Possible to get Unity3D on par with Unreal in regard to landscape design and quality?

Hi everybody,

I'm not sure if the title does justice to Unity3D, really. But my situation is as following: I own a Unity3D Pro license, like it quite a lot, and consider myself rather fit in developing with Unity and C# (though I only completed big 2D projects, so far). At the same time, I now got a client who wants me to realize his dream of an "awesome", artistic, surrealistic, alien-world, VR puzzle adventure. I think it will take a direction kind of like the recently released new Myst-game "Obduction" (which was created in UE4), but lesser in scope (since mostly only my client and I will be working on it), more "alien", and it will be VR-only.

Now, artistic work (modeling, sculpting, animating) will be done by my client, programming and putting everything together will be done by me.

Judging by the superficial impressions I got, I'm afraid the Unreal Engine 4 might be much better suited to do the job. If I am not mistaken, it would be easier to achieve the "awesome" looks (as much as that is possible in VR, yet) and to create impressing surrealistic landscapes with as little effort as possible. For example, just recently I saw how easy it is to distribute particles like dust or snow realistically in cracks and dents just by pulling a slider, and how much landscaping can be done in the editor itself. And my client is putting heavy emphasis on it being an atmospheric experience of a vast world, even more than it being a game.
It's not going to get mobile ports, so this won't be an issue either...

But on the other hand I consider myself a Unity C# Dev, which is something I'd like to deepen my knowledge in. It would probably take me a lot of time to get used to C++ (provided UE Blueprints won't suffice for programming the puzzles and interactions) and to the Unreal Engine itself, whilst I'd much rather would like to get another Unity project under the hood. And then there are the 5% royalties with UE, as well.

My question to you is: Would it be possible, recommendable and responsible (towards my client) to try to achieve what we are intending to in Unity3D instead of the Unreal Engine?
Are there extensions in the Asset Store that get Unity kind of on a par with the Unreal Engine in terms of creating vivid, impressive, atmospheric landscapes and looks, as efficiently as possible?

 

Sadly stock Unity built-in terrain is years behind most of the other engines and there is an official feedback thread buried somewhere which list a lot of the weaknesses and not many of them have been fixed or improved and the whole thing has been in the roadmap for ages like an abandoned card. Asset store has a lot of stuff that will fix the gap and even make better but naturally not without a cost. Example:

 

From a purely artistic aesthetic you can't beat unreal engine, especially for open world-ish games.

The scattering on the leaves, and better lighting/shadows make the difference.

 

It ultimately boils down to what you want to create and if you have the skills, expertise and resources to realistically achieve that vision.

Could you explain 'Landscape Design' a little bit more? With an example?

 

Outside of Mobile and 2D, Unity is grossly behind Unreal in pretty much every area.

You have to scour the depths of the Asset store to fill in all the holes, but it's possible.

 

We'll there are other ways to approach rather than use the in-built terrain system. Just use GPU instancing with mesh painting tool for foliage, using meshes with a stitching algorithm when you export tiles from world machine. You can still either hand paint / use colour maps or splat with meshes.

Quite a bit more work, but if you want better looking terrain in Unity that's the way to go.. I have some examples of terrain experiments I did in Unity:

http://forum.unity3d.com/threads/shadows-stuff.370855/

Personally, if you do it right then it's a toss up between the two.. Unity is just a lot more effort, in other areas though it's far less effort. So I'd invest some time into research.. I will say Unity deals with far more foliage in deferred rendering better than Unreal..

Unless you want to tackle the rendering pipeline in Unreal?!

 

Spending time in unreal 4 and Unity, I am actually using Unity to generate my Unreal height and splat textures.

I am using Terrain Composer, but I really want World Machine Pro, for high quality procedural terrains, the key is..erosion, and high quality textures.

So I tried Erosion Brush, for $25 I am getting good results,

Spoiler: Unity Terrain with splat textures ONLY.....512 height/splat...really!!

You need really good splat maps, having high quality textures is paramount.

I also have RTP Shaders (not used in above image), that also is pretty important, having height blending and texture packing, for 2-pass 8 splat renders!

Map Magic, Gaia, also look promising..but make sure there is procedural erosion of some sort.

So in short....Yes!, Unity can create some pretty fantastic worlds!!

