I never had to deal with any seams to be honest No matter how random they were, as long as texture padding was sufficient enough, they were never ever visible. The only time ever that seams mattered was when I was doing texturing in some software which did texture processing in UV island space, as someone above me mentioned, blurring in substance painter for example. But I’d honestly consider that a technical limitation which can ve solved.
I think it’s worth noting that you show a very particular type of asset there. There are assets on which you can afford to autounwrap and pack (although: with a better packer than built-in), but characters, other organic assets, things using trimsheets, assets featuring a lot of re-use, mirroring etc.
Manual unwrapping is still an important skill for those, and even on mechanical assets like yours I’d argue it’s important to at least know the basics so you can recognise a bad unwrap and head off any potential issues down the line, which in turn saves time.
Another good use of manual unwrapping (at least manually correct something generated by some automatic unwrapper) is when you create an asset for a moddable
car game/simulator. The “skin” must be usable in a 2d software in a way that stickers or letters will be easily undistorted and easily placed.
yes and this is what you ABSOLUTELY AND DEFINITELY should do too.
I give you an killer argument why:
I was an absolute shitty surface artist before 3D paintings programs came around. It would take me 7 hours to texture a simple asset and the quality would be straight up bad or maybe meh! and I would hate the process of doing it.
After learning Substance Painter (which took me half a day) I could do the same asset in under 1 hour and it would look good and it was fun doing it.
It made me an OKish surface artist just by changing the program.
How long did it take you to create the asset with Photoshop?
I could remake the same with painter in 15 minutes max if I push myself.
And in 20 I could probably create a version with more distinct details in it (without abusing vanilla smart materials that come with Painter)
I am not special and I am not implying your work is bad, but you could go through a similar transformation and increase quality and efficiency easily.
I can see the contempt dripping out from your comment. No hard feelings, but you should separate your negativity for the politics from the usefulness of workflows.
You surely know all this what I am about to write now, I just use your comment as starting point to explain stuff for others who read this (feel free to add, argue or simply ignore it).
One huge difference is coming from working in a team and working as a single dev/artist.
If you are alone, then you can do pretty much whatever you want, carrying the responsibility and problems on your shoulders.
Bigger Teams have dependencies. Often the UV’s are done by the modellers and the surfacing is done by separate artists. Things have to be streamlined in order to become practicable to manage.
That is not only true for teams but also single artists.
You brought up an example of an RTS game, where assets are small, compact and never seen from really close-up and therefore don’t have the same requirements than a model seen from eye-level in an FPS/TP game. Compared to these, RTS is quite different and that factor alone changes everything.
Its like Hero assets vs background assets.
The nature of RTS games means that you are doing basically nothing but background assets.
The more complex the asset and the higher the quality the more useless the Auto-UV workflow becomes.
UV’s are not only for texture mapping alone, they often drive other aspects, like VFX, damage, dirt, custom colours/textures/shaders etc, and Auto-UV workflow makes it much harder to implement these. Its setup/payoff. You have the “luxury” of being able to brute-force through the process, because for you there is no real payoff for doing otherwise. You truly would waste time for nothing.
But with high-quality assets there is a quantifiable payoff and the time you spare in the UV process creates more work downstream in the pipeline.
This is especially true for VFX pipelines for movies/TV.
What is spending an extra hour or two upfront vs a delay of possible days of work further down the line?
Context matters.
Also for people who want to work professionally, it is recommended to know how to do UV’s properly and to demonstrate this in your portfolio, no matter if you do it in your personal work or not.
It is not recommended to show off Auto-UV’s as it might imply that you don’t know better and can be construed against you, only show UV-maps that are nice and properly done.
Well, you are pretty much echoing exactly what I said. It’s mostly unnecessary for average indie artist, and somewhat necessary for an artist who works in high end production.
But especially high end production is probably space where it will be made obsolete next, because no one wants to pay person to do something an automated algorithm can, and those are getting better and better. If someone is a dedicated UV artist, they have some rough years ahead of them.
