(VIDEO) UE4's Geometry Mode is inadequate when compared with that of Quake 1

[=;473213]
Yes and no, because the 2D support is limited I think, the last versions Cryengine got tools, but too bad for 2D viewports, bad design.
[/]

AS far as Im aware CryEngine only has one view port, and thats 3D. I dont see how 2D support would be limited in any what what so ever, its literally the exact same thing that any 3D modeling program offers.

If 2.0 comes out year - It would be great to have it working in VR.

Here is a video of what I mean

https://.com/watch?v=4Cro1HdCoP0

[=;489900]
If 2.0 comes out year - It would be great to have it working in VR.

Here is a video of what I mean

https://.com/watch?v=4Cro1HdCoP0
[/]

Yeah add VR support to Blender too… :cool:

Hi,

I just found discussion and I agree with the OP, it would be great to have some improvements on .

You guys who are “opposed” to have better geometry tools I don’t understand why is that.

Yes, we know that the best selling games have artists and the game is created with meshes and all that…

Now, why don’t you check the Steam Stats? Steam Charts It doesn’t matter when you do it, count how many players are playing Dota, CSGO and TF2 and compare the result to the rest of the “best sellers” using non-BSP’s.

I’m a Sketchup pro and you can’t even begin to understand how infuriating is that something that can be made in Sketchup in an hour can take DAYS with the current BSP tools, and I’m not joking, I had a recreation of my city’s old town and I cursed the cr*p out when I tried to do the same with just the editor.

So yes, I had to create meshes and import them one by one and then relocate them, exhausting.

Why choose a slower and more tedious workflow when you can just do process in one step, with one program and faster?

I mean, if the tools could be improved to give me just a third or a quarter of the speed I have drawing in Sketchup I would be still faster because I wouldn’t need all the import and placement boredom and I could apply materials right there!

I don’t think we’re asking to make Paint Photoshop, we want to make Paint better, that’s all.

Edit: Also, nobody is denying that static meshes have better performance, AFAIK once you are happy with the geometry you can transform it into static meshes, can’t you? Then why use an external tool at all?

I think that we should just go the Source 2 route and use static meshes instead of BSP, but still allow for the same flexibility in creation as BSP.

[=;442556]
I’m guessing you haven’t either because the vast majority of top selling games use very little to no BSP and have a focus on high quality graphics. Lighting and rendering aren’t the only thing that makes something look good, the environments in Battlefront are organic and complex, something that would be difficult if not impossible to do with BSP. If you think level design isn’t good any more, then that’s your. Games are making more than they did before so trends don’t support your views.
[/]

You do realize the biggest selling FPS of last year, Black Ops 3, is still made with a **** load of brushes right? Its more brush than it is models. Go check out videos for the alpha mod tools just released, I’ll wait.

My biggest complaint about Epic’s geometry tools is the builder brush, I don’t want to have to find the builder brush, scale it up, type in sizes, drag verts, etc and then “make” something. Its super slow. At my last (we used UE3) I worked in Maya for blockouts because the BSP is so lacking and I still didn’t like it despite knowing that program very well. When you’re making geometry within the editor you have so much more context of the level, where the spawn is, lighting and shadows, quicker iterations, quicker testing (PIE), scale is easier to understand, etc.

You keep bringing up performance, but there are many games being made with UE4 that aren’t pushing the processing power of the platforms they run on and nor will they. Not every team is out there trying to break the next graphics barrier, where losing half a millisecond when using BSP if of any great concern. I don’t think Epic would actively rebuff simpler games or limit its accessibility because you want to spend less time learning Unreal and more time learning Maya. That seems completely backwards to me.

As others have said before, even if you can argue that there aren’t dozens or hundreds of teams that would prefer the simplicity of BSP for their small mobile game (or their retro shooter, etc), the crux of the matter is that it still takes too long to blockout and iterate in Unreal.

