im currently debating whether i should invest my time in learning and building my game in unity or udk.
first of all, im looking to build a game that looks and plays like Diablo 3, with its top down 3rd person view.
im attracted to unity so far only because the liscensing fee is really affordable (something like $1,500 for upgrading to unity pro) as opposed to UDK's hella painful 25% of your profits after you make a certain amount.
it makes me wonder why go to UDK at all? unless there's hidden costs im not aware of in unity.
is there any reason an indie company would use UDK as opposed to Unity?
in Unity i heard that it takes more effort to make your stuff look good, compared to Udk, but how much harder is it? and is easing that difficulty really worth that big chunk of 25% of your profits?
and whats the best looking game out there that was made in unity? so far ive only seen Shadowgun.
Replies
It's a lot easier to make a good looking game in UDK.
Programming is easier in Unity, and it is much easier to port to multiple platforms.
Try them both out and see what is most comfortable for you. There's hundreds of tutorials for each engine.
Most first-time indie games will probably not easily break the $50,000 boundary anyway.
Unity is great about 95% of the time. However, I find that some of the light mapping and dynamic lighting features you want to be using on a high end title are not fully flushed out, poorly tested, or suffer from strange implementation. Also, unity doesn't have a node based material editor (and strumpy is out of date). Aside from those issues, Unity is way easier to get started with and the scripting lets a tech artist do a lot without needing a real programmer.
Acegikmo is currently fixing that, ShaderForge will be available soon and looks amazing.
Which at that point is hardly significant.
Can you post a link? Googling Acegkimbo doesn't come up with anything.
And I found the node based editor that was referenced earlier:
http://forum.unity3d.com/threads/191595-Shader-Forge-A-visual-node-based-shader-editor
[ame="http://www.youtube.com/watch?v=XEbNnnz_tO0"]Raindrop Teaser 2 - YouTube[/ame]
If you or someone on your team knows how to code in any object oriented language, they shouldn't have trouble with UDK's UnrealScript. Make sure you're using UnCodeX to learn about inherited methods and variables so that you're not doing things twice. Unity really is a breeze though, and if you're just starting out coding, I would definitely choose Unity.
This is all of course without regard to licensing. If you honestly believe your game can make 50k, it's your choice. However, if you're unsure you're gonna make 50k, and you or someone on your team knows how to program, I personally would choose UDK.
But really, they're both free, try them both! Choose whichever one you'd be more willing to work with for hours upon hours.
This is the common issue with software also.
Blender -> free so ALOT of people use it, which means theres going to be more "bad" stuff to see rather than good. Many assume it isn't as powerful as other 3d packages because of this.
Unity -> free aswell as UDK but Unity is much easier to learn for beginners. The result is seeing alot of games below the quality mark you're expecting, which makes you question its capability.
In most cases the difference between UDK and Unity won't be hugely apparent unless it involves new next-gen features. UDK is obviously ahead in this department
Depends on the situation. For one developer just switching to the pro version for Mac/PC it's only $1500. But for a 4 man team working on mobile games (ios/android) you're looking at 4x (1500+1500+1500+500) = $20.000, or 20% of revenue. Not a bad deal, no, but significant, yes.
The biggest benefit of Unity versus UDK is platform support.
UDK: PC, Mac, IOS.
Unity: PC, Mac, IOS, Android, Windows Phone, Windows RT, Blackberry, Linux, and the chrome web app thingy.
Unity is easier to use and learn in my opinion, but UDK will come out looking nice out of the box.
If you have absolutely no game development experience, I would honestly suggest picking up Unity but if you have an idea of what you're actually doing that I would say go for UDK.
http://unity3d.com/gallery/made-with-unity/game-list
Lots of great looking titles and a huge variety of styles. Unity is as capable as any other engine, it's all up to the artist to create something visually appealing.
Unity also has a lot of inexpensive plug-ins that can really speed up your development. I highly recommend Playmaker.
Depends on my project.
*Scaleform is in Unity too as a plugin ($280-ish).
