How is game 3d content different?

The most obvious difference from personal work is that you have to weigh the wants and needs of the art director and balance that with the technical requirements/restrictions of the game you’re making.  Your voice is still in there and still matters, but it’s often muted by the needs of the game.  Not in a negative way tho – the art in a game needs to look consistent so it’s necessary that the team work together and creative something visually cohesive.

Personal work is often “gloves off”.  Do whatever you want, as long as it looks pretty and shows what you wanted it to show.  You would never be able to ship a weapon sporting multiple 8K textures in a real game … but it would sure looks spectacular in your portfolio!

How can an artist influence the gameplay? I mean all the elements such as lighting, movement, color could be used during the game production to achieve a better gameplay, better flow and so on?

This is often done in tandem with level designers.  They can use lighting to steer players where they want them to go, for example.  And they will make requests to artists for specific pieces or modifications to existing pieces if they need something specific to facilitate gameplay.

For example, a new version of a fence mesh that’s broken through to allow combat to take place through a specific area.  Or a hole in a wall mesh.  Or a broken version of something they can swap in after something specific explodes.

And artists of course will work with the designers to establish mood and setting for levels, which includes everything from post process to lighting to color schemes.  This can often be a different artistic discipline from prop creation however, depending on the size of the company.

How would you advice approaching the choice of tools?

My advice on this would be to learn as many tools as you can. Knowing what’s available to you helps you make the right decisions about what to use for any specific task.

For example, if you only know ZBrush, then you’re going to try and use ZBrush to solve every problem. That’s not efficient.

Not to say you have to master all the apps out there but at least having an idea of what each one is good at and how they work is beneficial.  It puts more tools in the toolchest, as I like to think of it.

Which tools would you recommend learning? What’s the best combo, that you think is used in any game production?

These days, the new hotness is definitely Substance Designer/Painter and CAD programs like Fusion 360.  I think those are going to come into heavy use in the near future.

Substance already is heavily used but I think that’s only going to increase.  And I think CAD programs are going to move more and more into artists toolchests as tools like Fusion 360 develop and mature.

What would you recommend to younger artists, who want to get into game development? How should they approach this task? What are the best ways that they can increase their chances of getting a better job in gamedev?

I’d recommend making sure the industry knows you exist.  Be visible.  Get out there, show your art, promote yourself.  Post WIPs where industry pros can see them, ask for feedback, act on that feedback, etc.

Get yourself in the loop and be seen.  Then when it comes time to hire someone, you’re not a stranger. They recognize your name from that Polycount thread or that 80.lv interview or whatever else you’ve done.

My other recommendation is to make sure that the end of school isn’t the end of your learning. Never stop learning. Always stay aware of upcoming technologies and apps and what artists you admire are doing.  If you’re not current, it’s harder to justify hiring you as you’ll need to be brought up to speed on something you should already know.

You’re trying to get a job as a professional. Act like one. Take it seriously.

Warren Marshall, 3D artist

Make sure to check out YouTube channel for game art tips and career help.

interesting links

ALWAYS BE LEARNING: KEEPING YOUR ART CAREER FRESH

BE VISIBLE: HOW TO GET A JOB IN GAME INDUSTRY?

BUILDING A TRACKER FROM ALIEN ISOLATION

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Allegorithmic, assets, cad, Epic Games, Fusion 360, game development, gamedev, indiedev, Substance Designer, Substance Painter, UE4, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Epic Games, game development, game engine, gamedev, games industry, indiedev, Quixel, Quixel Suite, UE4, Unreal Engine 4, Virtual Reality, VR

Feed enhanced by Better Feed from Ozh

During most recent talk at Gnomon School Helder Pinto (Blizzard) and his fellow colleagues EA DICE 3D artist Joy Lea and Naughty Dog environment modeler Martin Teichmann gave some super valuable tips for environment creators, who want to build great high-quality levels faster and more efficiently. Here are five tips from the masters. This article was originally published on Gnomon’s official website.

Arbor Vitae: a personal environment project by Blizzard’s Helder Pinto and Philip Klevestav.

During the evening, the recording of which can be seen on Gnomon’s Livestream channel, the panel demonstrated the workflows they have developed through their work on titles like Overwatch, the Battlefield series and Uncharted 4, before taking part in an audience Q&A. Below, we’ve picked five of their best tips for improving your games environment work, from which software to focus on to the key modeling and texturing skills you need to master and how to balance your level designs.

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Allegorithmic, game development, gamedev, Gnomon, indiedev, Marmoset, Marvelous Designer, Maya, Quixel Suite, Substance, Substance Designer, Substance Painter, UE4, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

3d artist Saimon Ma talked about his award-winning project, which was most recently featured in The Artstation Journey Challenge. It’s a great talk about the production of high quality 3d models: sculpting, texturing and post production. By the way it’s all rendered in UE4. Check it out!

Introduction

My name is Saimon Ma & I am an Artist at Sparkypants Studios. My current roles include environment concept development, modeling, and material work. I recently graduated from the Savannah College of Art and Design in Georgia, USA. This is my first studio gig. In the past I’ve done work for a indie projects part-time, but nothing big.

On my personal time, I do a lot of sketching and experimenting with techniques and styles. I try to do something different on my personal time than work, such as 2D work when my work is 3D, or organic versus hard surface modeling, and vice versa. I think it’s especially important to have unexpected inspirations for your designs when it comes to interesting designs, so working on different things is how I try to keep that happening.

Project

I had around three months of free time to work on personal projects between graduation and starting work. The Artstation Journey Challenge announcement made waves on my feeds when I was deciding on what to work on, and it seemed like a perfect opportunity. I also enjoy the sense of urgency you get from challenges & competitions, so this was right up my alley. The brief for the Journey Challenge did a great job specifying the bounds and limits of what you can make while giving you a great deal of freedom to come up with your design and piece. I knew I wanted to push posing this character, and work some kind of personal transport in there somehow. I also knew I wanted to make another character with real-time techniques, since that is what I enjoy doing more right now. Finally, I also wanted to work on my polygon efficiency a little, even though it wasn’t really important for the challenge.

The First Stages of Production

I decided on time travel as the main theme. The character would be some kind of maker, geeky type character with glasses who uses a personal transport. I also thought injecting some kind of punk aesthetic would help sell the geek and ‘high school prodigy’ feeling I was going for. I knew I wanted some kind of Arthurian subtext, so I’d need to work in an Excalibur somehow, too. My reference / mood boards consisted of a mix of cafe style motorcycles, punk fashion & Japanese school uniforms, WWII flight suits, NASA, and the DeLorean from the Back to the Future franchise. There’s also anatomy references, style references, and a bunch of photos of PVC figurines because they always have really dynamic poses. I made a bunch of mood boards out of my references, which I combined into a big sheet here. I also marathoned the Back to the Future Trilogy, as I wanted to capture some of the tone of that movie in this piece.

