UE4 or UDK?

[rtweed]

Question in brief: Should I switch to UE4 now, or should I wait until I have a working prototype in UDK and then switch? I definitely plan to switch at some point, but it's unlikely I'll have anything ready to release until late next year and could potentially switch engines virtually any time before then.

Where I am: I have been learning UDK on-and-off alongside a bunch of 3D modelling and animation tools. Most of my past game dev experience is with 2D games, although I did some bits & pieces in 3D years and years ago. I'm a self-funding lone developer so trying to fit all this around other work to keep the bills paid. With DK2 hopefully arriving soon, I plan to make a fairly big push to make some more significant progress before I get forced on to other things again.

Pros for UDK: Stable. Complete tooling. Lots of tutorials and documentation. Completely free (during prototyping phase anyway).

Pros for UE4: I know C++ and don't especially want to learn UnrealScript ⁽¹⁾. Better engine, full source access. Definitely what I will be using for anything beyond a simple prototype.

Cons for UDK: Not everything can be transferred directly to UE4. Possibly a lot of un-learning / time wasted learning things that don't exist in UE4.

Cons for UE4: Fewer docs, tutorials, etc. Some of the tools are still incomplete/nonexistent (AFAIK, but I haven't kept up to date with this). Greater chance of bugs. Monthly cost (would rather defer as late as possible). Possibly slower workflow ⁽²⁾

Neutral: Can do a lot during prototyping in terms of testing out assets, animations, etc., which won't be particularly dependent on game logic anyway. UDK might be better due to the cost, unless transferring assets is a huge pain. Working out things like final lighting models can really only be done in the production engine, but is also likely to come fairly late in the production process, long after the main assets & prototyping are done.

Unsure: possibly more direct control of Oculus SDK in UE4? Any significant practical advantages to that?

⁽¹⁾ I might bring in Lua or AngelScript to help speed up prototyping. Is this what others normally do?

⁽²⁾ I'd very much like to hear how people's workflow has changed since moving to UE4. Is there a lot more time spent compiling? Does that slow things down? Are other things faster? What's the asset import workflow like compared to UDK? Obviously most people will have a slightly different experience based on their skills, development priorities, etc., so would be good to get a few different perspectives on this for comparison.

[cybereality]

I would personally go with Unreal 4 at this point as you will likely get more updates and support.

[Sharpfish]

No need for a list.

Just go UE4 and be done.

[knack]

and i read few time ago have less royalti if you want make money

[spadak]

Also you can pay once for version 4.3 and stop your subscription and continue developing.
When you are ready, you can pay again and upgrade to latest version..

[rtweed]

"knack" wrote:
and i read few time ago have less royalti if you want make money

That's true and one of the factors in why I will ultimately be going with UE4. However, it doesn't affect the initial prototyping and development phase - you only pay royalties after you release something and start making money. On the other hand, UDK is initially free, whereas UE4 has a monthly fee. Not a big fee, but a fee nonetheless.

[knack]

"rtweed" wrote:

"knack" wrote:
and i read few time ago have less royalti if you want make money

That's true and one of the factors in why I will ultimately be going with UE4. However, it doesn't affect the initial prototyping and development phase - you only pay royalties after you release something and start making money. On the other hand, UDK is initially free, whereas UE4 has a monthly fee. Not a big fee, but a fee nonetheless.

19$, monthly if you want, i buy it 4 or 5 months ago, and cancel my subscrition while learning, i can update to 4.2. And when i receive my Dk2, i going to subscribe again for get 4.3.0 for the Dk2 support, i will fine perhaps few months more.

the are a minimun if you earn less than 50.000$ you not need pay royaltis.

[rtweed]

"spadak" wrote:
Also you can pay once for version 4.3 and stop your subscription and continue developing.
When you are ready, you can pay again and upgrade to latest version..

I have read this too. Not sure whether it's officially sanctioned by Epic or not, although I expect if they were worried about it they'd tighten up their terms a bit.

However with something like the Oculus SDK, there would definitely be a need for regular updates, which probably makes that approach less feasible. As far as I know, UDK is still being supported so updates like that will continue to be rolled out. Perhaps less frequently, but usually when software is updated more often it's also more accepted to ship with bugs and fix them later, so that likely cuts both ways.

Having done some digging around it looks like UDK doesn't currently support DK2 anyway, which might force me to switch if I want to get to work right away. It look like I'm in the first batch for DK2 although I haven't had a shipping confirmation yet, it says "processing" and payment has been taken, which I believe means it is ready to ship, it just hasn't left the distribution hub yet as some others have.

[falco]

"spadak" wrote:
Also you can pay once for version 4.3 and stop your subscription and continue developing.
When you are ready, you can pay again and upgrade to latest version..

I am seriously thinking on using UE4. Unity´s license is just too expensive.

[rtweed]

"falco" wrote:
I am seriously thinking on using UE4. Unity´s license is just too expensive.

If you're planning a commercial release and think you're likely to make (gross sales) more than about $30,000 (US) then Unity may work out cheaper in the long run, because they don't charge royalties. You assume all the risk by paying up-front though. It therefore only makes sense if you have a good expectation that you can generate sales significantly higher than that amount. Of course, that's for a single developer on desktop only: the break even point goes up significantly for teams and multi-platform projects. Of course it could be less if you go with the subscription model and can turn around a finished product insanely fast.

For me a 5% royalty isn't that much (to put it in perspective, Apple take 30% from the App Store) and it avoids having to pay a large up-front cost, apart from the monthly fee which is quite low anyway. If i had investors, a marketing budget and things like that then the up-front cost might not be significant in the overall production budget and well worth the potential increase in net revenue.

However, Unity is also a bit of a toy engine compared to UE4. The real killer feature of UE4 is full access to the source, so you can really do whatever you want, if it's within the bounds of possibility and you have the programming skill. As a programmer, this appeals to me; others may be scared of the complexity. No matter how good UDK, Unity or similar may be, there is always going to be a barrier that you can never escape, because you don't have the source code. Whether or not you actually hit that barrier, and how big of a problem it will be, depends on the game.