Yes.
But to be fair, if Ryse was on Unreal4 it would also do bad on mid-level hardware. CryEngine is just the most optimized engine out there… which is a shame because it has been abandoned since 2011 in terms of major development.
Chevy vs Ford
.NET vs Java
Windows vs Mac
Oracle vs SQL Server
etc…
1 Like
No comparison .NET and C# are just straight up a better Java
Every successfuly Engine has its strengths. Many engines are developed and modified just to get the last bit of performance for specific platform, for innovative features, a specific quality (like graphics) or target a specific genre of games. Sometimes the developers even work towards the best performance for a single game (down on the game-engines source code level).
Also, the teams who work on those hyped high-quality titles often consist of a professional core-team, with experts for details that you wouldn’t even think about. And if they don’t, they’d simply hire some.
Simply cannot be compared with indie-teams who have a low-budget and not the manpower to cover all that.
Which - of course - does not mean they’re not competent and talented.
But you cannot deny that some games are just thrown together without passion, sometimes even without proper knowledge.
1 Like
Rubber is a good product, but not all shoes using rubber are comfortable.
3 Likes
punch them. right in the face. really hard.
10 Likes
Well, lets see… is there ANY Unity game you can compare pound for pound to a release like Ryse? While there certainly are UE4 or CryEngine games like that (though also not as much as some fanboys think, at that lofty height of AAA-ness proprietary engines pretty much rule), Unity engine games that really try to be on the cutting edge created by big studios are not really existing AFAIK.
Even if they were, you cannot compare big AAA game A created with UE4 to big AAA game B created with UE4. Chances are quite high neither of those uses a stock UE4 engine. One of them might have heavely modified the renderer or other parts that give them a big boost to performance. That does not mean Unreal Engine 4 stock is better than Unreal Engine 4 stock… the same could be said if one of thos engines that got pimped was Unity.
Then, the games you CAN look at for devs trying to create “cutting edge” 3D games are mostly created by small teams. These tend to not have the resources to patch engine problems (thus having to work with stock engines), and they generally don’t have the resources to optimize their games just as hard. Thus I expect this game to have worse performance than a big boys AAA game in general, no matter what engine is used.
On the other hand, Unreal and CryEngine are used by bigger and more professional teams usually. While some people claim these to be just as easy to learn than Unity (which for them personally might be true, people tend to learn at different speed and view different things as “difficult”), the general usage pattern seems to point in the direction that on average, Unity is easier and more forgiving for newcomers, and lone wolves.
So on average, I would EXPECT games created with UE4 and CryEngine to outperform Unity games… because on average, bigger teams and more expierienced devs seem to work on these games.
As a counterpoint to your post, I have seen some trash CryEngine games pop up on Steam lately, seems like the lone wolf shovelware slingers have noticed the CryEngine engine. Now, these games clearly were built without care. But what is also noticable is that they run like pigs. Performance often is quite poor.
Does not mean much, I mean we are talking about “developers” who cannot even be arsed to swap the standard Crysis GUI elements for their own… but it does mean that even in CryEngine, if you don’t optimize your game, performance is going to suck.
So if CryEngine was used just as much by the bottomfeeder devs, the average performance of CryEngine games would most probably dip close to the average performance of Unity games.
Do I say Unity is on par or better than the other off-the-shelf catch-all engines? No, I am not saying that. Am I saying that Unity should stop optimizing the engine? No.
But what I am saying is that anectodal observations like yours are not facts, they are just anectodes. I can construct a counterargument in minutes that is just as wrong.
In the end until somebody constructs a real game in Unity AND one of the other engines, is just as versed in optimizing Unity games as he is in optimizing games created in the other engines, and presents performance numbers based on that, we are not talking facts.
2 Likes
The main problem if you can call it that is that those people have no idea how a game is made and what it requires to work, so all you can do is try to explain but still they won’t get it. Can’t do much about it if the other isn’t willing to learn, which is perfectly fine, not everyone have to know how a game is made, they just want to play.
So yeah, sadly with many examples of sloppy work and steam allowing all sort of junk, is inevitable that these talks happen.
All I would do is try to explain how the game work without comparison, in easy words if you can. If they get it good, if they don’t you shouldn’t give much of a shit.
Unfortunately this will never happen, because it costs a lot of money to make a game like that, and experienced developers who are trying to make a game have no interest in spending loads of time and money comparing engines for the ‘benefit’ of the Unity forums. They just choose what they know will work for their purpose. The only thing you can do as a dev is learn enough about various engines to know yourself what to choose for your own game.
Frankly in terms of optimisation, I think that it’s often overlooked how much an art in itself that is, how much you have to manage what goes on. I mean, I doubt in that Uncharted 4 trailer, that they just dumped all of that junk in the scene and pressed play - it makes my mind wither to imagine what the hierarchy would look like if it was in Unity. AAA games are sporting insane amounts of textures and effects, to get that sort of stuff running on a console for example is going to take some serious wizardry.
And that’s probably what the supposedly wrongheaded thread is all about, someone who has seen the results of people loading up endless ‘pre-alpha’ jitterware onto steam with no clue and/or interest in optimising things.
1 Like
Similar to one of my favorite quotes: “Facebook and Twitter have given voice to those who were never meant to be heard”.
2 Likes
This, i have seen a lot of unity titles that have had no effort put into optimization, and there is a lot you can do. A large part of optimization is not even a per engine thing but a art and tech art thing. creating your foliage in a way to reduce overdraw, writing good optimized shaders and getting the balance between polygons and shader complexity right, reducing draw calls, taking advantage of static batching, building your levels in such a way that occlusion culling is useful. Most of these things are things that have to be optimized per game and not on the engine side, and even require things to be designed in such a way they can be optimized.
