Hey there all,
I've got a question regarding UV space. Now to most of you pro's who've been doing this for ages and employed in the industry, this is probably an obvious answer, but I'm still very much a beginner so it's things like this I have to ask!
My question is should you always use all of the available UV space, regardless of resulting texture res on mesh components?
Whenever I do an unwrap, I try to keep all of my UV shells scaled to their mesh size/importance, and ideally keep resolution as consistent as possible. Sometimes this will mean I have a lot of leftover UV space, and I know efficient use of UV's is such an important thing to achieve with unwraps.
Here's a quick mesh I'm working on, it's a small saucepan asset that will be a very minor part of an environment. I've tried to explain what I mean in the unwrap. At first glance, it looks like the worst UV layout ever, but hopefully you get what I mean!
Many thanks
Replies
In general you want to keep pixel density fairly consistent in your entire scene from one prop to another as well.
If you don't mind a bit of mirroring and your engine can handle non-square textures, here's what I came up with for your UV layout.
Another option would be to shuffle things about to make a large open area then put another prop's texture in the empty space so they share the same texture maps. This can be done if the two props will always be together in the same scene. A knife or ladle perhaps. However if they would ever be separate from each other, then it would be somewhat inefficient to do that.
Utilizing shared UV's over a couple of assets likely to be grouped was a consideration I'd made, and in this situation that may work quite well. I guess the downside to construcitng assets and a scene the way I'm doing it for this piece is that a bit of guesswork is required as to how smaller assets such as this one will be placed in the final scene.
I always forget that I can use rectangular UVs in some cases too which should solve some problems.
Thanks again
So often you're really not saving anything by using smaller textures or a half sheet, unless you pack something else on the other half of the sheet, like another rectangle prop or retexturing the same prop to get a variant.
In my mind, if everything starts out pretty uniform, and you run into a situation like your example where you have obvious left over UV space, there is no rule that says you can't maximize your UVs for the sake of maximizing it.
As long as your important stuff is as big as feasibly possible, there's no reason not to waste UV space with your "lesser" important pieces simply for the sake of having everything a uniform scale.
This is true but I really don't think it matters for a portfolio piece and making a piece fit an artificial slot size when it isn't going into a game that has those sorts of slots seems like a total waste.
Additionally this is an environment prop so it is much less likely to fit into a pre-defined slot since those are typically for important objects (drivable vehicles, weapons, characters, etc), and instead share a memory pool with the entire environment, so making it very efficient is even more important since the more efficient it is the more clutter objects you can fit into the available environment mempool.
Also if this is a saucepan in a tiny kitchen scene I would seriously consider very heavy mirroring/tiling - unless the player is going to lick this thing or someone gets brained with it in a cutscene nobody is going to ever notice that it is mirrored/tiled along the edges, etc.
You can also quadrify whenever you can in order to squeeze out even more UV space:
For most of the straightening I tend to use TexTools Align options.
For major deformations however you can transform your UVs into 3D geometry and manipulate them with modifiers and modeling tools.Here is a quick mini tutorial on this:
If you're not familiar with TexTools, I highly recommend it ! You can download it from here:
http://www.renderhjs.net/textools/
I'm a big fan of TexTools and use the align tools a lot, I wasn't aware of the whole 3D UV thing though so thanks for that tip!
Is this just having the same checker patten size on all your objects?
I've heard having the same "pixel density" before but didn't really know what it meant.
autopsyS thats pretty much what pixel density is. there is a bit more to it tho if your interested.
http://boards.polycount.net/showthread.php?t=35204
generally you want to establish a pixel desity standard- that way when ever you make an asset you know what to shoot for, as opposed to just packing your UV`s so they fit your base page and having know Idea how that asset will compare to other assets in the scene. Saves you from having to go back and revisit things.
Yeah that would do it, given that you are applying the same size checker textures that your real textures would be and that the checkers themselves are a consistent size from texture to texture.
EDIT:
konstruct beat me to it, that's what I get for leaving threads open for a while before replying.