After some super rough thumbnails and iterations, this is the working concept sheet I ended up with after about three(?) days. I wanted to keep the character’s silhouette fairly simple and readable, since the bike is going to add complexity to the silhouette. In hindsight, I could have pushed her silhouette and color palette a little more at this stage. I knew there wouldn’t be enough time to make a super intricate rig for this character, so I kept the hair simpler and shorter so I wouldn’t have to fiddle with dynamics and IK spline rigs too much. Seeing the character in context with the bike was going to be important, so I worked in a sketch with her riding on the bike to get a rough idea of how the designs work together. I made sure to mix metal with a lot of fabric, plastics, and metal because I feel like there’s too many pure metal designs in sci-fi lately.

Mechanics of the Vehicle

The pivoting wheel design is based on the mechanism on the Dodge Tomahawk motorcycle, which raises one wheel and lowers the other to bank the motorcycle and help it steer. In this case, it’d also help with steering and balance because it widens the wheelbase and one wheel can spin slower to turn the bike. The rest of the mechanics are mostly cosmetic or there so I can make more interesting poses. The rig used is very simple, with the most “complex” features being the single-chain IK solvers on the shocks, handlebars, and pedals. Technical details like the wires, particle accelerator, and coils are there to help it look more DIY and like a time machine, using the same approach as the Delorean from Back to the Future. The coils on the shocks don’t retain volume when they retract and extend, but I wasn’t going to stretch them too much from the resting pose, so I left them as-is.

Model Creation

The character was mostly sculpted from a Dynamesh sphere in zBrush, remeshed in Topogun, and rigged in Maya. Little insert meshes were modeled in 3ds Max and Maya, but zModeler was used for bits like her watches and harness buckle as well. Clothing was modeled in Marvelous Designer in separate pieces, then heavily edited in Zbrush for better silhouette and folds. For her hair, I combined the use of GMH2 for the hair textures with Tom Parker’s technique for placing hair planes using zBrush. For her glasses, I transferred the normals from a squished sphere to the glasses lenses for better smoothing. Her skirt’s base mesh was modeled in Maya, then converted to an nCloth and simulated to get the folds, since I got better results making pleats this way than in Marvelous Designer. In the future I might figure out how to make pleats in Designer instead. I put some basic eye makeup and brows on the character super early on in the sculpting stage, since it affects how the eyes look.

Face at various stages:

Marvelous Designer screencap:
Early sculpts:
Later sculpts:

Topogun screencap:

Wireframe in Maya after remesh:

Nearly finished Unreal screencap:
I did a paintover after the blockout of the model as well, to give myself some direction for the final details and put myself back on track.

Renders for paintover:
Paintover:

The bike and sword was modeled in Max with a baked hipoly workflow. The High poly was modeled with SubD modeling. The bag and seat were sculpted in zBrush. Occasionally Send to Maya was used to help in UVing the bike, since I’m more comfortable UVing in Maya. The bike has two materials: one mostly baked normals material for most of the bike, and one mostly tiling material with no normal bakes for the frame and wires. xNormal was used for baking.

Bike High poly:

Bike Low Poly:

Pose

I spend a lot of time thinking about the pose, since it can make or break a model. I started with some thumbnails, and used Jeremy Ernst’s Animation and Rigging Toolkit to make a basic rig for the character, which I used to play with poses based on the thumbnails. I just picked whichever one I felt was the strongest, which is usually one of the most dynamic & natural poses that doesn’t break the rig. The pose works with the base to make an interesting diorama, so I made sure to think about them together whenever I can. In the end, I probably spent a good week getting a good pose and base. Since the rig isn’t perfect, I had to sculpt some corrections after posing the character with the animation controls. I did this nondestructively with blendshapes in Maya and brave rabbit’s SHAPES plugin.

Material Production

I used Substance Painter for all the textures on this model, including the masks for the sword. The only exception is the base color for the skin, which was painted with polypaint in ZBrush before being heavily edited in Substance Painter. zBrush’s Color Spray mode with the standard brush for polypaint can look pretty convincing for getting colored pores on your model. Painter’s library of materials and smart materials is great for getting a head start, but I wanted to push myself by relying on them less and making custom effects more, similar to the workflow in Substance Designer. In the end, I think it helped a to get more interesting materials. I used a Metalness/Roughness workflow. Normal baking was a combination of Zbrush and xNormal baking. Curvature & AO were baked using Painter’s internal baker.

For materials, I always start with the flat base values and colors for everything, make sure that works, then go into textural details and wear. Sometimes, the material and color choices change at this stage from the concept. For this reason, I didn’t use an ID mask for most of the model, and just painted where I wanted materials to be in Painter. When there’s livery, I’ll include that as early as possible too, since color changes how you read the shapes of a design.

Another thing I like to think about is breaking up mechanics into layers of materials: lower layers, and upper layers. If you think of something like a car, the chassis, engine, and the shell are all made of different kinds of metal, plastic, or carbon fiber for good reason. Thinking about it functionally like this means that naturally, mechanical things tend to have material breakup based on how deep the part is within the machine. Then things like joints and heat shields tend to have their own materials too, since they have to be resistant to friction and heat. Thinking about it that way makes it easier to decide how I want to texture a model, too. I ignore tiny things like screws and wires early on since they don’t contribute to the design and readability much at this stage.

For most of the fabric, I used a very subtle cavity mask as an overlay layer to get a little more color variation. Other colored overlay layers were also used to add subtle coloration.

I knew early on that I didn’t want this piece to look too grungy, but I did want it to have just the right amount of grunge to sell the materials and nothing more. That said, bikes get really dirty since they’re always oiled in the mechanical bits and oil attracts dirt, so with that in mind, I had a decent idea of where to localize the dirtiest parts of the model.

Plaid was really fun to make and not that difficult, you can get the pattern by using a few layers and the Stripes procedural effect in Painter. Skin pores is just playing around with different noise procedural fills to get the right density and variation.

To save time, some of the material folders were converted to smart materials and saved to the shelf. This way, they can be used between multiple material IDs and objects.

In Unreal, I used Material Instances as often as possible and gave materials some tweak controls to save time.

Substance Painter screencaps below. I like to use fill layers and folders with masks because I can change the base values very easily.

Hair is tricky to shade. I wanted a lot of control in the material instance, so I used a color overlay texture to get color variation in the hair, and a mask texture for the alpha. A Linear Gradient node was used to drive the colors of the hair, and I made a master material for the hair that allows different colors to be set at different points of the hair at different falloff rates. Up to 3 colors can be used. Other paramaters such as AO, Roughness, Spec, and Scatter can also be driven based on position on the gradient. The hair texture has the roots at the top and the tips at the bottom, which makes this possible. Since the hair uses the same texture space repeatedly, I baked AO into one of the hair’s vertex color channels to get unique ambient occlusion under the goggles and at the roots of the hair.
Texture maps:

Lighting

I used mostly dynamic lights and a few stationary lights to light this scene, because the hair shading model in Unreal Engine 4.11 has a shading issue with stationary and static lights from certain angles. I kept the number of cast shadows as low as possible for readability and performance. I started with a single main light, a rimlight, and a fill light, then kept adding accent and fill lights as needed. The result is a lot of tiny lights with small radii, but the overall look was already established by the first three lights. I kept the test cube that comes with UE4’s Advanced Lighting map in the scene throughout the texturing and lighting process to make sure nothing’s over or underexposed.

