3d generalist Thomas Pasieka talked about his approach to environment production, modeling and the creation of materials. He’s got a lot of
3d generalist Thomas Pasieka talked about his approach to environment production, modeling and the creation of materials. He’s got a lot of warm things to tell about Autodesk Maya LT and 3d Coat, willing to share some of his thoughts about producing great content in 3d.
Hi! My name is Thomas Pasieka and I am a 3D artist (generalist). I have been freelancing as a 3D artist for the past 13 years and continue to do so as I feel quite comfortable in my position. Over the years I’ve had the opportunity to work for some incredible people and companies such as Nike, Jaguar & Landrover, Disney, Intel, General Electric, Unity Technologies, Autodesk and Warner Brothers (Harry Potter Franchise) to name a few.
Currently I am part of a team creating a VR experience for Mattel’s “View Master”. On a personal level, I created 2 of my own games in the past using the Unity Game Engine. One was called “The Other Brothers” and was nominated as a Unity Awards Finalist in 2013 and was overall well received.
I have also been involved in many other “non-game” projects. For instance, I do a lot of architectural and “art” projects, as well, for various clients (due to NDA I can’t share these yet). Additionally, I do a lot design work on top. Recently I had the opportunity to design the album art for Jordan Rudess (Keyboardist/Composer of Dream Theater) new solo album “The Unforgotten Path”. In fact, I have been doing a lot of design work for Jordan Rudess as he constantly churns out new and exciting iOS apps! I have become his “go to” source for all things “graphics”.
In 2014 I was given the opportunity to work on visuals/video for “Dream Theater’s” live show during their World Tour. One particular track (Illumination Theory) was in need of a special “Animation”. At 4 minutes long, it was not only a challange, but also very important as the band would leave the stage for the video to take center stage. That was fun and exciting!
Well, every job comes with it’s own challenges. Sometimes, it can make you feel a bit uncomfortable because you may not have worked on a certain subject before. However, throughout the years I learned to look at all jobs the same way. Because, in all honesty, I am still using the same tools. It’s not like I have to all of a sudden use a tool/software I haven’t used before. So, with that in mind, it doesn’t matter much what the subject is. In the end, you’re pushing and pulling vertices, edges and polygons. I hear a lot of people say “Oh I haven’t modeled anything like this before so I am not sure I can do it!” That right there is the wrong mind set.
My advice is to stay calm and break the model down into chunks. Analize it and go at it piece by piece. Don’t rush it and keep a cool head. That has been my approach to pretty much all my work. It does of course help if you have a very solid understanding of the software you’re using so I study a lot and watch how others work pretty much every day. You constantly evolve, but you can only become better by “doing”. Even in my spare time I model or sculpt, paint etc. I pretty much “live” the life of a full time artist. I never really quit thinking/learning.
I have a background in 2D actually. Having worked for Advertising Agencies and Post-Production companies in the past taught me a lot about color theory and design. That is something a lot of young and aspiring artists perhaps don’t necessarily have, so I often urge them to learn design principles and color theory to develop a good eye for that. It, however, takes years and can’t be learned over night. Practice is key.
Good preparation is always key. When I am tasked to model a specific object or scene then the first step is to go and google reference photos and concept art. These serve as inspiration and guide. I then tend to straight head into Maya LT to model said asset by starting with the base shapes. I very often turn the model in the viewport to see it from all angles after finishing parts, constantly keeping an eye on overall readability and silhoutte. Once I am happy with the overall shape I add details and edge bevels where needed. Edge bevel really help so much as they catch the light and draw attention. Hard Surface models/objects really benefit from that.
Doing the actual UV’s is actually not as bad as many make it out to be. Maya LT has really evolved on tools/features in that regard and I do most of my UV’s in either Maya or 3DCoat. 3D Coat is really an awesome piece of software as well especially when it comes to Retopo, UV and painting. It now also features great PBR features to make quality AAA assets. Besides software, as I mentioned earlier it’s important to train your eye on all things color and lighting. Lastly, presentation is super important. For instance, don’t render or show your images on a pink backdrop! (unless it really makes sense). Go for the beauty shot and use DOF or other Post Effects but don’t go overboard either. Subtlety is key here. There are various options to present your work. Take stills or use the very useful and PBR ready Sketchfab website to upload and show off your work for everybody to see! It even works on Facebook!
