Pathetic questions. Please help.
Hello Polycount.
I am pathetic. This is not my real account. I am trying to practice up my 3D so I can apply for a specific 3D game artist job that I'd really like. I have two problems, one is that I need to learn 3DS Max asap (I'm a Maya user primarily) (I am working on this by watching Eat3D tutorials).
The other problem is that my brain does not seem to comprehend some of the most basic strands of 3D knowledge, and I have a few questions.
1) I (want to be) an environment artist. How does one go about unwrapping walls, sides of buildings, roof tops? I get that a tileable texture could work, but how do the UVs look for something like that?
Here is an image from a hand-painting texture tutorial.
He makes a horizontally-tileable texture, and it works well. How is that unwrapped? Is that cylinder (thing) unwrapped flat in a sort of long rectangle? Does it not matter if all the UVs aren't in the 0 to 1 square?
2) If it's a really big wall, and I slap on a tileable texture, do I just unwrap it to whatever and then increase the tiling/repeating of the texture in 3DS Max/Maya/UDK? Is that how it goes?
3) Modular pieces and texturing. I have looked at a number of UDK/UT3 textures and I notice that usually a small collection of assets all use the same texture file and each have their various spaces assigned to them. How does this workflow work..? How can you plan where all your UVs are going to go in advance like that? Especially when you take into account baking normals etc.
4) Philip K's 'Working with Modular Sets' tutorial. This confuses me but I don't know how to word it. This part I can word though, why does he do this step?
Here is the link, it's near the top: http://www.philipk.net/tutorials/modular_sets/modular_sets.html
He makes a high poly version. But then goes back to his low poly? What if he decides to change something to the low poly, won't that make the high poly completely redundant? Also what if he makes another piece, won't it be hard to try and emulate that high polyness again? This is probably very confusing to read, I'm sorry.
5) If you make the low poly, and then the high poly, and you bake it and everything is fine, but then later you decide "hmm I want to add a dent in here actually and maybe bend this a bit". Won't that completely destroy the compatibility of the high poly? That means you can't ever re-bake the normals, because the low poly is now a different shape, no?
6) I have seen people have a large texture atlas that provides textures for an entire scene. How do they plan this? Do they make the separate textures first and then stitch it together in the end? But doesn't that mean all the UVs would be all over the place?
7) How does one create a crate nowadays? A few years ago I would have made a cube with a few bevels and insets, then unwrap and texture. Would you now model each plank separately, sculpt, import high poly into max, create crate using the high polys, and then bake that onto a low poly crate? Or what?
These are all the questions I can think of at the moment. I have more, unfortunately, but I will save those for another day when I can remember them.
I love looking through all the environment art threads, but I am really scared that I cannot produce anything decent lately, now that I am out of school and need to get a job.
If you managed to read to this point, then I thank you for your time.
Best regards,
Pathetic
I am pathetic. This is not my real account. I am trying to practice up my 3D so I can apply for a specific 3D game artist job that I'd really like. I have two problems, one is that I need to learn 3DS Max asap (I'm a Maya user primarily) (I am working on this by watching Eat3D tutorials).
The other problem is that my brain does not seem to comprehend some of the most basic strands of 3D knowledge, and I have a few questions.
1) I (want to be) an environment artist. How does one go about unwrapping walls, sides of buildings, roof tops? I get that a tileable texture could work, but how do the UVs look for something like that?
Here is an image from a hand-painting texture tutorial.
He makes a horizontally-tileable texture, and it works well. How is that unwrapped? Is that cylinder (thing) unwrapped flat in a sort of long rectangle? Does it not matter if all the UVs aren't in the 0 to 1 square?
2) If it's a really big wall, and I slap on a tileable texture, do I just unwrap it to whatever and then increase the tiling/repeating of the texture in 3DS Max/Maya/UDK? Is that how it goes?
3) Modular pieces and texturing. I have looked at a number of UDK/UT3 textures and I notice that usually a small collection of assets all use the same texture file and each have their various spaces assigned to them. How does this workflow work..? How can you plan where all your UVs are going to go in advance like that? Especially when you take into account baking normals etc.
4) Philip K's 'Working with Modular Sets' tutorial. This confuses me but I don't know how to word it. This part I can word though, why does he do this step?
Here is the link, it's near the top: http://www.philipk.net/tutorials/modular_sets/modular_sets.html
He makes a high poly version. But then goes back to his low poly? What if he decides to change something to the low poly, won't that make the high poly completely redundant? Also what if he makes another piece, won't it be hard to try and emulate that high polyness again? This is probably very confusing to read, I'm sorry.
5) If you make the low poly, and then the high poly, and you bake it and everything is fine, but then later you decide "hmm I want to add a dent in here actually and maybe bend this a bit". Won't that completely destroy the compatibility of the high poly? That means you can't ever re-bake the normals, because the low poly is now a different shape, no?
6) I have seen people have a large texture atlas that provides textures for an entire scene. How do they plan this? Do they make the separate textures first and then stitch it together in the end? But doesn't that mean all the UVs would be all over the place?
7) How does one create a crate nowadays? A few years ago I would have made a cube with a few bevels and insets, then unwrap and texture. Would you now model each plank separately, sculpt, import high poly into max, create crate using the high polys, and then bake that onto a low poly crate? Or what?
These are all the questions I can think of at the moment. I have more, unfortunately, but I will save those for another day when I can remember them.
I love looking through all the environment art threads, but I am really scared that I cannot produce anything decent lately, now that I am out of school and need to get a job.
If you managed to read to this point, then I thank you for your time.
Best regards,
Pathetic
Replies
With regards to the texture atlas thing as shown in the UC pic you linked, those textures aren't actually joined together. You are allowed 4096x4096 overall in that case, so that can be a combination of various smaller textures, the artist has just put them together in a 4096 square in the picture.
Having said that, you can of course include different tiling textures within the same 0-1 space, and where possible, you should. For example, It's less expensive to call 1 larger texture than 4 smaller ones if you see what I mean. Generally speaking you would create these textures first, and then UV your model to make best use of them.
It varies on a per-project basis but all modular/tiled approaches require careful planning and clever use/reuse of texture space
1. If you look closely at the geometry of the tower base you can see that it is not perfectly round and has some edges. Id say that the artist used over lapping UVs for the base. In fact, if you look even closer you will see where the texture repeats (look for a specific crack and then see where it repeats itself). That is usually an indicator that UVs have been overlapped.
4. This is the process of normal mapping, where details from a high ploy model are baked to a normal map which is then applied to the low poly. The variations in color on the normal map determine the way light is reflected back from the surface giving the appearance of a higher polygon object.
In PhilipKs tut, it looks like he built the low poly model first, and then the high poly model. Whats not shown in the normal baking process.
5. If you make changes to the high poly geometry that you want reflected in the low polys normal map you will have to re-bake the normal map. Also, if you make changes to the low poly model after the normal baking process you will need to re-bake the normal map.
S6
so yes, alot of times your UVs will not be the nice tidy layout you see with characters: