The making of 'Bear TV'
Discover how Brazilian character artist Leticia Reinaldo created her moody
image, Bear TV.
This is a project I created for a class at the Gnomon School of Visual Effects. The goal of the project was to create a finished piece in 2 weeks (1 week modeling and UVs and the other week for texturing, lighting and composting). I chose this amazing concept piece by Cory Loftis because his work represents what I like most about doing in 3D – which is telling a story.
In this tutorial I want to show a bit of my pipeline and do an overview of the steps used in the creation of this piece. I really enjoy the entire process of creating a piece but the thing that I pay most attention to is conveying a GOOD story (funny, emotional, sad, happy, dark, and so on).
Choosing the concept
This first moment is very, very important! You really need to take the time to choose and analyze your concept before deciding to start a new project. Try to pick subjects that interest you (especially if it is a personal project). The importance of a good story is that people connect with your piece on a higher level and the technicality is forgotten. This is your chance to tell something through your artistic skills, so use it!
Studying the concept
After you have picked your concept, it is time to analyze the props, characters and details. The trick is, before you begin, just zoom in, go through your image and slowly start getting to know it. Walk around it with patience. It is also good at this time to have a notebook and start taking notes about the props and any ideas that cross your mind about how to approach specific objects. The more you familiarize yourself with the concept the better the creativity will flow.
Also, begin to gather references – you're always going to need them! With the list of props from your scene, it is easier to know what to look for now, so collect as much as you can. Try to get photos, drawings, maquettes, 3D pieces that inspire you or are related to the subject you are researching. Also try to be as organized as possible. Also, don't only think about objects, but also collect ideas for lighting, the color palette, mood, anatomy studies etc.
Let's start modeling!
For this piece, I decided to model in a type of zombie pose with the arms facing the front, because in the end it would be easier to match the pose from the concept.
In ZBrush I started creating ZSpheres with the overall volume of the character, focusing on basic proportions. After this part was done, I converted it to a mesh and started blocking the shapes, silhouettes and finding the apexes. Try to keep symmetry on as long as possible at this stage. After nailing the basic silhouette, I started to add secondary forms, building up muscle definition and transitions of sharp to soft forms.
Creating topology and UVs
While still in ZBrush, I decimated the model and brought it into Maya to start creating topology for better posing and animation. I used the quadraw tool (from the new modeling toolkit in Maya 2014), to define the flow of the loops. After the topology was done, it was time to start the UVs. I normally use Maya to create the UVs, but you can also use ZBrush for faster but less accurate results.
With the topology and UVs done, I went back to ZBrush to start posing the character in the chair with Transpose Master. This stage is very important from a storytelling point of view – the pose plays a big role in telling who our character is. Pay attention to contraction and compression, try to push the pose even further than the concept, and spin your model while posing to see if it reads well from all angles. Try to study photos of people in that pose or ask a friend to pose for you and take a photo. All this will help, as the more subtleties you add the stronger the character will be.
With all the stages presented before being done, I exported the objects as OBJs and sent them to Maya. I decided to use Mudbox as my main texturing software because of the ease of transitioning between Maya and Mudbox. In Mudbox, for the bear texture, I used the bw_bistrol2 stamp to paint some of the fur, changing the color constantly to create variation. This variation would help in the future with the V-Ray fur. For the props, I normally paint a lot of my textures by hand and at the very end overlay various images to amplify the effect.
This is the shader connection I normally use for cloth materials. Some types of cloth have a fuzzy matte feeling, and to get that effect in Maya I usually use this method.
Basically, you take a shader and instead of putting the texture on the diffuse color you plug a ramp node. From the ramp node you connect a sampler info node as shown in the image below. The idea is to have 2 colors on the diffuse and with the sampler info node you get the facing ratio, meaning that the top color will be facing the camera and the bottom color facing away. This, along with a rough specular and reflection, will give you the fuzzy feeling most of the time. Basically, we are cheating the physically based material attributes to behave the way we want.
For the fur I wanted to get a simple and very cartoony feel. I experimented with some Maya fur but couldn't get it to work the way I wanted and ended up deciding to use V-Ray fur. The V-Ray fur has very little control, which was a good thing for me because I was running out of time to finish the piece and needed something fast and simple.
To get a bit of complexity in the fur I decided to create several pieces of mesh, each of them with a different V-Ray fur assigned so I could control the randomness and direction of the patch. Also, I created some different fur in the same mesh, some with less fur but thicker and some with more fur but thinner. I used the V-Ray hair shader for it and plugged the texture painted in Mudbox to create the color variation mentioned in the texturing part. The fur's color is dictated by the diffuse of the underlying mesh that it references.
The lighting in this scene is very simple but carefully placed. Besides the character, the lighting is another important element that could be used to amplify the mood and story of the image. This scene could be very different if the overall mood was brighter and had less contrast.
At this stage, it's a good idea to research scenes in movies, other artists' work, photos, paintings and anything that has a similar mood that you wish to project. Experiment a lot to find the best way to tell the story (I'm repeating myself again, because this is important!) So, I decided to leave the scene very dark with a primary source of light coming from the TV, a secondary light from the lamp and a big fill light from the bottom left side of the image.
The compositing part is always fun! It's the time to push the lighting, add a bit of detail and fix some weird things that may have appeared during the render. I always generate a lot of passes which allows for more control throughout the compositing process. Multi matte passes help a lot for quick selections in Photoshop or NUKE.
The basic passes I always get out of V-Ray are in the image below. Basically, you will put them all in Add mode, except the AO pass which will be in a Multiply Blend mode. Clip the adjustment layers to your passes for a non destructive way of working. After all the adjustments, I like to top it off with a dust texture to give some atmosphere to the image. I finish off adding a lens correction filter and with this, you can add vignetting and chromatic aberration which helps to give a photographic look to your piece.