For the most part I am using Autodesks Maya LT! I am really in love with the “Game” and “Realtime” centric version Autodesk created. It has pretty much everything I need to create any kind of game prop or environment. The viewport with it’s various realtime effects such as Ambient Occlusion, Anti-Aliasing and DOF, to name a few, are awesome and very useful. The latest/current version (Maya LT 2016) also features Physically Based Shader (PBS) which is of great use if you’re using Unity or the Stingray engine. In fact, if you’re using Stingray you don’t have to re-assamble the shader/textures in engine – it will just be transfered over with everything being intact. As mentioned earlier, I do most of my modeling work in Maya. UV tools have also seen a great update and it’s actually rather fun to do UV’s in Maya now.
Depending on the task I sometimes use 3D Coat for UV’ing and Texturing. What I like about 3D Coat is it’s “Hand-Painting” features. It’s feels very natural and very Photoshop like in that regard. Then there is Substance Painter which is also ever evolving and I am using it quite frequently as well! Love the fact that you can create Smart-Materials which you can then later re-use or even share with others. Lastly, I am of course using Adobe Photoshop for certain tasks. As you can tell, I am jumping between software a bit. I am OK with that. I feel at home in all of them.
That’s difficult to answer really as every scene/object requires a bit of different setup. Sometimes the client has a specific look in his head which can be tricky to achieve especially in realtime. You also have to take into account if the camera or your viewing angle changes. Light interacts with objects in different ways. Surface texture all have different reflective or refractive values. In other words, it does help to read up on lighting and how light interacts. Scientific papers on Global Illumination, and such, are interesting and useful.
My game engine of choice is Unity. I use it 95% if the time, as do most of my clients. Having been a Unity user for the past 10 years (time flies when you’re having fun!) I have seen Unity evolve into one of the biggest game engines to day. A few years ago we had “Beast” at our disposal to bake light. It was really great back then and even allowed for baked and realtime light to be mixed. Now we have “Enlighten” which takes it all to the next step. Currently I am a big fan of the “Final Gather” feature in Unity 5.4 (beta) which takes a bit longer to render but the baked results are very smooth!
As for workflow, I start out with very low resolution settings. It makes little sense to user high resolution bakes from the beginning. That goes for realtime GI and/or baked lighting. I place lights where there is an actual light source in the scene. I often use geometry with emissive materials to light if I need a softer light. Lighting is a time consuming task but key is to really use small “preview” settings and slowly work your way up before going for high res bakes.
Starting the Production
Naturally it does help if the client has a good understanding of what he wants. I love when they provide concept art and reference images. This is usually the ideal situation. However, not every client knows what he wants and expects for me to fill in the blanks. In that case I hunt for reference images and concept art that deal with that certain topic. For a Scene it makes sense to break it up into managable chunks. Block it out first and arrange objects for a nice composition pleasing to the eye while at the same time being functional to the game. This is often referred to as “Whiteboxing” and that’s where I start my work before actually modeling the real deal. Once the scene feels good and makes sense, in context, I model, UV map and texture the various elements. Never forget the small details that make a scene believable. In my head I often have a little story and I ask myself questions such as “What happened in this place years ago” for example and try to add “story/detail” to the environment.
I have been an early adopter of the Oculus Rift. I had the DK1 and currently have the DK2 while waiting on the final Consumer Version to arrive sometime at the end of April or beginning of March. Very excited! I’ve been a fan of VR since the early days. I remember getting into Shadowrun (Pen & Paper Roleplaying Game) and ever since wanted to be in “The Matrix”! The tech and computers back then didn’t really have the power needed to make for a great VR experience. Today however we are there! We’ve got awesome tech and software to really make for a compelling VR experience.
I have created countless little demos and experiments (most of which get trashed quickly). It’s really a new world to explore and many things such as input have to be figured out. The Oculus “Touch” for instance is something I am looking forward to later this year.
Whenever I create art for VR I first take care of Scale. That is really not much different from creating art for any other game for PC and really should be the first step. In VR, things really look off/wrong if the scale isn’t right on. Your brain quickly notices inconsistencies and that can break immersion. Certain colors for instance (bright white) can lead to a lot of eye strain. I just started a more serious game in my spare time and it will be heavily focused on using the Oculus and it’s “Touch” controllers! I will share more in a few months!
I try to keep the experiences I create “comfortable” for the player. For instance, I try to never have the player look back and create strain or discomfort on his neck. In my opinion everything should play before the play in a 180 degree viewing angle. This does of course not always work as some games/experiences allow for room scale movement and interaction. Both the Oculus Rift and the Valve/HTC Vive allow for such experiences which is great. It remains to be seen what players want. At the moment everybody is trying to come up with great ideas gameplay wise and otherwise. In a sense, all current VR Developers are Pioneers exploring and creating new worlds!
Thanks for having me! Make sure to always have a fresh cup of tea!
Thomas Pasieka, 3d generalist
Make sure to visit Thomas Pasieka official website.