Unity / Unreal Engine / CRYENGINE showcases at GDC 2018

Here are the official videos for Unity, Unreal Engine and CryEngine that were created for GDC 2018 to show what games were made with each particular game engine.

PS: I wasn't able to find a Godot showcase from 2018, sorry.

 

Juan Linietsky's Godot 3 GDC material

 

Is Unity still the best engine today?

 

Only in the same sense that C# is the best programming language.

 

ya, so agree with that, and CG is also acceptable comparing with GLSL.

 

Just wondering, Instead of showing models+effect from artists, should showcase demo reflects how user friendly to cook a prototype and easiness to debug games?

 

I just came across another video, where they showcase (new?) Unreal Engine 4 features...

 

I do not think that there is any comparison to UE4 anymore.

The best looking Unity game simply does not look as good as the best looking UE4 game for some reason.

 

@Peter77 here is Godot report of this GDC 2018.
https://godotengine.org/article/godot-doing-well-gdc-2018


This is very useful to know

 

"The grass in the other engine is always greener."

I didn't want to make a new thread, so I looked for an engine comparison thread that I can necro and picked this one.

I have found a surprising number of "Oh god, why have you chosen UE4, it is such a crappy engine"-posts recently. And from a gamer perspective I can see where they are coming from, all the UE4 games that I care about either run like a** or have a very distinct "UE4 look" that I don't like. What really surprised me was that one post mentioned Unity in the same breath as CryEngine as being one of the "good engines". I don't know if this was one of you guys trolling, but with over 1000 hours in COD MW3 that person looked like a legit steam gamer to me.

Is the public opinion about the quality of Unity changing? Have you guys picked up any interesting anecdotal datapoints that support this hypothesis? Do you think this is caused more by the achievements of the Unity community as a whole, or by how poorly the developers of a few popular UE4 games optimize their games, dragging the public perception of the whole engine down?

 

I think public perception is pretty much the exact same for Unity, only UE4 is getting the same general treatment as more and more games get released by people who don't know how to get the most out of the engine or simply think they don't have to. I remember when all Unity games "looked the same" because the only real way around it was to write a bunch of postprocessing effects (Unity's were trash garbage at the time), custom shaders (Unity's were trash garbage at the time here too), or actually purchase some combination of the two. It's just that now UE4 is in that same position where everything looks same-y because, well, that's just how it is on this bitch of an earth.

 

I'm not a gamer but games in Unreal are really nice. Must be a player's perspective.

What is public opinion of Unity?

 

There is a very distinctive appearance to the default settings in UE4. One that immediately springs to mind is the early access game Citadel: Forged with Fire and I'm definitely with @Martin_H that it and other games like it have an odd appearance to them that stands out.

https://store.steampowered.com/app/487120/Citadel_Forged_with_Fire/

 

What about the other big engines like RedEngine4...

or Frostbite,

or RAGE,

how do they compare then?

Or are we reaching common ground where there is not much between the game engines now in terms of graphics DX11/12/Vulkan, Physics Physx and Sound as they use common API's and target the same hardware platforms.

Will Nvidia's RTX add another layer of commonality where all engines can have raytracing?

 

They are really good but also proprietary. We don't really know about the editor and working in them.

Frostbite is probably the most advanced with such amount of destructible environment and graphics fidelity. Also now with RTX.

 

Frostbite probably has the least amount of features out of all the engines. If you don't believe me, ask around

 

A few of my team were core engine developers at dice... they have less than stellar comments about its features and usability in development.

 

Ok, Unreal is the best engine then

 

That's surprising to me, and to be honest seems a bit like trolling. Not because I think Unity is particularly inferior to Unreal (although I think it still trails graphically despite closing the gap), but because Unreal does not come across to me as having any clearly visible deficit going on of the sort that all those Unity projects missing good post effects (or post effects entirely) communicated to gamers.