The sword material had a few components: the base shape, the shock diamond effect, the scanlines effect, and some layered UV warps, most of which are on different panners.

Here’s a short video showing the effect:

INTERESTING LINKS

LEARNING SUBSTANCE PAINTER 2

3D CONTENT CREATION TIPS FROM INDUSTRY EXPERT

CYBERPUNK ALLEY: AAA-ENVIRONMENT WITH ACCESSIBLE TOOLS

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3DS Max, Allegorithmic, artstation, game development, gamedev, Games, games industry, indiedev, Marvelous Designer, Maya, Substance, Substance Painter, Topogun, UE4, Unreal Engine 4, ZBrush

Feed enhanced by Better Feed from Ozh

This Unreal Engine 4 project was created by environment artist Aiko Shinohara. She did the textures, modeling, scene layout and lighting. The scene project was rendered in Unreal Engine 4.

© AikoShinoharaArt, 2016 YouTube Link

© video-editor for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: environment art, Epic Games, Technology, Unreal Engine 4, video

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3DS Max, envrionment art, game development, gamedev, games industry, indiedev, Marvelous Designer, Photoshop, topogum, UE4, Unreal Engine 4, xNormal, ZBrush

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: E3, Epic Games, gamedev, Games, games industry, Gears of War, Gears of War 4, the coalition, UE4, Unreal Engine 4, Windows 10, xbox

Feed enhanced by Better Feed from Ozh

Amazing 3d artist Joel Zakrisson gave a super detailed breakdown of his scene production. He discussed his texturing process, the way he models his creations and uses post-processing.

Introduction

I’m Joel Zakrisson, a 20 year old artist with a passion for 3D modeling. I’m currently studying 3D graphics at Future Games in Stockholm, Sweden, while simultaneously working part time as a VR consultant for Vobling, a company doing architectural visualization and VR-games to treat arachnophobia. I’ve been studying for a couple of years and right now I’m in the middle of finishing all of my personal projects before reaching out to some of the big game companies out there.

I really enjoy all aspects of 3D modeling but especially environment art for games and big hero assets. I love taking on huge projects that requires a lot of patience.

General Approach to 3D Production

When doing personal projects, I usually want to create really epic scenes with a lot of assets. To avoid feeling overwhelmed, I start by modeling the main asset and don’t focus at all on the environment until it’s finished. This way, I can also build up excitement and ideas for the rest of the scene while I’m working.

In the case of my Cog Ship Battle project, I actually started out with just a catapult over a year ago, without any thoughts of adding a ship or a giant sea dragon.

When my main asset is done, I generally compose a moodboard consisting of images with the feel I want to convey and the environment I’m about to create. When deciding what to put in the scene, I usually model directly from reference or freestyle around existing designs to fit my theme.

Something I really like and that I think others often rush is spending time and love on the presentation. While it’s often almost half of the work, I do think the environments and dioramas makes the models stand out from the crowd.

The next step is building a basic blockout of the scene in Maya, which I later move into Unreal Engine or Marmoset Toolbag. I like to present dioramas in Marmoset and environments in UE4 since it’s the game industry I want to work in.

I think it’s important to do many iterations of a scene and experiment with the presentation while always asking yourself “how can I make this better?”. My cog ship diorama is actually a remake of an older piece I made. The first version was already finished a year ago, but with some more experience I could see room for improvement.

I don’t think it’s necessarily always a good idea to move onto new projects, even if it can be tempting. If you push yourself to improve on what you’ve already done, you force yourself to get better. Otherwise, you might fall into the trap of taking every new project to the same level of finish.

I’ve always tried to follow a strong line of action with this composition, making sure all pieces get aligned along an S-shaped curve. This makes the observer’s eyes stay within the scene, always returning to the focal point at the dragon’s head. It’s similar to compositional planning in 2D work, but differs in the way that dioramas have to look good from all angles – which also makes it somewhat more challenging to do. So if I know I’m going to present my project in a diorama, I spend a lot of time analyzing my composition from all angles.

For the Locomotive Exhibition piece I used the environment as a circular frame, while having the exhibition attendees as a tool to draw focus towards the focal point. I usually don’t sketch it out like this, but I’m always aware of where I want the focus to be and experiment directly in 3D to see what works and what doesn’t.

Building Intricate Models

I do all of my hard surface modeling and baking in Maya. It’s my preferred program and what I’m working with the most, both in school and in my part time work. My workflow is similar to the usual high poly to low poly workflow that’s quite common, with the exception of a few things. I prefer modeling most of my details (such as bolts) directly in the high poly model since I find it just as fast as adding it in the normal map later. When possible I also try to use floaters, instead of modeling parts into the geometry.

I used to do all modeling with Maya’s marking menus, but during my time at Future Games I’ve completely transitioned to custom hotkeys and scripts, which speeds up my process a lot. It’s like night and day for my workflow, so I highly recommend binding as many keys as possible. In my opinion, if you use anything more than once, it deserves a key binding.

I mostly use Zbrush for the organic modeling and Marvelous Designer for cloth creation.

Texturing 

I almost always use Substance Painter for texturing nowadays. It’s really fast and simple to work with. I bake ambient occlusion, curvature and whatever other maps I need directly in the program, and use this information to generate masks. This is where all the small high poly details really pay off, since the masks adjusts to them.

Using smart masks and tweaking them in different ways is basically all I do. For the dragon I got some great details from the alphas used in Zbrush, which allowed me to easily generate color variation for the dragon’s height differences. I used a green base to match the ship’s colors and some red and yellow to make the head and various details stand out. Generally I start out by keeping the colors of different objects within a similar color range and push the colors when everything is more or less established.

The old version of the ship was completely made with tiling textures tweaked in Photoshop. This resulted in a lack of texture variation, especially since you can’t vertex paint in Marmoset. For the new version I combined the tiling textures into one map, and realigned my UVs after that. I used a normal map generated with Quixel’s NDO to get curvature and AO maps for the planks, which helped me create some accurate wear and algae masks for the base.

For the sail I didn’t have much information to go on from the baked maps, but I could still blend different procedurals to get a used and worn feel to the fabric. I had the patterns and colors in different layers, so I could adjust wear for them individually. Having small variations in height and roughness adds a nice touch as well.

I usually don’t paint much with the default brushes in Substance Painter, but I always use the Dirt 1 or Dirt 2 ones if I’d like to edit any of the masks, which I do with a paint layer.

