March 31st, 2022

Artist Focus - J Stein

J Stein’s work can be recognized from miles away. Known for his character designs, J has created a multitude of fashion projects featuring his star virtual model, ‘Hannah’. Based in Munich, Germany, J started out in photo design then moved into CG and is now wowing us with his photo-real CG renders. In this interview, J shares his personal take on Clarisse, why it’s a necessity in his character workflow, and how he creates such fantastic renders with Clarisse. Let’s get down to it!

ALAIA - J Stein




Hi J! Thanks for sharing your experience with Clarisse and its place in the workflow of your many phenomenal projects. To get started, can you tell us a bit about your background in CG?

Hi Isotropix! First of all, thank you very much for interviewing me and showing my work. How did I get started with CGI? I studied photo design here in Munich, Germany. After that, I had the good fortune to assist on a couple of productions that used computer generated imagery. My CG history began with XSI version 4.2. I took a 10 week course covering the whole program. We started with modeling a dinosaur. I remember, the guy next to me already modeled some high-res props for the dinosaur, and was already animating. On my monitor was a deformed sphere. I really had no idea what I was doing. ‘A polygon? Sounds interesting. But what is it exactly?’ I thought to myself. But I decided to keep studying. So, I grew up in XSI land. Now, I work with Clarisse and Houdini for both my professional and personal projects 

How would you describe the type of work you’ve become known for?

My work is all about the integration of CG elements in photos and product renderings. But only stills, I don’t currently do any animation. In my personal work, I’m heavily focused on my CG character pictures. This is where I spend every free second to get better, and that will keep me busy for some time! Professionally my work varies from stills to archviz and product visualization. At the end of the day, what I do is pictures. CG or real photography, or both combined.

Much of your work focuses on one CG character in particular. What was the inspiration behind creating this character?

It all started when the Managing Director of the photo agency I work for in Munich asked me if I could create a female model that could be used as a virtual fashion influencer. Fashion interests us because it contains a bit of everything; architecture, landscape, portrait, cars, props, set construction, and of course clothes. It’s a super exciting mixture. ‘Okay’, I thought. ‘Sounds great.’ So we did some brainstorming and decided to create a digital female character, called Hannah, and as you can see she now has her own Instagram account, where 100% of the work is rendered with Clarisse! Our inspiration came from some supermodels, female actresses, a little bit of everything, I think. 


How were you introduced to Clarisse?

I first heard of Clarisse in an XSI forum, where somebody had recreated the city of Amsterdam at the turn of the century and rendered it in Clarisse. I was totally amazed and thought to myself ‘Should this be the future? Being able to deal with this amount of geometries? That would be amazing!’ At that time, I was still working with XSI and Maxwell/Sitoa. Then things changed very quickly, XSI was discontinued and I knew immediately that I wanted to work with Clarisse and Houdini. A great decision, by the way!

How would you describe Clarisse to someone who hasn’t used it?

For me, it is a pleasure to work with Clarisse. Why? Here are some features I really like, in no particular order: The Polycount. What can I say? I’m still sitting in front of the computer and thinking to myself ‘That doesn’t exist at all!’ It’s just unbelievable how Clarisse deals with massive data sets. And by massive, I mean massive; billions of polygons! The preview mode in Clarisse is great. You can correct mistakes very quickly without waiting for the final rendering, saving massive amounts of time. No more stand-ins and proxies. You always see the whole picture, and that’s immensely important. The VDB display is great. You can position your VDB’s precisely without driving around half blindly in the scene. The displacement display is super fast, without rendering I can already see so much detail, even down to whether the skin pores are too intense. I may be wrong, but in my opinion the RAM consumption in Clarisse projects is significantly lower than in other rendering engines or software packages. The learning curve is also relatively low, meaning you can quickly achieve great results without spending 2 years training. And not to forget one of the biggest selling points, the scatter point cloud possibilities of Clarisse!


How did you find the learning curve of Clarisse after integrating it into your workflow?

I have to admit, technically I’m not on a TD level. Learning new software isn’t easy for me. It’s damn hard sometimes. But getting started with Clarisse was super quick, and surprisingly very easy. Of course there is still a lot for me to learn. What took me the longest was learning how to organize my projects in the most sensible way. I already knew some important workflows from XSI/Houdini, such as the TX image format, the point cloud techniques, Alembic, Disney shader. Overall, I find it very easy to get started compared to other 3D programs.

What resources did you use to help you learn Clarisse?

The Isotropix tutorials. And everything I can find on the internet about Clarisse. For beginners, there are some really good tutorials from Elephant VFX by Xuan Prada. Not free, but worth every cent.

We don’t often see such life-like characters created using Clarisse. What part of your workflow are you using Clarisse for? And what made you choose Clarisse for this type of work over other tools?

I use Clarisse for the entire workflow. I sculpt the character in Zbrush, mostly, and then I load the high-res Zbrush mesh directly into Clarisse to check the anatomy. My Clarisse Character Test Project has 10 cameras with different focal lengths and different light sets, so I compare before and after. It’s really important for me to see how the character reacts to different camera angles and lighting situations. It's really great to work this way because the loading times in Clarisse are super fast, you can quickly change things. When you ask yourself, ‘Does the head work better with Groom B under this light condition? Or this camera angle?’, it can be fixed with one click and you see the result instantly. You can quickly jump between ZBrush and Clarisse without exporting displacement maps every time. Just export your high-res mesh! 

Why Clarisse? I can’t imagine anything else, if I’m honest. Everything is super fast, the way it displays skin pores with displacement maps. It’s just wonderful. 




