3D texturing for concept art purposes
null
Hey everyone, I'm new here. Decided to join and get some feedback and help on a subject that I've been struggling to grasp lately related to 3d texturing in general. I'm mainly a 2d artist but I've been trying to learn 3d as well to expand my skills and incorporate it more into my process. I'm interested in 3d and how it can be used in making still images or concept art. I'm not interested in creating models or characters for games at the moment though that would be cool to learn as well later on.
So my question is how is 3d texturing used in the still image or concept art creation process. How does the process differ from creating textures for games or movies. I've been doing a lot of research but most of the information out there on creating textures is geared towards creating textures for games. It seems like a complicated process that probably isn't used if I'm only interested in creating still images. Most of the tutorials I've seen also go directly into the how process without explaining what, when and why assuming I'm creating the textures for game models which isn't my goal.
Basically i would like to know if i should be bothered with learning retopology, UVing, baking and all that process. I'm guessing this is all related to creating textures for games mainly. So if all i wanted is to create high resolution models and texture them for final still images or to be used in photoshop later on, what would be the process? and which part of texturing would you do in 3d and which ones would you just do in photoshop?
Also which tools would be best in using for that purpose. Right now I'm learning zbrush and keyshot but I'm wondering what other tools would I need to help me with texturing. I've been using zbrush polypainting and zapplink to create some basic hand painted textures, but i'm not sure what other tools or workflows I should be looking into. How can i make use of bump, opacity, reflection and other maps? would a program like 3ds max, 3d coat or substance painter be needed or more useful to help with shading and texturing?
Hopefully what I'm asking made sense and someone can give me some basic information on how to use 3d texturing for still image or concept art purposes. Thanks in advance
So my question is how is 3d texturing used in the still image or concept art creation process. How does the process differ from creating textures for games or movies. I've been doing a lot of research but most of the information out there on creating textures is geared towards creating textures for games. It seems like a complicated process that probably isn't used if I'm only interested in creating still images. Most of the tutorials I've seen also go directly into the how process without explaining what, when and why assuming I'm creating the textures for game models which isn't my goal.
Basically i would like to know if i should be bothered with learning retopology, UVing, baking and all that process. I'm guessing this is all related to creating textures for games mainly. So if all i wanted is to create high resolution models and texture them for final still images or to be used in photoshop later on, what would be the process? and which part of texturing would you do in 3d and which ones would you just do in photoshop?
Also which tools would be best in using for that purpose. Right now I'm learning zbrush and keyshot but I'm wondering what other tools would I need to help me with texturing. I've been using zbrush polypainting and zapplink to create some basic hand painted textures, but i'm not sure what other tools or workflows I should be looking into. How can i make use of bump, opacity, reflection and other maps? would a program like 3ds max, 3d coat or substance painter be needed or more useful to help with shading and texturing?
Hopefully what I'm asking made sense and someone can give me some basic information on how to use 3d texturing for still image or concept art purposes. Thanks in advance
Replies
- Learn 3d sculpting in a software of your choice
- Like you were thinking, don't bother learning retopo
- uv can be helpful in some cases, but you shouldn't go into much details with it, try to rely on auto uv whenever its possible
- Learn a render engine of your choice. A realtime one would allow you quick and easy iteration, and serve a very solid base for your paintover later. It would also allow you to do color grading and other post process effects in no time.
- Don't go much into technicalities regarding traditional 3d modeling, in your case, topology, polycount and stuff like that really doesn't matter. You only need a picture that you can paint over.
- You don't need perfectly textured meshes, nor all the maps that are normally required in pbr environment. You can rely on photo sourced textures in a lot of cases.
At the end, if you do things like this, your workflow would look like:
- sculpt some meshes, apply basic uvs and possibly texture to it
- put it into the render engine
- make your layour using your meshes, apply basic shaders on them
- apply basic lighting in the render engine
- render / make a screenshot
- paint over a little bit where its needed
- profit
You can do it differently, this is just a guide of what proved to work. Also, I'm totally sure, there are a bunch of another concept artists using similar workflows to this, I came up with this guy, because I know him and I worked with him directly, so I know how he does it.