@: I am on the same page with you (as I am sure many other devs too) and logically it makes sense to get these tools a go so that indies(and AAA-indies) can do small games or prototype quickly, for Android/iOS and mobile VR, but it seems that Android is at the bottom of the priority list (and most likely because UE4 hasn’t been used whole a lot by indies because Unity crowd paints UE4 to be some kind of monster when it comes to mobile; plus having Paragon in development doesn’t help several platforms anyway)

Ive watched thread for a year and a half be a fight for and against whether or not BSP should be used or not and waiting for Epic to develop a replacement (aka Geometry 2.0). Why not come together as a community and build the replacement, I’ve noticed trend of wait for Epic for features, with the occasional developer going off and doing it themselves ( and studios, etc.). I think there is enough skill in the community to work on and have a prototype up and going in a matter of months, once complete it could be submitted as a pull request to Epic or integrated with their current efforts with Geometry 2.0.

I know there has been no argument about Unity in thread, but I have noticed in others, people bring up the fact that Unity has tonnes of features that UE4 does not, is because the community is actively developing these options (albeit to sell on the marketplace). The same thing does not really occur in community, its basically submit a request and wait for Epic.

I understand not has the time to work on such a big project, but that’s why I think it would make a good community project, people who want to assist, can assist when they have time. And with the amount of feedback thread has received, I don’t think you will have problems finding help. Even I would be willing to lend a hand, even tho I have absolutely no interest in BSP level design, but I do have a good knowledge of the rendering backend of UE4.

Anyways thats just my 2c. Like I said, if someone starts a community project, I am happy to lend some of my time to get it up and running. Hell if there is enough interest in the concept, I’ll start the community project.

Wouldn’t it be easier to import Quake 1 map and convert Quake brushes into UE4 brushes? (Epic would still have to optimize BSP in general, but it’s a way less work than making whole set of BSP tools)

Trenchbroom was built by 1 very dedicated guy. It took years to get it where it is now. I doubt Epic has resources to do that (unless they just license the UI / workflow from Trenchbroom author). So maybe better run Trenchbroom / Radiant in parallel with UE4, and simply refresh imported .map inside UE4 to get updates?

One thing no one addressed yet: Special "nodraw" material to prevent triangles to be rendered - possible ? - Rendering - Unreal Engine Forums

[=;519609]
You do realize the biggest selling FPS of last year, Black Ops 3, is still made with a **** load of brushes right? Its more brush than it is models. Go check out videos for the alpha mod tools just released, I’ll wait.

My biggest complaint about Epic’s geometry tools is the builder brush, I don’t want to have to find the builder brush, scale it up, type in sizes, drag verts, etc and then “make” something. Its super slow. At my last (we used UE3) I worked in Maya for blockouts because the BSP is so lacking and I still didn’t like it despite knowing that program very well. When you’re making geometry within the editor you have so much more context of the level, where the spawn is, lighting and shadows, quicker iterations, quicker testing (PIE), scale is easier to understand, etc.

You keep bringing up performance, but there are many games being made with UE4 that aren’t pushing the processing power of the platforms they run on and nor will they. Not every team is out there trying to break the next graphics barrier, where losing half a millisecond when using BSP if of any great concern. I don’t think Epic would actively rebuff simpler games or limit its accessibility because you want to spend less time learning Unreal and more time learning Maya. That seems completely backwards to me.

As others have said before, even if you can argue that there aren’t dozens or hundreds of teams that would prefer the simplicity of BSP for their small mobile game (or their retro shooter, etc), the crux of the matter is that it still takes too long to blockout and iterate in Unreal.
[/]

Well that is more or less true and the old mod tools give a good tool (better than landscapes) since are for small zones a powerful tool to create walls and terrains, caves etc using the vertex

https://.com/watch?v=rhiWg1ajK3k

BSP tools are not that bad for its intended purpose.
Fixes i would like to see

  • Create (multiple issues. Creates flipped face, creates split flipped unflipped face, does not create face at all)
  • Faces are triangulated and loses verts resulting in geometry to be corrupt. Easy to fix by deleting face and recreating but considering above it can take a few tries
  • Brush Clip. Works pretty well except if split mode is enabled it occasionally deletes faces. See above 2 points.
  • No real grid snapping. Only object position based grid snapping (having real grid snapping will speed up bsp work)
  • align pivot to face (no such at present.)