Bottom line, the name of the thread is:
Is manual UV mapping becoming Obsolete?
My answer is - Yes.
I generally don’t have contempt for people working in high end production, but I do have contempt for a specific type of people, whom learned absolutely bare minimum skills to get a job in the industry, do the same manual task all over again every day, without ever striving to improve a single bit above the bare minimum that’s required of them, and when they are the first ones to get laid off when the company gets restructured, they are the first one to call for some employee unions and crap like that
i just realized that OPs question used the word “becoming” and I totally ignored that and assumed a statement about the present. My bad, I can be retarded sometimes.
I am a little more pessimistic about it because I believe that UVing as a workflow could have been much more easy (since ages) if there was an incentive and investment to make it so not even using techniques that are based on modern machine learning / AI but rather other clever ways.
But it never came. It seems to me that the industry as a whole never considered removing the need for stupid manual work but rather used human brute forcing (and effectively wasting countless man-hours) instead.
So while I agree with you, I would put the point where it truly becomes obsolete further in the future than you might do. Just a feeling.
Do they actually exist? I’ve never ever heard of them, but yeah, sucks to be them.
Gotcha, but this is a very specific contempt, are there many that fall into this category? I assume that those are getting weeded out by the rather bad conditions in the industry considering the fact that these kind of personalities could easily earn more money and being more lazy in another kind of job.
On the other hand the 3D worlds needs unions or crap like that in order to offset the crap conditions in it.
They used to exist, but better, more automated UV tools made them obsolete They were usually the type of people I talk about just below. Not everyone of course deserves to be laid off, and the crappy conditions in the industry are not excusable. At the same time, some of the people suffering those crappy conditions are suffering consequences of their own choice to have weak, expandable skillset which they didn’t invest into improving.
Others suffer for unjustified reasons, such as simply having agreeable, non confrontational personality. I emphasize with those.
Like Retopo “artists”.
Then came ZRemesher/QuadRemesher and Wrap3.
I absolutely HATE unwrapping and the day it goes the way of the dodo I am opening and emptying a bottle of champaign.
I truly hope that is earlier than my pessimistic view on it.
Well, I am making some progress on my game and ever since I started working on it, I don’t think I had to manually UV a single model
I mean, aside from small things like unwrapping tank treads for UV animation or selecting and projecting faces onto decal sheets. But manual UV mapping of the entire mode, yea… that hasn’t happened for a few years at least
I guess I use a bit of a hybrid method: Manual seams and unwrapping followed by UVPackmaster to efficiently pack everything for me and maximise the use of the available UV space.
You can save lot of space by working in a modular way, or using sub objects instances that will share same UV space
Also 3D coat or Substance painter can import material ID, it’s a huge speed in texturing and allow to test materials variations in a click.
And you can bake the object materials in one atlas
About UV, Auto unwrap by angle, and manually change UV when needed.
Might I simply suggest that: “newer automatic unwrapping techniques and texture-painting techniques simply make ‘manual UV mapping’ much less necessary than it used to be.” Yeah, when you still have to do it – well, you still have to do it. But, thankfully, those cases are now much less common. (Yay!)
I think that “Less Important” is more appropriate word than obsolete. Also CG is big field. Are model pure boxy hard surface, or hard surface which have more curved/organic shape… or this are human/alien/creature ( how many tentacles, horns, spikes he have ).
What is purpose of this model… background object or some hero object which can be seen form close. Cinematic model or simple mobile model…
Anyway I think that inspection and manual tweaking are necessary.
PS. These days big companies/studios more and more use cheap outsource anonymous studios from… countries where salary is very low. Like on “Industry Veteran” say: I’m good and fast with unwraping, but I can’t match dude which 10-12 hours per day work only on UV. Can this be considered as some kind of “automatic” UV
It’s their employer who has to invest in it. I mean sure, it’s the artist who does the learning, but they should be cautious not to be tricked into learning it in their free time. (Not implying that’s what you meant.)