Here is the link to his work:
https://www.artstation.com/mgreuli
The top 3 rows of concepts he made using the 3dcoat to unreal -> bit of paintover workflow.
Hope this helps.
I still do concepts for personal work as it's so enjoyable and frees me from the constraints and technical workflow that you have to adhere to when creating producton assets.
The beauty of it is that you can be completely free, as there are no rules that you have to follow, only your own knowledge of art theories and your creativity.
My workflow is pretty standard:
Create models using whatever techniques you want, and I mean anything goes here. Topology isn't a concern. Jam primitives together. Use rounded edge shading. Boolean to your hearts content. Kitbash from existing models/kits. And if you want to be really sloppy you can even fix bad surface shading in Photoshop.
Don't bother with UV unwrapping. It's unnecessary. You can use standard box-mapping. You can use tri-planar maps. You can use floating planar mapped decal geometry with an alpha channel. You can add specific details in post during the paintover. You can use Lazy Nezumi.
Texturing can also be done very flexibly. But generally this is achieved with multiple render passes and paintover/photobashed textures.
Rendering is used to generate multiple shading/lighting/texturing/effects passes. These passes are then stacked in Photoshop and comped using mask painting and blending modes. There is no limit to what you can do here. You can just create many generic shaders consisting of different metals/plastics or whatever. Painted/rusted/grimy/scratched/etc. Anything at all that you require. Then create a few variations. It's all very flexible. Bumped/displaced/procedural/tiling. Whatever. At this stage you can also create multiple lighting passes: key/fill/rim/GI. And comp them however you like. You can also render AO, or a Z-depth pass for depth of field, or different shadow passes. You can even create emissive and refractive passes. And if doing characters you can output SSS and hair/fur passes. Or simply paint them in post.
The fun part is in the actual comping/paintover. Here you hace the ultimate freedom to let your creativity out. By using masks you can paint in/out lighting. Paint scratches on your mask to reveal metal beneath paint. Slap photographs of anything from concrete to hair to roadsigns to human eyes.........paint out the bits you don't want in your mask. Mess with blending modes. Again ultimate freedom. Use whatever images you like. You can paint smoke. You can comp in matte images of real smoke. You can paint highlights and shadows. anything goes to reach your vision. And it's all 100% non-destructive. At any time you can go back and just render out more layers to comp.
Here is a video I did a few years back explaining the process using a 3d concept project I did:
And this is pure gold. A Zsummit 2016 presentation from Peter Konig. It shows the beauty of this process:
Since you can be kinda sloppy when 2d concepting and you can paint over anything, let me ask this in another way.
Is there a difference between texturing for a 3d still image with minimal photoshop paint over vs texturing for games or animations?
Is the texturing process still the same?..as in you would still have to use retopology, UVing and bake textures or is this more specific to texturing for games?...which texturing methods or workflows apply for creating a 3d still image and which ones don't?
For example do you have to retoplogy a model before UVing or you can UV a high resolution model if you don't need a lower res model?..and where does baking fit into place if it's used in the texturing process of creating a 3d still image?
Obscura: Thanks, I checked out Greuli's artwork. Cool stuff
Musashidan: Thanks a lot for the detailed explanation. I will def. take a look at those videos. I also appreciate your free videos on youtube. I remember watching some of your zmodeler vidoes when i was first learning it. I'm still learning it actually..i have a long way to go..lol
When I was learning about shading and texturing most of the tutorials seemed to be focused on texturing for game assets, so it's a bit confusing. Just want to learn basic 3d shading and texturing for still image purposes.
I'm using Zbrush, keyshot and photoshop. Would like to learn Modo in the future as well.
I checked out your artstation portfolio..you have some cool stuff..nice work..I also saw your thread on WXS-07 'Depopulator' and you mentioned using zbrush, substance painter and vray.
Would you say you can achieve similar results using just zbrush, keyshot and photoshop?..is it just different workflow using different softwares?..does programs like quixel suite or substance painter offer any advantages as far as texturing for still images or are they more geared towards texturing for production purposes?