UDK mainly shines in its realistic visual fidelity, as matching the quality of modern games built in Unreal in Unity would be a very large task. To be clear, that does NOT mean achieving a beautiful aesthetic in Unity isn't possible; there are many famous indie games that have super unique and cool art styles all built in Unity. I just wouldn't try to build something like BF3 in it (think Kentucky Route Zero or Blendo Games kind of style).
Also, outside of building the game to your vision and completing it, you should consider how important supporting multiple platforms is. Unity wins that hands down.
[ame="http://www.youtube.com/watch?v=Ln9F8Uf7HiI"][DLG] [ Unity 3D ] No Heroes - Massive Destruction Teaser 2 - YouTube[/ame]
[ame="http://www.youtube.com/watch?v=cfrL81qtbrk"][DLG] [ Unity 3D ] No Heroes - Destruction Part 2 - YouTube[/ame]
The destruction is exaggerated for demonstration purposes.
Then there's also the Line of Fire project:
[ame="http://www.youtube.com/watch?v=W69s2XsiSxY"]Alpha Project Line of Fire Update 4 - YouTube[/ame]
Then I tried unity and never looked back.
Took me 6 days from not knowing a thing about code to writing my own character controller just like I wanted.
Tons of tutorials, great community and asset store.
If you can't program I suggest playmaker. Visual scripting tool and a massive timesaver.
https://store.unity3d.com/products/subscription
Depends what your long-term plans are, I suppose, but you can do an awful lot with the free version, asset store, mechanim and the like to start building your game and then see which upgrade path suits you best.
I also dont plan to do the bulk of the programming.
the spending money on shaders(for unity) part seems like a pain. does that get quite costly?
i did see a demo of the houdini plugin for unity at Siggraph, and it seems interesting. i saw the demoer model a game level without having to bake and he did some procedual modeling stuff and put it in the level so fast.
but that costs more than unity itself, and it doubles if you want dynamics.
Even now the asset store has awesome things like Marmoset- http://www.marmoset.co/skyshop
...and there's things currently in development which really show the kinds of things that can be added to expand the capabilities when programmers can dig in to the engine - [ame="http://www.youtube.com/watch?v=Z1o_CCHhBTQ"]Voxel Cone Tracing Unity Plugin Preview - YouTube[/ame]
[ame="http://www.youtube.com/watch?v=1pfqCQQIbUM"]Unity - Screen Space Sub-Surface Scattering for real time skin rendering - YouTube[/ame]
Damn, I misspelled it. It's Acegikmo. And he's a polycounter.
http://www.polycount.com/forum/showthread.php?t=123439&highlight=shaderforge
wow that is certainly the best visuals ive seen in unity so far. i certainly want all those effects in my project.
though at this point im just too scared of the potential cost of all the plug ins, upgrades, and shaders needed to achieve that.
for now ill just grudgingly focus on udk, but ill be keeping an eye on all the future updates of unity. looks like unity has a lot of potential, just doesnt seem to have all the new stuff fully done or stitched together yet.
It happened...
2005...the year I was trying to figure out thief:deadly shadows editor, the strangest version of unreal engine 2.5...those were the days...
To be fair to UDK, it works well on current gen consoles and next-gen consoles as well as browsers. Unity works well on Wii U and is getting there for other consoles. Small-slice mobiles aren't anything to sniff at, but they're not as important as smooth PS/XB porting for a lot of devs.
If your game relatively small you can deal with it , but it is a real pain if you got some big stuff. Mechanim is a disaster then it comes to complex animation systems ,lightmaper and navmesh are in the 3.5 beta state, multuplatform is the issue ,in the real cases,memory management is a black box ,it would be better to check Unreal source licence ( Epic are kind about options).
Any way tools should be choosen by the purpose.
Allot of guys are comparing both engines to each other but in this scenario what's really important is what works for you and the fact is both would do a really good job on a diablo style game.
If that's the type of game your doing you wouldn't do massive photorealistic textures on your character etc because your far away or if that's exactly what you want to do then do it but with either engine your going to have to make optimizations and work around's no engine will give you that for free.
Secondly I don't think you should just assume that because you choose either engine your games going to look amazing. UDK has wicked materials no doubt you can find wicked materials for unity too no doubt but where the bulk should lie is on you as an artist. Once again YOUR doing the art so it with you that the flare and beautiful awesomeness should be coming from then the FX you add ingame adds to that not the other way around and both these game engines will give you that.
