Our migration from Unity to Unreal - the Good the Bad and the Ugly

I held out against getting ProBuilderfor the longest time. When I finally got up the nerve to purchase it, I was kicking myself for not getting it earlier. It's amazing how quickly it becomes an integral and critical part of the level creating workflow. As a prototyping tool it's perfect. As a full on modeling tool, it's ALMOST there. I have it now to where I pretty much just work on characters in 3dsMax and maybe the odd prop here and there.

The real kicker is being able to see the final result right in front of you as you're building. Eliminates any guess-work and gives you a 100% accurate image of what your level is going to look like.

I love ProBuilder.

 

Yeah! robust tools like this should be an integral part (my earlier comments were directed at say, bringing in an entire modelling package like Maya, which is a bit too far IMHO).

 

It's amazing that all the other engines have similar tools built-in and not required as a separate purchase. Also Probuilder basic sucks, it's good only for evaluation purposes (and I already know that "real deal" is good. Doesn't make me magically find $95 in my wallet (which is huge amount of money here in Poland, indie developer or not).

Things like that should just be part of engine. The only reason I'm not moving to another engine is because I'm too far into my current project. If Unity doesn't get its marbles together regarding this, I'm afraid I'll have to move to Lumberyard or Source 2 (S2 being primary choice, but you know, Valve Time...) by the next project.

 

You're aware probuilder free is pretty much all you need yeah?

 

Then I think we had a misunderstanding. I was talking about tool like ProBuilder the whole time. IMO Unity should buy the company behind it and integrate it into the editor. They wouldn't even have to write any new code, only maintain it so it is compatible with newer versions.

 

The fact I'm building flat colored game now doesn't mean I'll always make such. Also making good level without subdividing and merging is a PITA. Royal PITA.

 

As said, instead of "re-inventing the wheel". I checked out some of the latest features in Stingray / Maya LT / Max..

So it now has a live bridge, which you can sync camera's and watch the scene being created from the DCC in engine. Now that IMO not only sorts out any complaints (usually from people who don't actually do art, so not a clue why it affects them anyway) and it stops engine developers / integrators having to assign a lot of resources..

Pefect combo, complete DCC power with interactive engine engagement. That's a powerful combination...

 

From their comparison (http://www.protoolsforunity3d.com/docs/probuilder/#featureIndex - you may need to ctrl+f and type "feature" to get to the relevant part):

• Material Palette (ProBuilder Only)
• UV Editor (ProBuilder Only)
• Vertex Color Palette
• Smoothing (ProBuilder Only)
• Merge (ProBuilder Only)
• Mirror (ProBuilder Only)

Material palette is pretty much needed and as I've said without smoothing/merge/mirror making any reasonable-sized level is PITA. Not to mention there are more necessary stuff marked as "probuilder only".

 

Thanks! I didn't know Kholat was based on UE4, there is a good chance I'll get that at some point. On the other hand though, the reason why I did't suspect it's UE4, is that it doesn't look like something that wouldn't also be possible with Unity.

 

Live without ProGrids? What is this madness?

 

I played ARK on a laptop from 2014 without issue. Just had to disable a bunch of the fancy effects I'll try running it on the new machine later tonight and report back if a titan fairs any better.

There's a handful of flags you can put in steam when launching a game and then it runs fine, so it probably is all effects. I think running dx9 or dx10 disables some of the effects and you get a huge boost.

 

Totally fair, and I agree. Given all the things happening right now, Unity's focus should be elsewhere. Doesn't mean I can't fantasize about it though!

Another noteworthy thing about ProBuilder is how unbelievably active the developer is in his forum. There is not one single question, comment, or general piece of feedback that doesn't get replied to in a matter of hours. I honestly don't know how he does it, but that right there is some LEGENDARY customer service.

I don't know about should or shouldn't, but Unity is pretty good about figuring out what asset store tools are used most often and integrating those systems in some way for future releases. I know Rune got picked up for Mecanim after he did his Locomotion (which was the most utterly amazing tool I've ever seen, and FREE). Then there were the NGUI guys. Didn't they get picked up to help with the new uGUI? I remember reading that somewhere but I wasn't sure if that was just a rumor or an actual fact.

Anyway, the point is it may be the case that Unity is looking into integrating ProBuilder just like the examples above. Honestly, I can't see why they wouldn't at least investigate that. I'm sure there's a lot more bureaucratic red tape that goes into it, so I'll defer to their judgement, but I'm crossing my fingers.

 

Thomas from Chromatica studio, SSAO etc is working hard on the new Unity post FX as well.

 

Great article! Kudos to the OP for taking the time to explore the struggles and successes on both sides. This felt really fair and should provide ample opportunities for both engine teams to fix some critical issues. Grats!

