I'm out from UE4 to Unity

tegleg luvs this place :slight_smile:

he’s just mad right now.
took me a while & still am adjusting all the vehicle properties to get something that works, but mine looks & acts 100% better now and haven’t touched any code either.
just from using the templates & some ideas from the dirt buggy game.

anyway … yawn this one’s getting old lol :cool:

Thank you for this thread and for the honest feedback. As humbling as it can be, it’s still incredibly valuable for us to hear firsthand what is working well for developers and what is still in need of improvement. I know that everyone here at Epic is committed to making the engine the best possible technology for developers anywhere building literally anything they can imagine, and it’s just as painful on our side to know there’s parts of the tools that aren’t ā€˜there’ yet. I can guarantee you that if we had much more time and more engineers we would move even faster to address every issues raised by the community, but the reality is that we operate with many challenging constraints, much like any other software development team.

I believe the most important thing for any developer is to pick the right tool for the job and it’s the same philosophy we use internally when evaluating software we use in the process of making the engine. If UE4 isn’t yet the best choice for your project then that’s okay - we will listen to the feedback and we will do our best to incorporate that into our future development plans. We’ve worked hard to prove our commitment to the UE4 development community in the last 9 months and we will endeavor to keep earning your trust over the coming years.

We won’t always get it right at first but we’re also a bit stubborn, so don’t expect us to stop. :slight_smile:

Cool! Ray

but after spending so many years in my past playing WoW, I can’t resist the instinct to respond to these threads with:

Can I have your stuff! lol :rolleyes:

thanks for the response ray

Trust
this is actually a major sore point for me, after spending years becoming fluent with udk (ue3) the sudden death if it as far as ios games go was a big slap in the face to say the least. i have several ios games that cannot be updated due to changes by apple that have not been addressed by Epic, and by all accounts never will be due to the focus on ue4. and to top it off, the last experimental udk build (august) broke vehicles leaving at least 1 of my (released) games dead and unfinished.
will it be the same at the end of ue4’s life?

why i abandoned ue4,

**Performance **
being an indie developer with no budget any game i make has to run well on all kinds of devices, aiming exclusively for high end gaming specific devices is not an option.
during the beta tests of ue4 i was using an aging and it ran well enough to get the job done. as soon as the official ue4 came out it bloated hugely and ran incredibly slowly, causing near catastrophic meltdown on several occasions. it has even killed a friends brand new fairly high end graphics card just from using the editor.

since my focus is more toward mobile gaming any engine i use has ro run well on mid range and older devices. i know that this i a fairly high priority for ue4 development but right now even for the latest and greatest devices there are several major issues. this means that right now ue4 is not an option for me, maybe in the future it will be.

Documentation
blueprints are fine and all but, being primarily a programmer and having the source available it would be advantageous to have decent c++ and engine documentation.
i know you are working on this but right now the documentation and tutorials are almost exclusively on blueprints. having to back engineer the code to find out whats going on eats into development time to the point where i think, is it really worth it.

**Physics **
what can i say, its rubbish. the current physx api has some major flaws, this is not epics fault at all but thats what we have to work with right now.
the vehicles are unusable, if someone were to release a game with vehicles right now there would fairly quickly be some embarrassing glitches all over youtube, thus putting the game (and ue4) in a bad light.
on top of that a lot of the physx api is simply not supported in ue4. you say its because you want cross platform compatibility, in itself is not a bad thing but why not mark those things as ā€˜pc only’ or something so at least we can take advantage of things like fluids and soft body physics.

there is plenty more but those are the main ones.
please No childish abuse from community members, this is supposed to be constructive.

thanks

I may sound rude but have you ever wondered about the quality of your ā€œgamesā€?

and the childish abuse starts lol

@tegleg I still consider you a part of this community, most posts I have ever seen from you have been level headed and to the point.

I do get your frustrations or should I say problems with the engine right now, as I too, like to target as many platforms as possible and I share some of your other concerns as well.

I joke a lot at times, but it’s just to help relieve the tension I feel in some of these threads & to break the monotony of game creation that sets in sometimes.
Remarks I make are not targeted at any individual usually, and if so, they are not an attempt to insult them but to joke ā€˜with’ them and I hope I have never offended in any way, and sorry if I have.