Now let’s talk about some of your projects. Starting with one of our favorites 'BLACK AND WHITE’'. Many of your projects, including ‘Black and White’ are rendered in Clarisse. What made you choose Clarisse for rendering? 

Actually, all of my projects are rendered with Clarisse, which is one of the main reasons I use Clarisse. You load your geometry and you can see straight away if your idea works. No stand-ins or proxies. You work on your image and you see the image. I find that incredibly important. To be honest, I can't imagine going back to a proxy standing workflow, especially for characters with clothes, pose, composition, etc. I make many versions until I am satisfied with one. And in Clarisse, that works incredibly well and is done incredibly fast. Because I work as a one-man band, the time factor is important, including how quickly I can fix errors, test things and try them out.


Another one of your projects that we love is ‘KEEP IT NATURAL’. This project includes a stunning combination of character work, materials and environments. What’s your typical workflow for creating this type of project?

I find the environment in which the character lives is incredibly important, it supports the image and tells a story (if the implementation is successful). So I create most of the geometries in Houdini, all kinds of terrains, landscapes, VDB’s, and grooming. I then use KineFX for rigging. The character comes from Zbrush, and for texturing I use either Mari, Substance Painter or, more recently, the Quixelmixer. And I create the clothes with Marvelous Designer. I’m a big fan of libraries in Clarisse, especially ready-made projects with props, plants, assets of all kinds, that you can easily load and use to quickly test different ideas. In Clarisse, I begin loading my start project and that does a lot of organizational work for me. The start project is always relatively simple; one context for the character, one context for the environment, one shading layer master to collect all incoming shading layers. The lights and cameras are constrained to target locators. For me, it's easier to light a scene that way. The project also contains some master skin materials, two render nodes (highres and preview), most of the necessary AOV’s, Cryptomatte, some custom AOV’s (mostly makeup masks), and that’s it!

How is the workflow between Clarisse and Marvelous Designer? 

The overall workflow is super smooth. I export the character mesh in an A-pose as alembic (ABC) to Marvelous Designer. Once in Marvelous Designer, I do my pattern work and export a low-res obj. file to Clarisse. In Clarisse, I do a first check. I ask myself ‘How does it look, feel, and fit? Do I have to apply a quick shader?’ And things like that. If I think it could work, I export a high-res mesh quadrangulate to Houdini (I do the retopo in Marvelous Designer. It's a new feature and saves a lot of time). Most of the time, I use Substance Painter with UDIMS for the texture work. This is a nice workflow because all the important shaders from Clarisse are supported in Substance Painter. In Zbrush, I do additional tweaks on things like the cloth, folds, topstitches, belts, buttons, and zippers. I then go back to Houdini with a quadrangulate cloth mesh. Here, I can apply a thicken node to give the garment some thickness. If everything is ready, I export the garment again as an ABC to Clarisse.  

What's really great is that with the point cloud and scatter tools in Clarisse, you can scatter tons of additional details on your garment, such as pearls, flowers, and other embroidery. And if I have an animated sequence, I do a quick playblast in Clarisse to check if there are visible intersections between character and cloth. Overall, the workflow between Clarisse and Marvelous Designer is tres chic!

What specific Clarisse features do you find most useful in bringing artistic visions like ‘KEEP IT NATURAL’ to life?

It’s simply amazing how quickly you can implement different ideas. And as I already mentioned above, you work directly on the image with no stand-ins. The scatter and point cloud functions in Clarisse are also very important to create realistic images. The hairgroom import from Houdini also works great. I’ll stop here, otherwise readers will think you bought my answers!

How easy did you find integrating Clarisse into your workflow alongside other creative applications?

Super easy. The geometry comes as alembic or VDB from Houdini. The texture packages I use have presets for the materials we use in Clarisse, so no guesswork. And with ACES, consistent color management is relatively easy to set up.


Are there any features that would you like to see implemented in Clarisse in the future?

Today, I use the MakeTX script in the Clarisse install folder to convert my textures in TX format, but a built-in TX manager would be great. I haven’t had time to play with the Angie Early Access version yet, so I’m not sure if it’s in there. A physical sunsky system would also be great. There is a version based on the nishita sunsky, but I hope that we can get a better one.

What project, personal or professional, is up next?

In addition to new still renderings, I want to do more with animation and short video clips!


What would you consider a dream project?

In terms of content, I don’t actually have a dream project. If the finished picture triggers a reaction in the viewer, then it would be a dream project for me. That doesn’t mean that everyone has to think it’s great, but maybe just that it makes you think, like a short film is running in the viewer’s head.


Would you recommend Clarisse to other artists doing character work?

Yes, absolutely! Since Clarisse has had the random walk mode in the SSS shaders for some time now, I see no reason why you wouldn’t use Clarisse. Also because of the shader! I almost forgot, the standard shader; Disney, Arnold, soon MaterialX, I mean, what more could you ask for? 

What advice would you give to aspiring digital and CG artists determined to bring photo-realism and style to their personal or professional projects?

I think light is very important, especially in the character story. You can totally mess up the best anatomy and geometry with bad lighting. It's like in real life, how the best looking model looks suboptimal in the wrong light. I try to keep the lighting very simple. What I never use are CG tricks, light without shadows, only spec or only diffuse. I assign HDRI (high dynamic range image) to texture most of the lights, like the beauty dish, everything from the photo studio. Those are the little things that I find very important. 

Thanks again, Isotropix. That was good fun! 

Thanks J for sharing your Clarisse experience with us. Character work is always a fun topic and something we don’t always get to see made with Clarisse, so your art always stands out to us at Isotropix. We love seeing your projects and hope there are many more #MadeWithClarisse to come in the future! 

Check out more of J’s Work: ArtStation, Instagram, Website