Lastly I think your thinking pretty much only about the art bare in mind on a diablo style game there are tonnes of functional implementations that your going to need to do so put in your homework about how easy that part would be to do and how much support there is forums, tutorials and specific tutorials for what your aiming to accomplish.
By the sounds you know how to do 3D art so your fine with either engine in that respect so great but how is your scripting, and understanding of game development as a whole. You may choose something that allows your game to look ace but when you come to getting stuff to work you have to smash through titanium to do things especially simple things.
Think more about the functionality and less about the art because your game will look nice whatever, why because your the artist you'll make it look nice whichever engine you choose and that's your job as an artist however as your looking to make a whole game by yourself atm you'll need to be more than an artist which means you'll need to take more into consideration.
Either way look forward to seeing your progress dude and don't forget to update cuz there are a tonne of people perfectly capable of help you out dude.
So keep at it.
Yeah and that new agreement will drastically change the amount of money involved.
It's more like I don't want to choose an engine and then later regret not choosing the other engine. Sort of similar to how I read in some other forum that some guys that learned cryengine regretting it because it's not friendly to indie devs.
And thanks for the input , definately some stuff I should consider! For the scripting side, I do want to have someone else or 2 to do most of the scripting, so this isn't totally a 1 man project. I just want to bang out alot of the visuals first. But man, this is way in the future, I'm lightyears behind this stage at the moment.
I hate to say it but logically speaking, most of the time artists get the least say in engine choice because all they're doing is making art. At the end of the day, you can't have a game without programmers and so for a smaller team that has very limited funds, you'd probably want to have to programmers pick the engine.
I know with my own project that I'm currently working on, we started off on Unity and were going to switch to UDK for the better visuals but didn't because programming would become a lot more difficult. Looking back on that 1-2 year old decision, I'd say sticking with Unity was one of the best choices we could have made.
I can't speak for everyone but from Antichamber talks that's not our experience.
UDK and Unity are both highly capable, but they each are a different philosophy of "how to make a game". Which one resonates with you, and what your end goal is, will determine what the actual best choice is.
UDK has more out of the box functionality than Unity. Unity can match it with addons, but the addons aren't always quite as homogeneous as they're created by people not on the core Unity Dev team.
UDK's AnimTree system seems to be a bit more capable and flexible than Mechanim, but less accessible.
UDK's built in shader system far outstrips Unity's for quickly setting up complex shaders and raw power.
Unity's shaderLab implementation allows for more exotic effects more quickly than UDK, simply because you have to write them all out. The node based shader addon helps close this gap, but again, add ons.
UDK's input handling is more mature and flexible. Its easy to allow players to reconfig their input prefs while playing the game.
Unity (stock) only allows this to be done on game startup and presents a rather ugly menu. Again, add-ons can fix this.
Language wise UnrealScript might as well be C#. The syntax is nearly identical.
The difference comes in how they interface with the engine. UnrealScript does need pre-processing and packaging before its directly used. So this slows things down a bit when iterating in code. (last I checked you need to re-bake and restart udk when making significant script changes) Unity can just compile and adapt on the fly.
UDK's entire setup is a modern version of a classic software development paradigm. Code drives the whole thing, the audio/visual aspects of the game are called from and specified by the code. (even if the code is created by Kismet nodes) The most basic game Object is an abstract concept from which concrete versions flow.
Unity's approach is far more modular. A basic game object is a concrete container that you add functionality to. Code still drives things, but you don't have to dig into the deep guts to understand enough to make a wide variety of games.
One other concern is the end project size. UDK games seem to always be just plain huge. GB worth of data. Unity games seem to more readily publish smaller end packages.
Edit: Unity has a much more flexible camera. UDK's camera system doesn't do orthographic.
And just to put the cat amongst the pigeons, you could go for the Cry Engine, the engine is 100% free "like UDK until you get to the £50K mark, then its -25% royalties but then that still very good for what you get". Plus you don't have to do any light baking. happy days for your artists.