I barely use ID-maps anymore, because of the Polygon Fill option. It’s really fast and flexible when blocking out materials for faces, UVs or whole objects.

My texturing overall is a pretty ordinary PBR workflow, and I don’t really use any particular shader tricks.

Post Processing

When working on my projects I always get inspired to add even more and more stuff. If you include small additions like particles, characters or small animals in the scene you can easily enhance the storytelling and the sense of scale and life.

Since the seagulls and fish are so small, I didn’t want to put too much time into making them. Sculpting these minor details in Zbrush would’ve been crazy. I solved this by tracing two images of a seagull from the side and the top with Create Polygon Tool in Maya. Then I merged them together while using the same images as textures.The same process was used for the fish.

Similar kinds of tricks goes for the humans. I’m not really a character artist, so I used sliders to generate a suitable, but still unique human in a free program called MakeHuman. The actual work was instead made in Marvelous designer, where I sewed my own medieval clothes.

The rain drops and the splashes are basically all planes composed in Maya, with a few texture variations.

I often get asked about the water base of the scene, which basically is a tessellated water image. The trick here is a generated height map in Bitmap2Material, that’s applied on a circle shaped plane in Marmoset.

To make the underlying sphere I used the same height map in Maya and extruded the geometry down and in towards the middle.

As for the water materials I used a glossy non-transparent surface for the top, and a transparent blue-green material for the bottom. I used the same normal map for the bottom as for the top. Having local reflections enabled made a huge difference to the water reflections.

Lighting

I believe the lighting is really important when working with mood and making your finished props look their best. Experimentation is something I do quite a lot, trying out lights from different angles and different HDRI images if I’m using those.

I do have some principles that I often follow, such as using warm light and cool shadows, or the reverse. This is done with a strong warm directional light in the front, contrasted with one or two subtle cool lights in the back, or with post processing where shadows are tinted blue and highlights orange. While it doesn’t work for all kinds of scenes, it’s something that usually works great when showcasing hero assets.

I used a variation of this set-up for the Locomotive scene in Unreal Engine, but with the main directional light coming straight from above. This was mainly because I wanted a spotlight affect appearing on the floor, created by the shadows cast from the ceiling’s main dome.

Advice

I was a beginner myself not that long ago, and I’ve found that finding my passion projects and sticking to them have been key to both my own development and some nice portfolio pieces. When working on something you care about you have no choice but to conquer your challenges in one way or another. So don’t focus too much on what you think other people want to see, but instead what makes you excited, especially when you’re just starting to learn.

While the subject matter of course is important, a great presentation weighs just as much. Think about a scenario, make a small diorama, it doesn’t have to be a complete environment if that’s too much for you. The good thing about starting off small is that when you’re in the middle of it, adding new things to an already existing environment gradually becomes easier.

Try and take shortcuts where you can. Small pieces of the presentation shouldn’t take too much time, even if it might seem complex, like a character in a scene where it’s not the focus. Instead, spend your time on the assets that matter the most in the scene.

Finally, to increase your speed I couldn’t recommend it enough to use hotkeys. Not only does it save you time from looking through menus and marking menus, it also impresses your friends and clueless onlookers. Here’s a link to my custom Maya 2016 hotkeys if you’d like to try them out.

Joel Zakrisson, 3D Artist

interesting links

BUILDING INTRICATE DIORAMAS IN 3D

SIMPLE AND RICH LANDSCAPE BUILDING IN UE4

MOODY SCI-FI ENVIRONMENT CREATION

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Allegorithmic, b2m, gamedev, indiedev, Lighting, Marvelous Designer, materials, Maya, Photoshop, Quixel, Substance Painter, texturing, UE4, Unreal Engine 4, ZBrush

Feed enhanced by Better Feed from Ozh

Gearbox Studio Quebec has a very interesting open position for the environment artist. If you’re into Unreal Engine 4 – this is your chance.

Gearbox has a long history developing awesome games with Unreal Engine. Borderlands series was actually created with the help of Unreal Engine 3. The most recent mix of DOTA and FPS – Battleborne – was actually developed with UE3 as well. This time though the company is building a project with the next iteration of the engine.

The requirements are pretty standard for an environment artist position. You need to know your way around Unreal Engine 4, have to work with 3DS Max, Zbrush, Photoshop, Substance – the usual suspects. Here’s the full list of requirements:

RESPONSIBILITIES

Create high-quality, realistic environment assets and content including props, architecture, geology and organic elements;
Collaborate with Design to make sure environment assets support gameplay;
Achieve quality and style benchmarks established by the Art Director;
Collaborate closely with the Technical Art Director and Lead Artist to understand and, as necessary, establish content creation processes in line with technical constraints;
Create blockout geometry, establish layouts, apply materials and details.
Stay up to date on new modeling tools and trends, as well as the latest innovations in the industry.

REQUIREMENTS

5 years of experience in a similar position;
At least 2 published multiplatform titles;
Deep understanding of video game art production techniques and processes;
3DS Max expert;
Advanced knowledge of Zbrush, Photoshop, Substance;
Experience on Unreal Engine;
Enthusiastic, self-motivated team-player who works well under deadline pressures.

Yes, and be ready to move to Quebec if you land the job.

Source: quebec.gearboxsoftware.com

© Admin for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3dart, design, environment art, gamedev, gamejobs, leveldesign, production, UE4, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3d coat, 3DS Max, composition, design, Epic Games, game art, Game Design, game development, gamedev, Games, games industry, indiedev, Lighting, materials, Photoshop, Quixel, Quixel Suite, SpeedTree, UE4, Unreal Engine 4, World Machine, ZBrush

Feed enhanced by Better Feed from Ozh

Game Artist Nikola Damjanov presented the breakdown of his first project in Unreal Engine 4 at Noredus Blog. Developer talked self improving, planning and modeling, texturing, post-processing and other stuf. How does one developer make a step from using Unity to learning another engine? Check this breakdown to find out.

Self Improving on a Large Scale

I hope this article finds you in a good mood because sometimes that is all it takes to harness motivation and creativity. I would like to start with a small disclaimer: this was my first project in Unreal Engine (UE) and I do admit that par of the workflow could be better and/or more optimized so if you have any comments, please do share them.

So let’s start. The dawn of this project was more than a year ago as an internal game prototype. We wanted to try a game with PvP gladiators and with a rock/paper/scissors mechanics. We had a week to do it so I kit-based and slap-dashed a lot of stuff just to make it in time. And, of course, the visual fidelity of the game was not the target here but nevertheless I wanted it to look at least nice-ish.

screenshot from the initial mobile game prototype

The game was scrapped for various reasons but I never really could let it go. And then, a year later, I stumbled across the old files while I was thinking about my next personal project. That’s when I decided to revisit the scene and redo everything from scratch, the way I had wanted it and seen it.

