Basically, the guys we had in house on staff were just focused on low poly and fixing bad topo.
Max, Maya, Zbrush and substance painter all made auto-magic re-topology tools. Whatever that shit out went through minimal editing before it was baked and considered good. Lots of broken "difficult" areas like bad topo around tits, bad topo around the corner of lips. Weird as hell lighting artifacts thanks to that.
You know how you'd have topo that is like /\ in the middle of a symmetrical line? If the retopo pass made it / / they'd not even bother to turn the triangles. Oh, giant 8 edge n-gon? Fuck it. It's flat. Surely the engine will get it right. Let's not solve that pre-emptively!
The dumbest topo tool is zbrush though. Smear red and blue to indicate where you need more texture density and hit the easy button. Others in the office didn't even know you could premark texture seams to make the tool at least a little more intelligent.
I first learned Maya back in the day when I was a little noob and didn't know a thing about 3D then I made the switch to Blender. I'm not surprised by the poor technique ( personally I know about good topology in relation to quads ) but what frequently surprises me is how studios are so stubbornly sticking with completely outdated software with their game dev.
It's as if the CEOs can't be bothered putting the money in to train their staff properly or overall they're just too lazy to make the switch even if it would really benefit them. I'm somebody who had to make the switch from Unity to Godot but it was worth it. Yet these morons even for new projects seem to refuse to learn new software and insist on sticking with their horrible buggy internal engines or whatever crap they've been using since almost 20 years ago.
As somebody who does take pride in my work, I simply can't process that level of laziness and incompetence because the only way you'd let any of this go on is if you're an arsehole that doesn't care about your own work.
Most people heading up studios had a backround in QA or production. So management has a familiarity and connection to the tools that are legacy. They can easily call bullshit when a hireling starts saying "Oh, my texture bake for a gun is going to occupy my computer for the next 8 hours. I'm going to leave it to cook, see you tomorrow boss!"
That's the big reason. Getting new tools for the staff would mean having to learn themselves.
Another thing is the "good enough, lets just get it done" mindset. That'd have a guy struggling and grunting to drive in some wood screws by hand because he doesn't want to climb through a messy garage to retrieve a powerdrill.
Theory of mind issue; you're thinking if you can do it then they can do it. Consider yourself a unicorn.
Linux has ALWAYS been an option. Every time windows is looking to impose a really shitty change, people threaten to go to linux. Do they though? Or they threaten to change cloud services... but they usually stay on azure.
You could have a manual transmission offer 30% better gas millage. People would still stick to automatic because they just don't have the mental bandwidth to learn manual after a lifetime of automatic.
Not justifying it, just saying, humans preserve brainpower whenever possible.
That's the big reason. Getting new tools for the staff would mean having to learn themselves.
Pure laziness as far as I'm concerned but your information is very valuable and it explains a lot of what's going on. If I'm a single indie developer and can switch engines so can they, they've got no excuse. It's not just that this software is legacy, it's utterly deprecated by today's standards, moving from Unity to Godot for example with my own experience I was shocked at the amount of bloat Unity had by comparison. Even simple things like loading up the damn project took seconds by comparison and it was purely because of the amount of junk Unity has in it's back end. That's a whole other topic I rant about regularly though with regards to game engine choice.
I mean for fuck's sake, an example of a fantastic feature Godot has been looking at introducing is the idea of a drop down selection for multi-threading. This means you can have individual objects in the game running off separate cores as opposed to the first one with the click of a button instead of having to go into the code yourself and write up the multi-threading behaviour which is remarkable. I'm going to have to do some testing on it when I get my project properly playable.
Legacy is legacy, y'know? More ready made tool and scripts, tutorials and video learning. What makes them better than blender is just the fact that there's more money behind it. Whenever blender comes up with something truly unique, the major players just copycat them.
Opensource tenders to be by the people, for the people, but like most projects that are quasi-communism, there is no real motivation to have high production values in training seminars or said training in the first place.
If you are a self guided learner, this is fine. If you have the free time, this is fine... but where commercial software wins is when the industry makes a major shift. Like in recent memory, the change to PBR materials. When you have to change the entire way you bake materials to accommodate ray tracing or change the way you model something to accommodate nanite, there will be professional training in person you can buy from one of the big 3.
With blender. . . you don't really have that crack team of "We need to train the entire studio on nextgen pipelines THIS MONTH" groups.
