Split decision, UE4 or Unity
polycounter lvl 12
I've been doing some initial research into which engine to start R&D'ing here at work and I'm at almost a 60 Unity / 40 UE4 split. So I was wondering about getting some opinions. I work in an architecture firm that specializes in sports venues, so our projects are usually are on the large scale.
What has me leaning more towards Unity is that I'd like our end product to be reachable to the widest general audience possible. Things such as mobile and web (and the ease to get there) are more important than overall quality. Don't get me wrong, I want to do this in highest as possible quality. However I also don't want it to be so high quality that it becomes impractical for anyone who doesn't have twin GTX-970's to run it. The arch viz that has been produced so far for UE4 looks drop dead amazing, but I'm not sold on the practicality of that workflow for general public use.
I haven't seen too much yet on UE4's HTML5 or mobile capability and limitations. I've heard that it can struggle on less than high end mobile devices, is that still true?
Thanks.
What has me leaning more towards Unity is that I'd like our end product to be reachable to the widest general audience possible. Things such as mobile and web (and the ease to get there) are more important than overall quality. Don't get me wrong, I want to do this in highest as possible quality. However I also don't want it to be so high quality that it becomes impractical for anyone who doesn't have twin GTX-970's to run it. The arch viz that has been produced so far for UE4 looks drop dead amazing, but I'm not sold on the practicality of that workflow for general public use.
I haven't seen too much yet on UE4's HTML5 or mobile capability and limitations. I've heard that it can struggle on less than high end mobile devices, is that still true?
Thanks.
Replies
Tomorow: 60UDK / 40Unity
if graphical fidelity is your focus then go with UE4, easy as that.
Unity has a built-in webplayer plugin which can be nice, Unreal Engine 4 has been working on html 5 support (so it won't even require a plugin) although I haven't tried it yet: https://docs.unrealengine.com/latest/INT/Platforms/HTML5/GettingStarted/index.html
Honestly Unreal Engine 4 seems to scale pretty well and has been receiving constant major updates compared to Unity's months and months for an update. Unity's wide variety of plugins on their store helps offset this, whereas Unreal Engine's marketplace is currently trying to catchup. (most being art assets and animation, not too many plugins that extend out various aspects yet)
Scaling wise, having prebaked shadows will make a drastic difference in performance versus dynamic shadows (that goes for both engines)
There's also matinee support in Unreal to consider which can be excellent for quickly setting up cinematic walkthroughs of your architecture scenes, and it's pretty easy to set up basic movement controls with blueprints.
It'll probably come down to what your programmers are mostly comfortable with though, unless your team isn't implementing major features/just movements and fly throughs of architecture.
I'm always worried when it comes to presenting real-time 3D content to a client like that since god knows what toaster they're working on, aka videos being the safe bet, at least as a backup.
Ipad is probably the best for presenting something like that since it's hardware standard if you know what model the client might have.
UE4 has an edge for quick projects that need to look awesome, but aren't very sophisticated code wise, OR really awesome looking projects with really sophisticated code. This spread is because you're either dealing with Blueprints (beginner friendly) or C++ (relatively complex). Blueprints are surprisingly robust and one Epic developer I spoke with said the ratio of C++ to Blueprints for some of their projects is as much as 60:40. The down side is that anyone who uses Blueprints needs to learn them first because they're proprietary to UE4.
Unity has an edge if your project doesn't really fit into one of UE4's templates and you have Java or C# devs on staff who would be more comfortable simply adopting Unity's API rather than learning C++ or Blueprints.
On a scale of complexity from 1 to 4, my order would be UE4, Unity, Unity, UE4.
That's all purely my opinion though.
You said your projects are on the "large scale" but are you referring to code complexity or are you speaking in terms of something like art? C++ scales way better than C# for larger projects and so if you are working with a lot of code then you'd get better performance with C++, however if you aren't going to write a lot of code then C# in Unity will perform at a suitable level.
The good thing with picking Unity would be that your project would start out as more of a blank slate so you can improve the overall quality as you go along rather than having to tone things down from the very beginning to reach your ideal performance.
So the last time you used it was Unity 3 or 4? 5 just came out. I've got UE 4.7.1 and Unity 5 here, I don't notice any huge performance difference; personally I've had better luck with C# vs blueprints.
Please... Just... You're either a brilliant troll, or just Pancakes.
Regardless, all the advice on this thread is solid, save Pancake's.
EDIT: Not editing this post, but I get that I was rude, and am more articulate here.
I'm more impressed that you built the same game in both unity and blender game engine!
If you're a great artist then use whichever one you want. If you're not, go with the tool you know better or the tool you want to learn the most.
If it doesn't do what you want of it, there would be no loss. Then you can then switch over to Unity.
They both do large scale well, and they both have impressive web options. Though UE4's HTML5 is a bit younger I was impressed by its performance.