Gigi

 

The only constant is there's no "perfect engine", I thought the article was extremely well thought out too. What is exciting, with all the competition, they seem to be exploring new and amazing ways of doing things. Engines are more powerful than ever and once the ball gets rolling on the newcomers, it'll be an exciting time.

 

It uses a 4.7 if i am not wrong, some things have improved in last releases.
Also they use not a lot of optimisation and their custom real time lightening using their own iplementation of lightprobes that can do IBL lightening combined with distance ambient occlusion with high settings, this is different than using lightmass. They are also reasons why the game is demanding.
You can take a look at 40:57
www.youtube.com/watch?v=NFM1s91Nseg

 

I don't know, I find the process extremely inconvenient, because as far as I can tell there's no function to grab one side of object's bounding box, and drag it, scaling the object accordingly, with snapping (imagine 3d rect transform, pretty much). There's probably $35 package on the store which addresses that, but why bother? Something like that should be builtin.

Ouch.

Thanks, but I have no interest in ark. I hate jungle in video games. Had enough of it since playing the first FarCry and Tresspasser (which also had dinosaurs, by the way).

 

But is it that far fetched though? There is already a ton that has to be done in Unity to set up animations, and with mechanim, I don't see why you couldn't take a generic rig and animate it directly in Unity. Then with a generic rig, why not drop it on a model to rig it there without having to deal with importing to the modeler or miss-naming/miss-identifying rigs in Unity.

As much as I think Unity shouldn't go down this road just because the editor isn't really designed for the workflow, I don't reject the idea outright because there would be some benefits to it all being in one (like seeing the real output from an in-game camera while editing UVs/materials).

 

Given people are barely capable of handling a complete 3D package, I'm not sure how that makes game development easier. The complexity of these packages are fairly insane.

Far better is fantastic live interop with an app per monitor, for example.

As everyone's moving toward that, you'll probably be doing this too.

 

I also have no interest in ARK, i pointed the video to show the tech they made could be one of the reasons why the game is so ressources demanding.

 

People that can handle comlexity should have the perfect tools, IMO.

And spend eternity waiting for the engine to reimport something every time you make a change.

 

That would be a thing that should be optimized away. As far as I understand it, the type of workflow they talked about implied instant live-link/export from 3D application to the engine editor. And since the usage paradigm between different pieces of 3D software can be pretty different, I think having the modelling toolset external and interchangable makes more sense than e.g. forcing everyone to use the same set of integrated in-engine toolset. Blender's approach might be great for you and me and we wouldn't mind it being in Unity, but a Modo user has different preferences. I believe Modo doesn't even have a modifier stack like blender does. To me that's crazy, I bet it goes both ways.
I wholeheartedly agree though, that the toolset unity offers for basic moving things around, is severely lacking and in dire need of being improved or replaced.

 

If you look how CryEngine Designer tool is capable or making complex modeling and UV, i think Unity should really bring an advanced Designer tool that really speed up your game creation.
1:10

Unity could bring the same Designer tool, extruding faces, beveling for example are not very complex things.

 

I am shocked Unity haven't prioritised a Blueprint type tool. It's still on the "Investigations & Design Work" stage of the Roadmap. It would stop many people migrating away.

If they're trying to democratise development, that's the #1 feature that will do it. The assets could also be easily shared amongst users, or even put on the asset store.

As for adding things to the editor in general, I think it's only going one way. 10 years from now editors will have more features built-in, not less. They shouldn't be reliant on third-parties for anything, if possible.

 

Or just buy company behind ProBuilder and integrate it into main engine, which initially could be as simple as making PB free on AS (not talking about Baby's First level design tool that dares to call itself Probuilder, actual probuilder), and provide it with unity onwards, just mkaintaining it so it works.

 

The guys have done an amazing job with probuilder.

and check out Polybrush

edit: oh a bit pricey... I would've purchased if it was 60.

 

Heck no. Something like that can be an icing on the cake, but not the primary feature. unreal is attractive because of visual quality + performance you get out of the box + source code access, not because of blueprints.

 

Interesting. I wonder how well this and probuilder could work together. They are practically at the point where there is no point in selling things piecemeal, since all of it is pretty much integrated together.

 

It will be some primary feature for all people that don't know how to use a modeling software or are used to work with Source 2 or Hammer for example.

Until that happens, Lumberyard will have it out of the box and perhaps will get lot better editor and engine features.

 

They do have a bundle but PolyBrush does not appear to be part of it.

https://www.assetstore.unity3d.com/en/#!/content/15447

 

I was talking about blueprints. Not about CSG level modeling tools and such.