The initial concept I had was to remake just a part of the arena and use it as a backdrop for a gladiator’s fight scene. I had a static scene and two characters in mind at that point, and had an idea to render everything in Marmoset but I quickly ditched those plans. Shortly after I started rebuilding the scene, I immediately searched for a more engaging camera angle than the one from the game prototype but that was hard; I was limited by the small part of the arena and however I tried to move the camera, something was missing.

initial size of the arena

That’s when I decided to build the whole thing. I thought it would be interesting to present it as a moment in time but with the freedom to pan and rotate around, to explore the scene and absorb more pieces of the story by experiencing the environment, atmosphere and the characters. So I switched the idea of a static shot with one of a diorama. But the project was derailed again when I realised that Marmoset was not the best option to render scenes at this scale. I love Marmoset for presenting characters, props or small sections of scenes but I felt like this was out of its scope. I just couldn’t get the quality or flexibility I needed. That is when the second big change happened and I decided to switch to UE.

scene assembled in Marmoset Toolbag 2

I have a lot of Unity experience since it is my everyday work tool but I never had the chance to play around with UE on a more serious scale. I’ve seen some mind-blowingly impressive real-time stuff from other artists and that gave me enough motivation to learn something new and to try it out.

scene assembled in Unreal Engine; immediately showing more depth and atmosphere

This is the general story of a chaotic road that led me to do the scene in the way I did. It was filled with trial and error but I liked it since it made me try new things and learn as I went. I find it interesting that sometimes when I do personal stuff, it’s pretty straightforward — I know what I want to do and I just do it using the steps that are well known to me. And that is perfect for times when you are focused on the end result. But every now and then, I just let the project take the steering wheel at guide itself. It’s much more time-consuming and frustrating, but I found out it’s also much more educational and fun. You get the end result but the journey is enjoyable as well.

OK, now that we know what we want and how we want to do it, let’s go through some of the actual work steps.

Planning and Modeling

Because the project was going all over the place in its initial states, these two stages eventually got fused. The models themselves are extremely simple. Since the entire scene is mostly composed of rocks, I just started with basic shapes and simple poly-modeling — nothing fancy. Considering the initial plan to render everything in Marmoset, I laid out almost the entire scene in 3ds Max. I suggest you be smarter than me; if you are planning to make something like this in a game engine, just assemble the scene there, use instances and save yourself time and performance. I was too lazy to rearrange everything again because I knew I had a lot of work to do and wanted to focus on other tasks.

basic models used to build the scene

Now that I look back, I would definitely add more geometry to certain places; that is one of the drawbacks when you let the project go wild. You start with one thing in mind and end up with something completely different. Things get lost in the translation because when you are making real-time assets, technical limitations are important so try to plan ahead and avoid making those limitations into defects.

Sculpting

After knowing what the scene would look like, I separated every unique asset to a separate file, made UVs and imported them into ZBrush. My workflow there went like this: import a low poly model, subdivide many times with smooth off to get more geometry but preserve the shape, and a couple more times with smooth on to soften the edges. That gave me a good enough starting position.

Rocks were made using Clay brushes to add volume then Trim brushes to remove it. I’ve repeated and layered those steps couple of times until I got the stone look. Towards the end of an asset, I’d use Dam Standard with a low radius to define the major shapes, add more depth to crevices and insert a crack here and there.

stone sculpts

For metals, I had the same start but didn’t sculpt them too much. Instead, I layered two or three surface noises (Erosion, Dent or Corrugated)  of various shapes and sizes, and then used the Trim brushes in between the steps to add large scale surface deformation. I was aiming for an old iron look and that did the trick.

metal sculpts and example of a Surface Noise

Cloth had a slightly different treatment because I only subdivided it with smoothing on since I needed soft and rounded edges with a nice flow to them. So I just creased the edges I wanted to stay sharp. I heavily used the inflate brush to make it feel like the cloth was sagging and details were added by using very nice cloth folds alphas from Ahmed Teka. I didn’t sculpt any surface details or the torn parts because I knew I would add them in the texturing phase.

cloth sculpts and mainly used alphas

Texturing

One of the few things I knew at the very beginning was that I was going to texture everything using Quixel. I’d used it before on stand-alone assets and I wanted to test it on an entire scene. The quality of the scanned textures that Quixel has is second to none but the software could be better on the usability side. Since Quixel doesn’t have high to low poly baking, I exported the normal maps from ZBrush but everything else was baked in Quixel — AO, curvature, object space normals and position gradients.

various baked maps; tangent space normal, object space normal, curvature, ambient occlusion and position gradient

First thing I did was a bit of exploring and look development. I started with a stone part since that was the majority of my scene and played around with textures. I was experimenting a lot of things like the color/darkness of the rock, how rough it was, whether it was more or less covered in sand…etc. This was a pretty playful part so I suggest you take your time to try everything. Quixel is your friend here since you can easily layer materials using their powerful masking system.

look development for the stone material

When I got the look I was aiming for, all the rocks were practically finished. I created a smart material from the lookDev and that gave me a preset defining all the different mixed materials, their attributes and masks. After that, it was just a matter of applying that smart material to all of the stone elements with some occasional individual tweaking for base color and amount of sand.

Image 11 – stone smart material used on the stairs and final diffuse map

Metal was much simpler since it just had an iron base and some coating on the top. I was afraid of how the transition from a metallic to a nonmetallic surface was going to look like in a metalness PBR workflow, but it turned out looking good.

metal smart material used on the torch and final diffuse map

Cloth was maybe the most challenging part. (Not for the texture though, that was straightforward: I used the cloth material for the base, then added some lighter color overlayed on the exposed parts, darker color multiplied in the occluded areas and a washout color for edge wear. I also added some additional dirt in the end).

cloth material used on the shades and final diffuse map

What I had trouble with was the torn parts of the cloth. I went through a couple of trial and errors before I got something that fitted the scene. First, I tried to use Quixel to make some grunge masks and then just paint them in wherever I needed them, but that just felt uncanny. In the end, I opted for a photobashing approach. I gathered a lot of references of various torn cloths like denim, flags, curtains and shirts. I then cut out interesting parts, desaturated them and played around with levels to get the best possible alpha map segments. From there, I continued arranging them as puzzle pieces, trying to match the scale and give enough details to the cloth. But be careful, it is easy to go overboard!

references I gathered for making the cloth’s alpha and the final result

The ground, well, I did that twice. At first, it was a simple plane modelled in max and modified it to look like there was some slight height variation and as if there was some dirt buildup around the edges. It was textured as an 8K map, combining two types of sand (coarse and rough), dirt and pebbles. All of that was mixed with hand-painted masks. From a distance, it looked OK but the plane was too large for one 8K map to look good from close up. I had an idea to split that plane into parts and have them use separate textures but that was a pain to do in Quixel. Anyhow, in the end, I chose the third — and probably the smartest — option; I rebuilt the ground in Unreal Engine using it’s terrain tools. I exported all the different maps from Quixel as unique tileable textures and then made a splat shader in UE that mixed those maps based on vertex colors. There is a tradeoff to this technique though — you get nice and crisp textures (even in close-up) but the complexity of mixing them is directly dependent on the amount of geometry that terrain has. So, as always, you need to find a balance between geo resolution (performance) and the complexity of mixing (looks).