In my opinion, the Unreal sizzle reel shows a huge range of different aesthetic styles. I definitely agree that the default post settings do have a bit of a specific style, but that's to be expected. And what I would say is that the heavy-duty default post probably looks exceptionally bad on games that are otherwise not well presented, whereas Unity's presentation for these sorts of games is usually a bit more 'humble', it doesn't make a sub-par visuals look like they are trying to be glossed over.

For Unity, I think the best way to go is still a bit of a stylized look. Games that aim for too much blockbuster presentation (such as D.R.O.N.E or Seeking Dawn) fall a bit short of where I would like to see them, whereas games like In The Valley of the Gods or Vane look really great. It's debatable how much of this is due to Unity but I can't think of a real exception to this hypothesis.

 

In the same sense that C++ is the best programming language.

Unity's still the most popular for indies:

• Steam Indie genre page. Check the Trending, Top Sellers and What's Popular tabs. All of them prominently feature Unity games as the majority of games that use third-party engines, such as Rust, Slay the Spire, Hollow Knight, and the Forest.
• GameJolt. Most games are Unity. I can't remember the exact page, but they also released a pie chart recently showing the engine distribution, and Unity was on top.
• Most people use Unity in Global Game Jam. 4000+ for Unity and 200+ for Unreal was the number I can recall.
• Google Trends. Unity is more popular than Unreal, which is more popular than CryEngine, according to Google.

Unity's also the most popular for 2D, mobile and XR.

Where does that leave Unreal? Unreal does do indie, mobile, 2D and XR just like Unity, but it's less used due to Unity's existing community, mobile support, 2D support (Paper2D is dead) and such. The only market Unreal is better than Unity is in console AAA games, where Unity has near 0 games. However, if you're going to do AAA, from what I heard you should learn C++ instead, since most studios use custom engines, unless you're planning to work for a studio that uses Unreal specifically like Bandai Namco or Capcom.

And speaking of good looking Unity 2018 games, GTFO:

 

To me it's not always which engine is best, but which one you like to work with the best...That's why I use Unity...

 

My theory why this is happening is because Epic Games made UE4 accessible to a much broader audience than it was ever before. It now attracts not only professional game studios, but untalented people as well.

It seems you had to pay several hundreds of thousands of dollars to Epic Games to license Unreal Engine 3 (1). Investing so much money in game tech was probably reserved to professional studios only. A professional game development studio should be able to create high-quality games. That's why many Unreal Engine games were of high(er) quality in my opinion.

Today, with UE4 being very accessible, many people are able to create games with it, just like it was always possible with Unity. Thus, developers with all types of different experiences create UE4 games nowadays, which causes the perceived quality of UE4 games to worsen, because of every high-quality UE4 game come <number> games which are of lower quality.

I believe Unity is on its way to attract more professional studios lately, because of their latest technology enhancements. If Unity Technologies keep working on tech that is interesting for pros, I believe the public perception of Unity is going to improve, not because the technology got actually better, but because more professional studios are going to use it and deliver more higher quality games than we see today.

 

Nothing to do with the Engine, but I thought the Cryengine video was extremely uninteresting compared to the other two. It's like a montage of unfinished, generic shooters.

 

It had an indie tier that had a $99 fee that you paid when you published a game with 25% royalties.

 

UDK! Ah, good times. See, Unreal had been free for a long time and it didn't kill Unity. I actually tried UDK because it had real time shadows in the free version (back in 2011), but quit because you had to restart the editor for every script change, which was insane. Also, Unrealscript.

 

The deficits that the UE4 games that I tried have, is not visible in videos. It's a) abysmal performance that forces many players to play on medium or low settings, and b) their antialiasing often looks like a blurry mess. It looks fine in videos but whenever I see it ingame I think "eww, maybe I should just turn AA off altogether". And I'm someone who doesn't even complain about FXAA and usuall takes any kind of AA over no AA.