P-

 

A foliage study not bad at 4FPS

 

You're on a Mac. It's expected.

 

Based on your screenshot, it looks like the CPU is the bottleneck. You should investigate with the profiler to figure out what is using so many CPU cycles.

 

My understanding was that Unity's VR support beat Unreal's. I might be off base here. But it's worth thinking about more then just visual capacity. You might get a better visual out of Unreal, but it might not be able to run on your target device.

 

They're bringing improvements to their virtual reality support with every release (just look at the latest link below for a few examples). If they are behind Unity then they won't remain that way for very long.

https://www.unrealengine.com/blog/unreal-engine-4-13-released

 

Ask you client for some test art, make up an environment example in both Unreal and Unity and show it here, and maybe some experienced people can tell you how to bridge the gap and what can and can't be done in Unity. For a project of this magnitude, foregoing experimentation would potentially cause you a lot of trouble.

Personally, I don't think Unreal can be beaten in graphical quality, but I'd love to be proven wrong and I still hope to be in the near future. However, you have to take into consideration the potential difficulties of Unreal (which I know nothing about - you'll have to ask someone experienced in it) and how Unreal will compare in all the logistical aspects of game production (workflow/time spent doing stuff/engine updating etc) - which Unity have pretty much down pat as far as is possible from what I can tell.

Also, if your client is an experienced artist, they probably have some amount of experience in Unity and/or Unreal and perhaps you could get them involved in tweaking some test projects if you're lacking artistic taste with lighting and image effect settings such as color correction and so forth. In the end your client's opinion is all that matters since it's not your project, and you don't want to start showing them stuff a month or two down the road and get a disappointed response. Get them involved as soon as possible, and if they play around with it themselves and clearly prefer one engine over the other, and you can't make a good case for why the other would be preferable for other reasons, you're just going to have to go with that.

 

Seems to me your looking for the best quality artistic presentation - but the catch here is VR. There are going to have to be thorough optimization done where the visuals are scaled back to allow for a smooth experience.
IMO - if you go all out in Unreal - ultra super duper graphics - you are going to need to scale back a lot further - which might cause issues since you are more experienced with Unity and C# than the other.
My bet is you could create a beautiful looking experience that runs well in VR much easier in Unity than Unreal. VR will be what holds the visual grandiose-ness back - not the engine.

This is also key - the ability to create high quality content - that runs on VR is a high bar. The skills required to create that high quality visual content isn't moderate. Though since it's fps - might not be as complicated as it would be with high quality characters. Enemy characters will need to match the visual quality of the environment.

Sounds like an interesting project!

 

Wow, thank you very much for all your responses.

You are right! At least initially, I should do some experimentation im both engines with whatever I will be getting from my client. Then see what looks better to us, and how my general feeling towards the technology is.

My client is NOT (yet) very experienced in 3D-stuff, let alone engines (@billy). Actually she (it's a she) is a "real-world" sculptor, now attempting to realize her ambitious artistic visions in the far less limited virtual space. So, in regard to technology she is solely depending on me. If I say it will be Unity (or Unreal), she'll trust me in that.
Though she is just starting out creating 3D-game-content, we did years-long projects successfully before, so I know she has the endurance, willpower and the funding to actually realize what she is looking for..

@ Andy-Touch : I'm afraid I cannot really explain what we will be looking for in terms of "Landscape Design". From what I saw in her few sketches so far, it will probably be less about "earthly" landscapes (with "realistic" terrain, foliage, etc), and much more about peculiar, "alien" geometry, weird colors and materials, otherwordly weather conditions, particles flying around, etc. Perhaps it's going in the directions of "The Solus Project" or "Mind : Path to Thalamus", which happen to both been created in Unreal Engine..

Okay, I understand that VR poses limitations in regard to how much visual potential can actually be used, and that Unity is currently ahead in terms of VR-support and -capabilities.

This is going to be a tough call! But you are right in that there will be no way around trying them both..

 

Learn and use Unreal 4, you'll got better and faster results out of the box for this kind of project.

 

After spending at least 1 year to get up to a competency level enough to develop the project you are looking to create.

 

I mean terrain and vegetation Tool, shader editor, available open world tools, if you already have 3D art you get results quickly and it will look awesome. Making the gameplay is another matter.