comparison of ground texture details; top – unique 8K texture, bottom – splat shader with three 1K maps;

Assembly

Now I had everything I needed to start packing it in UE. After importing all the models and textures, I created three materials for the entire scene. All of the materials used the standard 2-sided metal/rough PBR base with default UE settings but with some slight differences. Cloth material, for example, used the masked blend mode (for transparency) with a subsurface shading model and the terrain had a custom built mixing shader based on the layer blend node.

configuration of the splat shader used for the ground

This point was a huge milestone for me because it was the first time I had seen the entire scene, together with textures, assembled. And, even in its vanilla state, it looked good compared to Marmoset. I was really happy that I made a right decision; it also boosted my motivation to undertake the big amount of work that was in front of me. The scene was there but it was static so it was time to breathe some life into it.

scene with just the textures

Wind

The first thing I tackled was animating the cloth parts and it was not as easy I thought it would be. I can’t elaborate on the details because my knowledge of the subject is basic and there are some nice YouTube tutorials to get you informed. In short, the workflow goes like this: you need nVidia’s Apex Cloth plugin that you can get for free from their website. Then, you need to create at least one bone and skin with the cloth object so that UE can recognise it as an animated object. After that, you apply the Apex cloth modifier to the cloth object and manually paint in the force influences on the vertex level. Of course, those vertices that are connected to something and should remain static must have the influence value of zero. You then separately export the skinned cloth object and an additional apex file that basically holds the cloth vertex information. When you import the object in UE, you will need to supply that additional file as well and then UE will consider that object as a physically animateable cloth object. All you need to do after that is to add some wind to your scene and things will work — it did take me some trial and error for things to work as I wanted, though! I left the UE’s wind at the default settings but I had to use really high force influences in order for them to behave realistically.

Mesh Details for setting up cloth in UE and how to define simulation-ready vertices in 3ds max

Foliage

The plants are actually the oldest trick in the book — just simple intertwined planes with a transparent texture. They were placed around using the UE’s foliage painting tool, which is really easy to use and powerful, and works for everything. You are not limited to foliage, as you can easily populate your level with props, cars, bricks….etc. An extremely useful feature!

 foliage painting tool in action

The plants were not physically animated using the wind but the animations were done in the shader using the SimpleGrassWind node. As the name suggests, it is simple to setup: you need a numeral value for wind intensity and a gradient map for wind weight that matches your diffuse UVs — black for the parts you want to be still and white for the windy bits. Then just plug the result into the World Position Offset of your material node and you are good to go. This approach is much simpler and resource friendly than the cloth one but you don’t get the same realism or the variation. In perspective, you could make a more complex shader that will modulate the wind intensity/weight for individual plants and improve the overall variety, but I didn’t do that.

plant shader with SimpleGrassWind

When you start painting out foliage, it’s very easy to go over the top; trust me on that. Just be reasonable because this kind of set-dressing should benefit the overall composition and not flood it.

Lighting

The arena has several lights but the most important one is the directional light that simulates the sun. That is also the only light that casts shadows. Other lights are placed and the torches simulate fire. At first, I played around with real-time lightning since I had an idea to present the scene in some daylight changing time-lapse fashion but I opted out for a baked solution because I knew I wanted slow camera movements, close up shots and crisp shadows.

daytime cycle

I hadn’t planned this step ahead so some tweaks were needed. I had to manually adjust the lightmap resolution on an object basis so I could get a more unified texel resolution across the scene and thus lightmap consistency.

I added a BP_Sky_Sphere that creates a procedural sky dome based on the position of your directional light. Then I created a SkyLight that uses the entire captured scene as a light source and had one SphereReflectionCapture in the middle of the scene for reflections. I kept the bake settings at low for the sake of iterating quickly and when I was satisfied with the light/shadow position, I just turned everything up to 11 and waited.

scene without and with light information

Particles

This was another thing that was completely new to me. I had the practical particle knowledge and experience from using them inside 3ds Max and Unity but UE’s Cascade felt different from the start. I’m still not sure I understand how it works to the fullest. I had three types of particles in the scene: fire, dust and sand. Fire is a modified version of the emitter that comes with UE’s standard package. It has suited me well from a testing and learning perspective, plus it looks good. I often find that reverse engineering, and modifying existing assets, is a good way to learn and explore, so do that when you can. The same goes for the dust — I just tweaked the settings of the lifespan/velocity/color/size to suit my needs and layered them across the scene. I took the things I learned and made a sand-storm that goes over the entire scene. It’s a very subtle effect from a simple particle system that has a smoke texture in 4 stages, a high velocity and a low opacity.

Cascade setup for the torch fire

Post-processes

After everything was there, it was time for some post processing. This is that crucial final step that will unify and polish everything you’ve done so far. Think of it as the icing on the cake. What you need to do is add a PostProcessVolume into your scene and mark it as Unbound so it affects the entire scene and off you go. Go wild, play around and find something that suits your style and your vision. Color grading is your biggest friend here; you’ll see how quickly you can change the mood and atmosphere of the entire scene. Color grading in Unreal works with LUTs (lookup tables) that you can easily author in Photoshop. Grab a screenshot of your scene, insert a default LUTs in the same PSD file and then use any of the Photoshop tools to adjust the colors — levels, curves, color balance, photo filters, gradient mapping… anything. Once you are satisfied with the result, just export the LUT piece of the image with all the adjustments and use it as a color grading LUT in UE. Simple and powerful!

default LUT (top) and the modified one (bottom); changes look minuscule here but as you can see on the next image, the effect is drastic

That is exactly what I did. Aside from the color grade, I also added some bloom, lens flares, ambient occlusion and screen space reflections.

scene without (left) and with post processing (right)

Sound

In order for the video to be a complete and immersive experience, you need sound. Just watch any animation that you like with and without sound, and you will see the kind of impact it has. I actually spent two whole evenings just searching for the perfect music and I’m still not happy with my choice! From all the WIP screenshots I made during the production, I had the animation’s storyboard in mind and I knew what I was looking for: an epic sounding track that starts slow and dramatic, and that then abruptly switches to an upbeat tempo. Every time I find something promising, I play the music in the background and walk around the scene in UE and simulate the camera movement just to get the feel. In the end, I found my best match in YouTube’s audio library. It’s a very nice resource for royalty-free sounds. The only other sound that I’ve added is the wind you sometimes hear in the background.