At the end of the day it will come down to your preference in workflow I think. They both do their job well, it is a matter of which does it in a way you are comfortable with.
You know what? I've never said one god dam thing to you. And here you are accusing me of something. All I gave was my honest opinion. And if you've got a problem with that. You might want to think of a more intelligent way of expressing it other than name calling.
I can go into more detail if you want, I have shipped multiple Unity projects (although with unity 4) commercially.
Or anything with a fair amount of popularity, for that matter.
If anything, Unity has far more "Fanboys".
I'm still using Unity. This is mainly because I've already learned so much about Unity, and I prefer C# over C++. Also, the technical art requirements I'm working with are much lower. But at the end of the day, it is a tool I'm working with, and I will reach for whatever tool I think is best for the job. I've already installed UE4, and fully intend to dig into it when time allows.
Now that both engines are free to experiment with, it has become either-or scenario. I no longer have to concern myself with financial investment, just time investment.
If we're looking for visual quality, UE. Generally these projects are intended for Sales Centres. We spec the machines and we build everything based on knowing exactly what hardware we'll have available. I'd be very fearful if we had to web-deploy any UE project as they generally have quite large footprints.
A typical project would be an exterior and/or interior of a 6,000 to 8,000 seat arena but can go as large as our MTS Centre project. We wouldn't need a full walk around, it would be a limited space view of the concourse, suites, loge boxes, locker rooms, etc. We also get a few requests from owners or big money donor's with views from their seats. We also show various arena configurations such as basketball, hockey(if there is a team there), concerts, and trade shows.
We also do large sports complex's but those in the UE/Unity app would more functional views of various cameras and highlighting spaces. Such as, here are the coach's offices, here are the bathrooms, etc etc.
Obviously, sports arenas are done all of the time in games. So I don't want to reinvent the wheel here with my process. But having a tool like blueprint to place the seats would be a huge time saver. We use the Rail Clone plug-in in Max extensively and we can create an entire arena bowl in a few hours with just a simple set of splines.
It seems like Unity should be the way to go, but UE4 might be the better tool in the long term development. Right now, I have no deadline on this. This is just R&D to see what we can accomplish. Maybe the best bet is to just get both and try them out on very simple scenes. It's been a while since I used both. I last used Unity 3 and UDK.
We had looked at using CL3VER(http://www.cl3ver.com) since it's close to what we want to accomplish but their pricing structure, and to some extent the quality, just doesn't work with our goals.
Thanks for all of the solid advice. It really gives me a lot to think about.
Visuals out of the box, UE4 does look better but if you put time and effort into Unity it can look just as good.
That's a perfect job for Houdini and Houdini Engine.
Really with any package or engine, max and maya can both easily duplicate objects around splines, and Unity could do it via a editor script, while unreal can do it via a construction blueprint.
Houdini Engine and Unity is pretty cool, but I am expecting much more from Houdini Engine and UE4.
I'm sorry if I came off as overly rude in that post. In hindsight I shouldn't have made it. But I don't believe that your no-holds barred, UE4 is the absolute be-all end-all of every engine ever made attitude is conducive to being used as a source of advice, or discussion, and you've used it repeatedly throughout the course of game-engine related discussions, even when the only thing to compare is the theoretical possibility of a feature that may or may not exist.
When you draw your evidence against an engine from experience that's several years obsolete, right after a major new release has come out (and for free, even), it begs the question whether or not you're being serious.
I've been using UE4 for my current hobby-scene since the thing went free, and I'm really impressed with it, relative to where it was when it first came out and I love its toolset. But I disagree with some of its design decisions and elements of its workflow in the same way I do with Unity's, and Horde3d, and JmonkeyEngine.
When comparing large software packages like game engines and frameworks, that touch so many different areas of the game development process, no one solution can really be the be-all-end-all of everything or even most things.
Yeah, I've been really taken off-guard by how Epic's jumped into that. Definitely worth considering if you're on the fence after doing an initial comparison.
I have opposite opinion. I was able to prototype base mechanics for my game in Unreal in less than hour.
I actually spend more time testing and iterating over mechanics, than on actual coding (and blueprinting).
Yep and thats why we got choices, since i find i can work faster in c# than blueprints, for others its the other way around. Also depending on the gametype a upside to unreal is there are templates already implemeanted in your choice of c++ or blueprints.
Everyone is entitled to their opinion. No need to call names. I haven't really gotten in to any ready made and branded game engines since I am working on mine and waiting in the slightest glimpse of hope for ubi-arts game engine, I would prefer to use unity because I am already studying c# and you can write scripts in c# compatible with unity and according to the indie game industry, many of them use either unity or xna game studio (even if xna may be discontinued five years from now.) To me, it would be much easier to get used to compared to udk.