Sergio Raposo Fernández discussed his recent fantastic work and shared his modeling and UVs workflow.
I am a freelance artist from Madrid, Spain. At the moment, I work as a freelance, but I would like to work for any studio and collaborate in feature animation films. I studied 3D in Trazos Academy, in Madrid. But all my life I’ve been learning painting with my mother, a wonderful traditional artist.
My goal in this project was to learn Mudbox and Substance Painter, because I had never used them, and to improve the result of my artworks. For the theme, I choose a magical interior with a lot of details and in-cartoon style. The most important reference for me in this project was the Gargamel´s tower in the film "Smurfs, the lost village" by Sony Pictures Animation because I love the visual style of this movie. Also, I used Pinterest to find images that inspire me, especially medieval interiors to know what was the style of the furniture and the architecture of that age.
I modeled the whole scene in Maya, and then I exported objects to Mudbox to sculpt them, for example, the walls with the bricks, the furniture or any assets, as the cauldrons, the books or the bottles. Then, I made the retopology of all these objects in Maya with the Quad Draw tool. The workflow was tedious and hard because it is the first project, where I used Mudbox. The topology of all objects is clean, without triangles.
The UVs were made in Blender. Really, I am constantly moving between Blender and Maya, exporting assets in OBJ format according to my needs, because they both have tools that I like. And for the UV work, I like Blender more. But for modeling, I prefer Maya, because I can use the Wacom tablet for modeling when I can’t use it in Blender, and modeling with the mouse is impossible.
Once the retopology is done, and the UVs of the objects in low poly, I start texturing in Substance Painter. First of all, I bake the textures from the high poly mesh to the low poly mesh, always in 4k resolution. Talking about the textures that I use in my materials, they are always seamless, and I usually download them from this website. In order to make the different maps, like the normal map, I use Crazybump software. I use the default smart materials often, and then I modify them. I have learned to create materials by studying the existing smart materials, analyzing them. But if I create a new material, I begin by creating a fill layer and adding different textures in the Properties-Fill panel. There in the projection options, it’s very important to select Tri-planar projection to get a material without stitches.
Then, I usually add a black mask to this fill layer with a default smart mask. Substance Painter has a lot of smart masks, so I rarely make my own mask. I only have to play with the smart mask values to get the result that I want. I add such filters as "HSL Perceptive", or "Levels" to change the properties of the textures, the color or the contrast, for example. And I use "Alphas" and "Hard Surfaces" to make stamps, for example, for the book covers (view image attached). And I used Particles to make the liquid on the cauldron (view image attached). Finally, I added dust to all the objects, but in Blender, because when you rotate the object, the dust changes its position (not like in Substance, because it creates a static texture which doesn´t change if you rotate the object).
The render was made in Blender with Cycles. I love Cycles because the nodes system offers a lot of possibilities. There is only one thing I don’t like: it is slow. But now, Blender has a very useful option called denoising, which cleans the render, and it makes the process faster. For the previews, and while I was working in the scene, I used GPU render, but the final render was made with CPU because this scene is very big for my GPU. And the lighting is very simple, with an HDRI. I always use HDRIs for the lighting because it is more realistic. In the windows of the tower, I put Portal lights to reduce the noise and to get more clarity. Finally, I choose a PNG format with Alpha background, and I added to the scene a different background image, more according to the scene. The post-production was made in Gimp by adding volumetric rays, correcting colors and adding depth to the field and fog.