Cameras

Now comes the part where you can sit in your director’s chair. The fun part of this is finding out the angles and compositions that suit your scene. Pay attention to the golden ratio, symmetry and the foreground/background balance but do unleash your creative freedom — it all comes down to these beauty shots. I chose my camera angles based on what I wanted to show (cloth, metal, closeups, entrance, fire…) and what I needed (inside for the beginning, outside for the middle and bird’s eye-view for the end).

compilation of various camera angles that I thought were interesting

Choosing the shots was not that hard but animating them inside UE was a nightmare. In its current state, Unreal’s Matinee is just a pain to use. The way you select cameras and create their keyframes is very un-user friendly. I reset the animation or placed keyframes on a wrong camera so many times that I lost count. That’s why I’m really looking forward to Unreal’s new Sequencer that will hopefully remedy this problem and bring a much better UX and an NLE approach.

my Matinee setup with 14 cameras/shots

Final output

Once you set up your cameras, you are actually one click away from the rendered frames. And since this is a game engine, you won’t wait long. I rendered everything out in 4K and used AfterEffect to put the pieces together. I wanted to keep the straight look from UE so I didn’t do any post-production in AE.

For some additional beauty shots, you can check out the Artstation link here and I highly suggest you watch the final video in HD full-screen and with audio:

As a final touch, I leave you with this panorama.

Panorama of the Arena

Article by Nikola Damjanov, Game Artist at Nordeus Blog

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: game development, gamedev, indiedev, Marmoset, Photoshop, Quixel, Unity, Unreal Engine 4, ZBrush

Feed enhanced by Better Feed from Ozh

The creators of Strength in Numbers Studios have talked about their debut game Tuebor and discussed the difficulties of indie development.

Introduction

Strength in Numbers (SiN) Studios first opened its doors on April 1, 2015. The name originated as our CEO believed that as our team grew larger, SiN Studios grew stronger. Plus, it has a fun little acronym.

Tuebor is SiN Studios first launch title.

Many of our senior staff have worked for AAA studios and have shipped titles such as: Fallout: New Vegas, Dungeon Siege III, Alpha Protocol, Call of Duty: Ghosts, Halo 4, Halo Combat Evolved: Anniversary, Halo Reach, Call of Duty: Black Ops, Galactic Civilizations III, Saints Row, Saints Row 2, and Red Faction: Guerilla.

Many of our junior staff have also shipped indie games from various projects.

The CEO, and Founder, of the company had owned a LAN Center for almost 9 years, which is where the game design originated from.

Tuebor

Tuebor is a 3rd person, cross-genre, competitive, multiplayer, action game where players will conquer, defend, and survive against hordes of mutants, cyborgs, and an evolving artificial intelligence system in a post-apocalyptic, dystopian future.

Our goal was to capitalize on features and aspects of games people enjoyed, while discarding the features that were less than optimal. Everything is end-game content so players can jump in and enjoy the experience without having to spend months getting to the good stuff.

The game is heavily influenced by sci-fi and cyberpunk with inspirations from such great franchise as Aliens, Blade Runner, and Star Trek.

Building Multiplayer Element

One of our primary goals was making sure, throughout the creative process, the game remained fun for players. Each map, game mode, and characters’ abilities have been play tested repeatedly over the course of the development cycle.

The closed alpha testing we have been doing over the last few weeks has helped bring fresh eyes on the game. We’re continuously tweaking things that don’t feel right or don’t give the user a good/fun experience.

Our concern too was to make sure the levels didn’t just feel “grindy”, but also had a strong element of fun, or challenge, or excitement about them.

Game Engine

We are currently using Unreal Engine 4 for our game engine, and a custom “learning and evolving” AI engine that we have been adapting and licensing from Michigan State University.

Useful Tools

The tools are easy, just learn everything about Maya, Zbrush, Unreal Engine, Visual Studio, AWS, PhotoShop, After Effects, Agile Methodology, SharePoint, MySQL, NDO, DDO…

But seriously, the production pipeline got into the best position when we were able to hire an experienced full time project manager/game producer. Having someone be able to organize and assign tasks to keep the pipeline flowing has been immeasurably helpful.

Avoiding Mistakes

The entire process is a learning process, whether you’re a junior developer or have shipped a lifetime of AAA titles, so there are always bound to be mistakes.

The big mistakes came early on in assuming we had enough design documentation. Once we learned that we needed to step up the documentation we were able to keep all teams running more efficiently. Having people wait for the design team to flesh out the information wasted more time that would have been prudent. The fix for this was having the design team work overtime (nights and weekends) for several weeks to get the pipeline unclogged.

Monetization

We are approaching the monetization on three fronts to allow gamers to tailor their purchases of Tuebor based on their personal preference.

Right now (subject to change) the plan is as follows:

• Free to play: to allow players to get their feet wet and see if they really like the game
  o F2P accounts will have limits to weekly content such as game maps and heroes onrotation.
• $15 Skull Ball Edition: this will include all of our Skull Ball game maps and 10 of the30 heroes
  o This account will also still be allowed access to the same weekly maps and heroes
• $30 Champions Edition: this version will unlock 20 of our 30 heroes and all of the game modes and maps.
• $45 Heroes Edition: this version will unlock all 30 heroes and all game modes that ship at launch.
• $60 Ultimate Heroes Edition: this version will unlock all 30 heroes and all game modes that ship at launch, AND all heroes and game modes that we release for the entirety of the game life cycle.There are also vanity items in-game to purchase to snazzy up your hero in very unique and fun ways. These will be sold via low dollar amount micro-transactions between $1 – $5 for the Tuebor enthusiast that wants to pimp-their-hero out like a baller.There are no pay-to-win elements of this game, you cannot buy items to make you stronger, faster, better.

Emily Springer, Marketing Manager at Strength in Numbers Studios, Inc.

INTERESTING LINKS

LIGHTING SCENES: REALISTIC LIGHTING FOR AUTODESK MAYA AND MODO

3D CREATION WITH MAYA, UNITY AND 3D COAT

STUDYING LIGHTING & MATERIALS IN UNREAL ENGINE 4

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: After Effects, Agile Methodology, AWS, dDO, game development, gamedev, indiedev, Maya, monetization, MySQL, NDO, Photoshop, Quixel, Quixel Suite, SharePoint, strength in numbers, tuebor, UE4, Unreal Engine, Unreal Engine 4, Visual Studio, ZBrush

Feed enhanced by Better Feed from Ozh

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: 3DS Max, Allegorithmic, game development, game industry, gamedev, indiedev, Photoshop, sci-fi, Substance, Substance Designer, Substance Painter, UE4, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

As  you must have heard, Nicolas Pirot has recently released a great project, showing Dynamic Decay in UE4. It’s a great way to show how you can do real-time changes in the environment with some Bluprints magic. Specially for 80.lv, Nicolas talked about his project and the way he achieved these amazing results. You can purchase Dynamic Decay project for just $5.