That's true, though CryEngine happens to have been used for some of my Favorite AAA shooters like Homefront Revolution or Prey. I actually like its look the most out of all current engines I think.

In my specific question I don't care about that, this horse has had its proper beating. I'm wondering more "what the everyman thinks". Just a few months ago I told two friends about Insurgency Sandstorm being made with UE4 and they both seemed to think UE4 is the most amazing game engine since sliced bread, and I said "I don't get it, I have yet to see a game not made by epic themselves, that looks good and runs well on that engine", and no one ever can give me any examples when I ask for them. And now that Insurgency Sandstorm is in open beta for preorder buyers, their forum is riddled with complaints about performance issues and the old Source Engine game running better.

I think it is interesting to have an eye on these events because I think it is not that far fetched that someone without any kind of deeper tech-knowledge, who buys 3 games in a row made with one engine and they all ran poorly, will develop a bias against that engine that factors into steam-review outcomes or even purchase decisions. When Unity was used to flood the games market with S***ty asset flip games, we've seen that happen against Unity, but now it seems CryEngine and UE4 have also had their fair share of terrible asset flip games, taking some of the heat off Unity.

 

Definitely, after PUBG and Insurgency: Sandstorm I've seen people started hating on UE4 like they did on Unity in the past. Also Fortnite on Android was quite a disaster too from what I heard.

 

I think the new splash screen has made a huge difference to perception.

No longer is the first impression of a bland grey screen. Plus it has changed from a Unity game to an individual game enabled by Unity. Judgement has changed from what the engine is capable of to what the developer is capable of, but Unity still gets marketing (marketing that now gives a better impression). A win for all.

 

Where Unity falls behind competitors is not always about tech or performance.

Currently Unity offers a lot of ways to do things but not a lot of easy ways to prepare things. We need to be able to prepare assets in-engine in an easier way, such as h-lod and Lod generation built in. Convex hull decomposition for mesh colliders built in. Things like that would help us a lot.

 

Also, improved data streaming/loading between levels. I saw a lot of players were dissatisfied about the time needed to load another level.

 

Nice to see cryengine up there too.

 

Yeah I totally agree about the AA, it's way too heavy. And I also think that Cryengine looks the best of all - everything looks razor sharp without looking like it's made of aluminium foil.

But was performance not an issue before? I think as @Murgilod pointed out before, maybe it's just become a lot more accessible to people who do not know how to optimize well, or how to tune the aesthetic properly. It's certainly much of a heavier load on cpu/gpu by default than Unity is. And I've heard people say (I think it was @ShadowK) that it runs a lot faster than Unity in something like an open world scenario. So it seems possible to make it run like a champ if you want to.

 

My main pain point with Unity is source access. It's not a good enough reason to switch engines but I can trace almost every major pain point to things that would have been relatively straight forward to fix with source access. Not every case is like that, mine just happened to be in the relatively easy to fix category. And Unity wouldn't due to priority. Which includes cases like they have a fix sitting there but the overall system is complex enough it could end up changing behavior in unexpected ways. Which I think is completely valid on their end not faulting them for anything there.

 

That's great, we're finally getting source access to everything that matters. Being shifted over to C#, SRP, ECS, package manager etc.

 

I'm just curious, what do you think about Unity's TXAA and SMAA, compared to the other two?

 

I haven't used it much myself, I usually just go for FXAA. Judging by Book of the Dead, it looks pretty good, although they used so much depth of field and chromatic aberration it's hard for me to tell.

 

Thought I will reuse this thread...

I just found "The State of Godot" GDC 2019 talk on youtube, perhaps more people than just me find it interesting as well...

They seem to support different programming languages and even came up with their own GDScript. Why would anyone do this?!?! Boo boo!!

What I think is pretty cool is that they're working to run the editor in a web browser. For serious development quite some non-sense again, but I can see why this is of interest for schools.

I also learned they receive support/donations from companies such as Microsoft.

