I have noticed that when building larger projects in ProBuilder the construction tolerance is not great. A simple plane is not completely level, though the rotation shows no tilt. I checked that the walls are of equal height, and the bottom edge is at zero, along with the plane, but clearly it is not. I have included a couple of images here for you to see. This is the same room, different sides - about 30 meters across the room from each other.
I have also noted that after creating many extrusions, some extrusions along some planes are not extruding exactly along their normals, but at slight angles. Now that I know this, I sometimes catch it, and can correct the angle, but at times, I have noted the error after much building.
I suppose I should just use another 3d program (Blender), but ProBuilder is very much more intuitive and nice to build with (when it behaves itself).
I am wondering if anyone else has noticed this issue. On DIscord, I mentioned this, but most people don’t do large-scale builds with it, or are probably not concerned with a truly accurate build since it is technically a prototyping tool.
It is a prototyping asset, not meant for production/developing, its been said here many times, lol. Personally i see why you posted the issue you ran into (alot of others have noticed it, and hopefully made a bug report on it), its all you can do.
PB is a good asset, and has its place. My question is: why would anybody want to invest alot of time with it?, or figure out probs with it… its just a prototype tool. You have better things to do with your time :).
I mean, I always saw that as an excuse for it being kinda bad.
Does it have features that make it really good specifically for prototyping, or does it do an okay job at everything until complexity goes up a bit and then things fall apart? If it’s the former, then you are right, if it’s the latter, then it’s just a bad feature that is only good to fool around with.
See also: Unity is said to be great at prototyping.
Doesnt mean its bad, and its not. With PB the higher complexity (like making a large level with it) it doesnt scale well, but then again its not supposed to. PB is great for what it does. Just prototype things out, then replace things with real game objects and develop your game. Can you use PB stuff in a real game, sure, hell i can make anything with a mesh/mat/texture.
Probuilder is there like a introductional tool to Unity. You can make floors/walls/ramps/etc. to rough out an idea. PB is a tool, Unity is good at prototyping and so much more, it depends on your usage of the tool, Unity itself, and your goal…
My point and question is, are there parts of its feature set that made trade offs in favor of prototyping and if they tried to make it a more fully featured tool, it would become worse at prototyping, or is it simply that no parts of it are fast or robust or well thought out enough, so it scales like ass, at which point the prototype label acts more like an excuse?
Like for many years it didn’t have the ability to remove edges (I don’t know if it has it now), was it because it’s a prototype tool, or was it because it is (was?) kinda ass?
Ive used it a few times because some of my students liked it, and it did help them prototype out ideas, but i also explained its not a game making asset or anything you should count on for production/development of a game.
I understand what your asking, but im not on the dev team to know what their plans are for it, as its not an asset i spend time working with, researching, and following the development of. I just know what its purpose is, and it does it well enough.
But i can tell you this: after class one night i decided to test it my way, lol. So i deved out a 1024x1024 terrain (flat) and started building out a full environment with it. The more objects i added (optimized objects) and the bigger the scene got, you could see and feel the impact. But once again why spend time on a prototyping tool, when you should be developing a game with real objects and real ideas… and lastly, real results.
Brother, we dont know their intentions with PB accept what they tell us. All we can do is report possible bugs, and hope they get solved, lol
LOL there are tons of models created with probuilder that are used in production. You’re supposed to export your models when you are finished with them. There is literally nothing that would make it “not production ready”. It creates models and those can be exported as fbx, obj, unity mesh, or whatever you need.
Is it the best choice for everyone? Probably not, but it has nothing to do with whether it game make production ready models. It is ALSO good for prototyping.
EDIT: The more they update probuilder the worse it gets imo. They keep trying to make it something it’s not and really just need to stop. I can’t stand the new shape creation tool.
People bashing it for some of the unfounded reasons they do is comical to me. Overall, it’s one of the most intuitive hard surface modeling programs I’ve used. My opinion is that people want to gatekeep 3D modeling
There may be some games produced with PB, and i agree its a great modeling prog. You can export models from anything, and as i stated PB is good, and has its place. PB like any other software will have bugs, but i dont think the intention of PBV is a game creator type asset. But devs will use it the way they see fit, nothing wrong with that…
I think there might be a miscommunication. We were saying the probuilder has been used to make models in many games. Nobody was referring building games soley with probuilder…it’s not a game engine like Unity. This is a (hopefully) obvious distinction.
You can make games in blender from what I hear. Why one would do that I don’t know.
The original post is about hard surface modeling with probuilder. It is perfectly fine for that (that I think we are mostly agreeing on and if we don’t, then that is also ok with me). I have no idea what caused OP’s problem they describe. It could be something just like the transformation rotation of a parent game object and have nothing to do with Pro Builder. I’ve done this before and was just like “whoops” going to have fix that etc.
There’s likely something OP was doing that caused this. They mention “large levels” which leads me to believe they “might” of been creating entire levels as one big game object (which is not a good way to go about it regardless of what modeling program you used for game levels). It’s much better to build modular pieces that snap together, then merge them according to geo location (using the same material of course). Probuilder can sometime do strange stuff when you get over 50,000 vertices for a single game object, but that’s on the upper end of the traditional threshold of 64,000 vertices.
I enjoy Blender for sculpting, rigging, and animating. There’s endless plugins for it and has an enormous community. The sheer magnitude of what can be done with a program like Blender is where alot of people get overwhelmed using it. I don’t know how or why I would, but I believe it is possible to make games to some extent using just blender (plus maybe some plugins? I don’t know).
I will say that the 2.8 + versions of blender really cleaned up the interface to make it more user-friendly. I think it can be risky to invest too much time into any single software these days. You never know what’s going to happen. It could be outdated in a few years or be replaced entirely. Because of that, I think it’s best to focus on transferable skills and concepts such as making 3D models that run very fast in realtime environments. It’s not what you use to make 3d models for gaming, but HOW you make them that matters.
That’s funny and quite an extreme stretch, particularly when talking about hard surface modeling. I’m almost tempted to challenge you to race doing hard surface modeling with custom uv’s. Runtime speed, design time, and aesthetics would count to see who wins. The final result has to be in unity. I think I’d smoke most people who have your opinion like that lol.
Is your entire model and environment on a scale the size of the room in the image, or is it a small part of a much larger model/ space, such that you have the room placed at a large-valued position (like 3000,3000,3000)?