1 Like
Which is exactly WHY these studios pay all those people working on their game instead of a single dev, and why they are not going with off the shelf engines, and if they do, modify those to such an extent you could call it a different engine.
ALL the off the shelf engines would most probably crumble under the weight of such a project… some maybe more than others, but not even CryEngine, at stock, is really up to such a task. But none of this engines has to. They are catch all engines created for the average project, not specialized engines for any high-end AAA project.
Its individual devs misunderstanding what general purpose engines are, and what their job as a game dev is, throwing all their assets into a scene and hoping for the best, that create the insane idea that its an engines job to make the game run fast.
In the end, that is the devs job. If a game runs slow, its the dev of the game to blame first, not the engine.
1 Like
Why should it be that?
As said previously, AAA would get source access and could be able to rewrite half of the engine.
I really don’t get why small indie and hobby users care so much if AAA devs use Unity.
I wasn’t making an argument. I clearly stated that observations were “personal experience”, and I ended the post asking a question, not asserting fact.
That question remains:
- Is Unity on par with other engines in terms of off the shelf performance?
- If not, where are the big differences?
The AAA rewriting half the engine thing seems to negate meaningful discussion. So what about at mid-scale indie range - like $2-3 million.
Anyone have any real ideas on where these things stand in 2017?
I actually thought that “Would that game have been made without Unity?” was the most important point I made.
By the way my point had more to do with the OP post than anything else. I don’t really know how well Unity would handle Ryse. I just think it’s a good bet that many Unity games are unoptimised and that contributes to threads popping up bringing it into question as an engine.
The thing is that AAA-style, graphics heavy, openworld games are not usually done in Unity. So it seems quite possible that it might have some problems with that sort of thing. On the other hand, it seems quite possible that things that other engines are ‘capable’ of would preclude them somewhat from efficiently running smaller games on low-end hardware.
For one thing, (in my limited understanding of it) I think Unity’s graphics pipeline is somewhat clamped to low-end hardware. No doubt the reason they picked enlighten, despite its various problems, is because it can run on fairly recent mobile hardware, whereas voxel ray tracing would very probably not. A lot of graphics-related stuff in cryengine and unreal might necessarily take advantage of things that a lot of the devices that Unity often targets simply do not have.
I seem to remember ShadowK mentioning recently that Unity’s foliage performance improved greatly recently. I’m not sure what kind of testing it was. But considering the terrain system is ten years old, it probably still has a ways to go before competing with cryengine in that area.
Let’s not get carried away on the whole “a custom-built engine is always better for AAA” train. If the dev team has a lot of experience with that, sure. But there are some cases where building a custom engine did more to hinder a game than help it (exhibit A, Square Enix).
1 Like
The thing is, there are very few AAA games made with any of the major game engines, if you google for it. I tend to think that a customised game engine would very much help when building a lot of content for a specific kind of game. The kind of optimisations that are needed for Elite dangerous are going to be a lot different from what you would need for a game set inside a jungle. Generic game engines (which would even describe unreal to a great extent), are never going to be optimal for any specific need without a lot of work.
Even for me, customising Unity with editor extensions (and blender as well) is a huge part of what will make the kind of game I want to make possible, by making things much more efficient. I have no doubt that AAA companies really shape their engines for what they need. When you’re making such a lot of stuff, a bit of efficiency compounded is going to go a long way. Especially when you’re breaking up the work amongst a lot of people, to reduce information overhead you’re going to want to channel things along a very specific workflow.
It’s really a trade-off between flexibility and optimisation, the way I see it. And when you’re going to spend 100 million, it’s probably going to be worth spending a few million on making the engine spit out exactly what you need.
whats terrible is for a little while, I heard “Unity sucks”
so then I was trying to learn the Blender Game Engine LOOOL
or there was like panda something … and some other crap … those REALLY suck lol
buut … maaybe Unity had a different price structure back then?? wasnt free??
or something??? … I forget… been a looong time
The answers are:
BMW, Rust, Linux, and MariaDB. What do I win?
There’s a ton of AAA games using off the shelf engines. Sometimes they just pay extra so they don’t have to say what engine they use. There are literally hundreds of AAA games from the 2000’s that pretend they’re using a custom engine, when the reality is they were using the Unreal Engine with varying levels of customization. Part of me suspects Capcom’s RE Engine is a modern version of this as it has a lot of similar visual techniques (and artifacts!) as UE4, but UE4 devs have also been fairly open about their work so it might just be an artifact of them building on those talks.
The problem with Unity and AAA is just because it’s always been a bit behind in terms of “AAA” graphics features that are built in, and a lot of veteran developers refuse to use engines without access to the source code. There’s good reasons for this, as sometimes a bug that is preventing a game from shipping might take many man-days to diagnose and subsequently work around that access to the source code might reduce to one programmer spending an hour or two to fix. A few years ago it was essentially impossible to get access to the source code, were as today they have it as a built in option for the top tier, and their upcoming Scriptable Render Loops already have a lot of modern graphics features implemented so it’s something they’ve been working on improving.
That being said there’s a lot of us veteran devs who are using Unity now for smaller projects. My experience has been that Unity lends itself best to smaller teams where custom engines I’ve worked with in the past are designed heavily around allowing multi-user editing, or at least easier segmentation of levels. Again this is something Unity has been working on with stuff like multi scene editing, and of course nested prefabs 
…
Full disclosure, I’ve never used Rust or MariaDB, I dislike using Linux, but I do like BMW.
6 Likes