Anyway, judging from this video, Godot is not (yet) a competitor of any of the commercial engines available today.

 

Godot is years away from being a competitor to any kitchen sink game engine and that's just assuming that kitchen sink game engines do not advance in any meaningful way from their current state.

 

Godot cannot and should not compete any more than blender "competes" with maya. It's a + in the world, but not an exclusive.

There is always a need but I am not interested in an engine chiefly developed by volunteers. I want staff in a day job, paid extremely well to work hard on something that's max top end. Unity is that for me.

As for running in a web browser, quite inevitable. Everyone already considered this including Unity. Its not an if, it's a when.

For something like godot, it's a very easy "sell".

 

They give their reasons in the history section on this page https://docs.godotengine.org/en/3.1/getting_started/scripting/gdscript/gdscript_basics.html

I think a scripting language that is built from the beginning to be solely for game development could be a very powerful concept. I was a big fan of Skookumscript for UE4 but the integration with that engine was a mess at the time. Sadly Godot did not take GDScript far enough imo and it's just a basic scripting language that doesn't shine like skookumscript could have.

 

Blender for sure has turned some formerly paying customers of Max/Maya/Cinema4D into Blender users. I know several people that I'd classify as "casual but commercial" users of 3D software (e.g. illustrators using it to make 3D blockouts etc.) that either switched or are considering to switch to Blender. I've even heard a small art outsourcing studio that works on AAA games consider to switch the entire studio over to Blender and only keep one or two 3DSMax licenses to run some proprietary exporters. And unlike me those aren't Blender "fans" they've been using Autodesk products their entire careers. I don't think you're giving blender enough credit for how much competition it really is. Rumor has it that Autodesk is S***ting their pants enough to spread lies about artworks created with blender still being bound to the same open source license that blender is. Source for that comment is in this video:

Spoiler

relevant part starts at 53:00 till about 56:00

 

Fun fact: Unity will soon have modelling tools and replace Blender

 

Not to mention that the skills you learn in most of them are largely transferable. Hell, here's a controversial statement to say the least: since 2009, Blender has had a dramatically easier to learn interface than the absolute clusterfuck that is Maya's UI. I haven't gotten quite used to the 2.8 UI, but it's a decent improvement over 2.49 in a lot of places that were newcomer unfriendly.

Godot feels when Blender just got particle interactions. Sure, you could maybe use it, but compared to everything else on the market, it'd be a huge uphill struggle compared to the slight incline it is now.

 

Steady, never dissed blender or anything. But it's not going to replace anything at all simply because ultimately when change happens as it always does, the money is at the front of change.

In Unity's case they have full time staff, very smart expensive staff, solving things like optimised raytracing, DOTS, etc. GODOT isn't even on the same page performance wise and never will be until hardware makes software optimisation obsolete.

So it's obviously not a competitor but a +, a choice, a bonus a tool for the job, horse for the track, car for the supermarket

 

Pics or it didn't happen.

 

Yet, in line with the statement about user interfaces by @Murgilod, Unity's staff are the ones responsible for that horrible new asset store and the decision to replace the forums with Lithium.

 

Nope. You're mixing the engineering work with the managerial decisions. These are two very much different things. With the listed technical stuff they don't have too much wiggle room to decide on managerial level, they can play with how many engineers work on each of things and when exactly, but they can't say no. There is too much hype around them.
The social media/website/community stuff was managerial decision, and sometimes wrong ones, sometimes not, but they certainly didn't come from the same engineers who are working on the ray-tracing features.

 

Not necessarily true. Managers should be technically an experienced staff. But unfortunately, I know in fact that this is not always the case. And that includes big industries outside game dev. Not because they knows nothing. Rather opposite. But carried expertise may not necessary fit, to desired outcome. Hence example of results, as we have seen in Asset Store.

 

For normal everyday companies I would be inclined to agree with you, but Unity's staff blur the lines between the two with a good portion of their management being people who have experience working on games and game engines.