Destructive Time-lapse Project

The project currently is focused on the material deterioration, although it has a lot of other interesting things going on as well. The concept was an idea I decided to undertake as part of my graduation work as a game graphics student. I’ve always had a strong interest in visual storytelling through interesting environment design in both games and film. As a student, my graduation work was the perfect opportunity to combine my love for weathered environments with a strong interest in Unreal Engine blueprints and Substance Designer materials.

The unique aspect of this tech showcase is giving the user the possibility to completely transform the environment during runtime. Right now it’s based on a set of open parameters, but the gameplay possibilities using this type of workflow are endless.

The scene has a bunch of various materials, which are changing in realtime depending on the surrounding circumstances. Could you talk about how is the whole thing organised in Unreal Engine 4?

Everything in the scene is done using UE4 blueprints and materials. Both been absolutely fantastic to work with, giving me as an artist the tools to create awesome things, even without a coding background. This combination of visual beauty and backend ease of use is why I chose for Unreal Engine 4 over other popular game engines.

The framework I’ve built creates a link between the UI, the level blueprint, and the actual materials/shaders, in order to pass on input information to the materials.

One of the challenges I faced early on was finding a way for materials to be editable in runtime, and to control them without having a hardcoded sequence of material changes over time. The issue was that regular materials or material instances could not be edited once the game starts running, which was quite problematic at first. The solution came in the Shape of Unreal’s Dynamic Material Instances. Even though they can only be created and assigned during runtime, they have the major advantage of being editable on the go. Once I got them integrated in the system, I could create material instance as normal, and use the UI input as parameters.

The materials themselves use a custom home-grown greyscale map I’ll be referring to as a decay map. As pictured in the accompanying image, a decay-map is a greyscale map that defines different grungemasks, depending on an input parameter. This input parameter is a scalar input value, calculated using the UI values defined by the user. Using some clever shader tricks a certain area of the decay map will be converted to a greyscale mask, which can be used to grunge up areas. The input parameter defines the lower limit in the range of brightness that should be interpreted as masked. When the input parameter is 0, all values between 1 (255,255,255) and 1 are masked. The more time progresses, the larger the mask range becomes. This newly created mask is then passed on, and the masked areas that were once clean materials, become totally grunged up. This process repeats a few times for each different effect, as to prevent obvious tiling and repetitive damage.

The materials were ALL made using Substance Designer 5, Substance Painter 1 / 2 and Bitmap2Material. Working with Substance is particularly great because of the immense amounts of control I get as an artist in creating and fine-tuning details in my materials. Some of the materials such as the clean tiles are fully procedural in substance designer, while others (such as the moss) use base-maps created in Bitmap2Material.

How did you approach the production of the materials here? How did you approach the creation of the materials and their gradual deforming? How did you change the qualities of the materials to make them feel more influenced by time and other factors? From the visual perspective.

For each surface, a set of materials was created. Each set represented the far end of a certain type of destruction (plus a clean set ofc). All surfaces have at least a clean set, a grungy one, and a mossy one. All materials seen in the scene use these material sets and combine them over each other to ensure a convincing result at any stage of decay.

Each material was sized to 1K, and uses the standard PBR material workflow. Each set has 3 maps, one for BC, a BW one for roughness and a normal map. The metalness was handled by using the decay mask in combination with a set of shader nodes, which saves 3*1K maps for each surface.

Once I’ve got the different versions of each map, they go into a single dynamic material instance, which can then be created, assigned and edited during runtime.

Could you walk us through the process of the creation of changes in the materials? How did you manage it all there with Blueprints? It just seems like a crazy task.

The big catch was to divide the huge task of user input based dynamic textures into smaller bits. The first step was to get the user input from the sliders, and to convert it into something I can use in blueprints for events and the likes. Once I had gotten those values, I used dynamic material instances to pass them on to the dynamic material instances, which can only be created during runtime. Once the materials had the 0-1 values, I could start playing around with them, the same way regular material instances work.

The real challenge in creating the shaders was creating something that both looked good, and blended in a natural and logical way. This is why I’ve created the Decay map, which has served it’s purpose well.

Could you talk about the way you’ve worked with the post-processing and fog? What did you use here and how did you add it all into the general system of the environment?

After I had completed basic versions of the materials, I moved on to create the atmosphere. While weathered textures do a lot for the mood, things such as atmospheric depth fog, colour grading, and lens effects also contribute massively.

My favourite of the above was of course the exponential depth fog. Usually used for large-scale outdoors environments, it worked like a charm on this indoor scene. The main benefit was again that I could edit a ton of things in runtime, such as distance to camera, strength and colour. Using a set of relatively simple blueprints, I’ve rigged up the colour to the heat/humidity in the scene, and the opacity threshold and strength to all parameters.

As for the lighting, the same reasoning took place and led to some experimentation. After researching on how to efficiently tinker with movable/dynamic lights in runtime, I chose to use the LinearColour Lerps, in which I blended between white, and the colour the situation required. The same thought process that went into the fog, which both worked like a charm.

The best thing about this is that, while the blueprints may be very simple, the gameplay possibilities are immense. The environment can quite literally change around a player, depending on his/her moral choices in the game.

How did you work with the destructible meshes here? Could you talk a little bit more about how these elements are integrated into the general flow of things?

The props were added to add more of a dynamic feel to the scene. Some events were scripted to be exactly the same every time something happened, while others were dynamic and physics based, to improve the diversity of the outcome.

The destructible meshes were mostly fragile props such as wall tiles, lights and bottles, which were set to become visible physics-actors after a certain point in time. The tile themselves were part of a larger system of events too, as they were attached to a specific decal above them. The point of this was to alter the base textures so that every time a tile dropped, a strong stain would become visible. I wasn’t too happy with the way the tiles were falling, so I added invisible physics props to bounce them around a bit, and create the illusion of dynamic tilefalling.

Nicolas Pirot, Game art student – Hard surface artist

interesting links

Building Temple of Utu: Blizzard Devs On Great Environmental Design

Borderlands and Gears of War Artist Explains Quick Landscapes

Guillermo Moreno Alfaro on Building MediEvil with Unreal Engine 4

© artem for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Allegorithmic, b2m, Epic Games, game development, game industry, gamedev, indiedev, Substance, Substance Designer, Substance Painter, UE4, Unreal Engine 4

Feed enhanced by Better Feed from Ozh

This is a preview of a tutorial on creating a snowy forest landscape using Unreal Engine 4 and Substance. The package costs $29.99 and includes the following:

• An Unreal Engine Demo Scene
• 4 Bonus Timelapse Videos for extra learning!
• Additional Grass Ztool and all scene source files
• All 12 Standard Tutorial Videos for Forest Snow Ground
• Bonus Timelapse videos for creating grass, scene layout, and simple paintover

© PurePolygons, 2016 YouTube Link

© video-editor for 80lvl, 2016. | Permalink | No comment | Add to del.icio.us
Post tags: Substance, Technology, Unreal Engine 4, video

Feed enhanced by Better Feed from Ozh