Basically, the guys we had in house on staff were just focused on low poly and fixing bad topo.
Max, Maya, Zbrush and substance painter all made auto-magic re-topology tools. Whatever that shit out went through minimal editing before it was baked and considered good. Lots of broken "difficult" areas like bad topo around tits, bad topo around the corner of lips. Weird as hell lighting artifacts thanks to that.
You know how you'd have topo that is like /\ in the middle of a symmetrical line? If the retopo pass made it / / they'd not even bother to turn the triangles. Oh, giant 8 edge n-gon? Fuck it. It's flat. Surely the engine will get it right. Let's not solve that pre-emptively!
The dumbest topo tool is zbrush though. Smear red and blue to indicate where you need more texture density and hit the easy button. Others in the office didn't even know you could premark texture seams to make the tool at least a little more intelligent.
I first learned Maya back in the day when I was a little noob and didn't know a thing about 3D then I made the switch to Blender. I'm not surprised by the poor technique ( personally I know about good topology in relation to quads ) but what frequently surprises me is how studios are so stubbornly sticking with completely outdated software with their game dev.
It's as if the CEOs can't be bothered putting the money in to train their staff properly or overall they're just too lazy to make the switch even if it would really benefit them. I'm somebody who had to make the switch from Unity to Godot but it was worth it. Yet these morons even for new projects seem to refuse to learn new software and insist on sticking with their horrible buggy internal engines or whatever crap they've been using since almost 20 years ago.
As somebody who does take pride in my work, I simply can't process that level of laziness and incompetence because the only way you'd let any of this go on is if you're an arsehole that doesn't care about your own work.
Most people heading up studios had a backround in QA or production. So management has a familiarity and connection to the tools that are legacy. They can easily call bullshit when a hireling starts saying "Oh, my texture bake for a gun is going to occupy my computer for the next 8 hours. I'm going to leave it to cook, see you tomorrow boss!"
That's the big reason. Getting new tools for the staff would mean having to learn themselves.
Another thing is the "good enough, lets just get it done" mindset. That'd have a guy struggling and grunting to drive in some wood screws by hand because he doesn't want to climb through a messy garage to retrieve a powerdrill.
Theory of mind issue; you're thinking if you can do it then they can do it. Consider yourself a unicorn.
Linux has ALWAYS been an option. Every time windows is looking to impose a really shitty change, people threaten to go to linux. Do they though? Or they threaten to change cloud services... but they usually stay on azure.
You could have a manual transmission offer 30% better gas millage. People would still stick to automatic because they just don't have the mental bandwidth to learn manual after a lifetime of automatic.
Not justifying it, just saying, humans preserve brainpower whenever possible.
Pure laziness as far as I'm concerned but your information is very valuable and it explains a lot of what's going on. If I'm a single indie developer and can switch engines so can they, they've got no excuse. It's not just that this software is legacy, it's utterly deprecated by today's standards, moving from Unity to Godot for example with my own experience I was shocked at the amount of bloat Unity had by comparison. Even simple things like loading up the damn project took seconds by comparison and it was purely because of the amount of junk Unity has in it's back end. That's a whole other topic I rant about regularly though with regards to game engine choice.
I mean for fuck's sake, an example of a fantastic feature Godot has been looking at introducing is the idea of a drop down selection for multi-threading. This means you can have individual objects in the game running off separate cores as opposed to the first one with the click of a button instead of having to go into the code yourself and write up the multi-threading behaviour which is remarkable. I'm going to have to do some testing on it when I get my project properly playable.
I do some hobby 3d rendering in blender. How does that tool compare to the professional ones?
Legacy is legacy, y'know? More ready made tool and scripts, tutorials and video learning. What makes them better than blender is just the fact that there's more money behind it. Whenever blender comes up with something truly unique, the major players just copycat them.
Opensource tenders to be by the people, for the people, but like most projects that are quasi-communism, there is no real motivation to have high production values in training seminars or said training in the first place.
If you are a self guided learner, this is fine. If you have the free time, this is fine... but where commercial software wins is when the industry makes a major shift. Like in recent memory, the change to PBR materials. When you have to change the entire way you bake materials to accommodate ray tracing or change the way you model something to accommodate nanite, there will be professional training in person you can buy from one of the big 3.
With blender. . . you don't really have that crack team of "We need to train the entire studio on nextgen pipelines THIS MONTH" groups.