I think that is about it.

Then again i only use BSP to block out the level and then start meshing.
there are so many mature 3d apps out there i honestly do not see why someone would spend the time and money on developing a 3d suite inside a game when the tools are already available.
But then again… thats just my personal opinion and Epic is mature enough to make their own decisions and justifications.

[=;442639]
I made handy list!

People in thread who are in favor of improved geometry tools as we’re describing:

me
dishwasher





rpotter
sonkim
soulroll
wanderer_eternal








arbopa
ruhton



taz0


skytram
**motorstep

**People in thread who are **against **or even just unconvinced **about **improved geometry tools as we’re describing:



errvald
kurylo



**networking

**25 to 7, yet the entire last couple pages has been argument, so you guys constitute an insanely vocal minority​.
[/]

Make that 26 … Unity even worse for built in level design tools.

I’m not a professional developer, but I’ve dabbled with a lot of level design programs in the past (Quake/Hexen/Unreal) and I have to agree that UE4’s current implementation is woefully lacking.

I understand now that for AAA games levels mostly consist of static meshes created in other 3D suites and imported in, but for quick iteration and testing you really need a fast, intuitive, and powerful BSP tool. It saves the level designer time and frustration which means work gets done quicker - who can argue against that??

I’d like to get back into level design after a rather lengthy hiatus, and while I will most likely still create assets in an actual modelling program, I’d very much prefer to be able to quickly create a rough draft of an idea, iterate on it, and then use the final result or parts of it to create more detailed static meshes. Probuilder for Unity looks exactly like what I want, though I’d rather get into UE4 for various reasons. Thankfully it looks like work is underway to creata UE4 version as well, beating Epic to the punch to something that should’ve been done ages ago.

So, not only Unity has Pro Builder, but also .

Kinda upsetting, but no surprise there are more devs/indies using Unity than UE4.

Hey ,

I recently came across addon for blender.

https://.com/watch?v=gM1N6WUQ_gc

That said - Give it a look at. It is very similar to Doom ED, DoomBuilder and SLADE.

Give it a look at - https://-lucas.itch.io/level-buddy

Edit: Sidenote - It is still in development. So There may be bugs or incomplete features.

Here lies a problem - you don’t want for your level to be made as single mesh. Even a single room better be made of several primitives, because of the way UE4’s culling system works.

So CSG is still better than Bender’s method. Btw, there is a plugin that imports maps made with Hammer / DarkRadiant / TrenchBroom. So far, it’s the only decent method of making CSG maps and getting them into UE4.

[=;544575]
Here lies a problem - you don’t want for your level to be made as single mesh. Even a single room better be made of several primitives, because of the way UE4’s culling system works.

So CSG is still better than Bender’s method. Btw, there is a plugin that imports maps made with Hammer / DarkRadiant / TrenchBroom. So far, it’s the only decent method of making CSG maps and getting them into UE4.
[/]

Alright - Interesting. Did not know that thanks :slight_smile:

Well if there was a vote as a content creator I would have to vote for better editing tools in UE4 but as a world builder adding editing over functionality to UE4 would be like trying to write a book using a spreadsheet app.

I feel that the better solution could be handled via the plugin route and create better DCC pathways through technology convergence and cloud base networking solutions and make use of the tools already being used by more than a few if not free DDC editing applications.

A good example

http://www.projectmessiah.com/x6/index.html

Messiah can make use of an FBX service to move keyframe animations back and forth between target and host applications and there are a few applications that takes advantage of the “send to” pathway so just like you can author animation between Messiah and UE4 the same DCC pathway could be used with say a Hammer and UE4.

So, what’s wrong with making maps in Hammer/DarkRadiant/TrenchBroom and instantly importing them into UE4?