Unity on the other hand you have to pay out all the time for little features and pay for each platform you develop on at around £1500 for each. Unity is also not free for education so if you need fresh talent they will more then likely know UDK and the CryEngine.
Ryse: Son of Rome
Also CryEngine (4th Generation)
On August 21, 2013 Crytek announced that their next CryEngine would not carry any version number. The reason for this decision was the fact that this new engine bears almost no similarity to previous CryEngine versions. The new CryEngine supports next generation platforms such as PlayStation 4, Xbox One and Wii U. So for others not to get confused, it will be called as CryEngine (4th Generation) overall.
Source - http://en.wikipedia.org/wiki/CryEngine
Title made with the CryEngine3
ArcheAge[31]
ASTA: The War of Tears and Winds
Cabal 2[32]
Corrupt Oasis[33]
Crysis 2
Crysis 3
Crysis[34]
Enemy Front
Everybody's Gone to the Rapture
Evolve[35]
Festung Europa[37]
Fibble Flick 'n' Roll[38]
God Slayer[39]
Icarus Online[40]
Lichdom[41]
MechWarrior Online[42]
Monster Hunter Online
Nexuiz[43]
Ostrov[44]
Panzar: Forged by Chaos[45]
Project Heart and Soul[46]
Resistance and Liberation
(CryEngine 3 reboot)
Shadow of the Eternals[47]
Sniper: Ghost Warrior 2[48]
Snow[49]
State of Decay[50]
Stone Rage[51]
Titanic: Honor and Glory
Tour Golf Online
Traction Wars
Untitled RPG[52]
Warface 360 Edition[54]
Warface[53]
http://www.polycount.com/forum/showthread.php?t=125500
Good luck.
http://youtu.be/5RlSmkSbleI
Not sure where you got the idea with the 50k threshold and the 25% royalties for Cryengine ?
As far as I know, there is no public info about Cryengine's license available. If you want to enter a license agreement with Crytek, you have to sign a NDA.
Besides that, it seems a lot people who have been working on their Cryengine games for years struggle to get a license for their games now, or even get in touch with Crytek.
Seems you're more a cheap beta tester for Crytek rather than a games developer.
As far as Unity is concerned, yes you have to pay upfront for this engine and for sophisticated features. I'm not a huge fan of that business model, because you don't get a unified workflow with lots of third-party plug-ins from different developers, with different level of support.
It's also known that Unity is a pain for bigger projects, because of some limitations and issues such as their garbage collection.
But compared to UDK it might be cheaper, considering you pay once. If your games makes significant money, UDK's royalty share might be the worst deal.
There are almost no indie games using CryEngine; this should probably hint at something.
Lol, first hit when you google CryEngine Licensing:
http://mycryengine.com/index.php?conid=70
Yeah ok. But the whole deal is still obscure, isn't it?
Plus you shooting in the dark with your project. If Crytek eventually dislikes it, you're screwed. No?
But I admit, I actually can understand Crytek here.
They don't want their glorious image of Cryengine dragged down by ugly hobby projects. At the end it's the games that matters. One reason why Unity has such a bad reputation in the graphics department. Because there are thousands of bad lit games with Unity's ugly default shaders. That lead people to the conclusion Unity is only capable of graphics straight from the 2000's
Oh and I'm not a big fan of LUA, but that's perhaps something personal, others might love it.
anyways, I love both Unity and UDK and I've been using them both for a several years. One thing that really lacked in Unity was a proper lighting solution. I personally really like the UDK way of lighting scenes.
The recent builds of Unity does have beast lightmapping but the defaults seem to be not so optimal, with some tweaking however, it is possible to get similar results. Here is one of my old scenes in both Engines. It's not a fair comparison AT ALL, just a personal test. So don't take this ss some childisch Unity vs Unreal video please
UDK: https://www.youtube.com/watch?feature=player_detailpage&v=CYrSiq1rRnM#t=35
Unity:
https://www.youtube.com/watch?v=PhxRl81rgsc
..and bam, proving everyone's point: nothing up front about licensing terms. This is acceptable for established triple-As and people already doing business. It isn't for indies who need to know what they're getting into before they start evaluating the platform, who don't want to have to already be at a business negotiating table before they've even started working on their game.