And if you have decided to leave or already have, more or less, then you will be missed and we’ll be sorry to see you go and we hope to see you return some day.

If you can’t handle criticism you might as well stop showing your games in public.

@Errvald
oh dear, please try and be a little bit professional on here.

Talking from an artist point of view I can say that UE4 solved for me most of the problems that I had with Unity, which is game logic.
Blueprints made UE4 a complete tool to work with, and the logic behind the nodes is pretty clear and, most of the time, surprisingly easy to use.

Not knowing scripting Unity for me became a continuous search into the asset store to see if the script I need was already there or I needed to pay someone to create it, mostly because I don’t have time to learn scripting language.
I know that there are some visual scripting tools for it, but with Blueprints I really feel like ā€œat homeā€, considering that the logic is quite similar to Softimage ICE visual scripting.

Graphic power, Blueprints, Marketplace with plenty of cool scenes to study, continuous update of the engine itself and a ridiculous price ( basically 20$ if you want only the engine ) make it perfect for me, and most important, a tool which I can use by myself without relying on someone else and be stuck in a project.

Just my 2 cents from an artist point of view

Anyway best of luck for the switch :wink:

The thruth is, there is no best engine. Each one has good and bad things:
UE4,good source code, community, and graphics. Bad for its limited documentation and generally almost ā€œbetaā€ stage.
Unity, bit more stable, good docs, fine for small games and 2d. Bad becouse its not very high end, and no source code.
Cryengine, good graphics, good level editor. Abysmal coding and documentation, very unflexible.
Project Anarchy, quite cool for mobile games, but completely untested and really really young. Has potential to rival Unity in mobile, very cool c++/lua programming model, only inferior to UE4 but better than anything else i saw.

You cant chose a best engine for everything, if i were to make a 2d game or a mobile, i would chose unity, but i find UE4 better in other cases.

project anarchy is quite young but its basically an editor on top of havok’s (previously Trinigy) Vision engine, and that has been going strong for a good few years. it has excellent documentation and is very stable. after the initial ā€˜farce’ of installing all the required sdk’s its actually quite nice to work with.
its main problem that has become apparent here is that nobody has heard of it.

RE: Documentation -

The current documentation looks like it might be generated from internal code documentation. This is fine, but the developers need to take time to document as they go or this documentation gets stale the same as printed stuff. Generating docs on the build provides an illusion of timeliness because, well, it’s generated every build! But if it’s not updated in the code it’s not updated anywhere else… Customers love short release cycles but allowing an extra week or two for a pass is worth the wait in my opinion, and doing this every release allows for incremental improvements in the documentation in lockstep with the engine itself.

Documentation is the toughest thing - everyone wants it but no one wants to write it! This has been a problem across many engines for many years now and all we can do is continue to push against the inertia.

I can say that the thing that got my attention the most when i tried it, its how good is all of their documentation. Very nice AI library and good animation systems. Seemed solid overall and if i needed to make a mobile game that i cant use UE4 for, i would use it before Unity.

I am using only paper2d and blueprints and my experience so far is this.The capabilities and the scope of the engine are simply unmatched. It gives me the opportunity to create a game with outstanding visuals and game logic with as little effort as possible. Blueprints for me are simply revolutionary and they will change the industry to a more artist/indie/newcomer user friendly experience.

My criticism to the guys is that when you are very smart and try to create something so powerful as UE4 you forget about the simple stuff .What I am trying to say is that you need to pause for a sec and create internally a simple game without you touching any code to fix stuff. Within a week it will expose all the shortcomings of this engine and it will give you a better understanding on how difficult it is for a single or very small team of developers to create something with your engine.For example as I have said before why isn’t there a way to play an animated loading screen with just drag and drop?Or what is the point of creating a super duper new feature that will have very little to add to my 2d game when rendering of sprites is choppy and we all know that this is the first thing a player will see?

I am very proud to be part of this family, the level of commitment you have is extraordinary and over the months I became better, I understood your workflow, I adapted and now I am more efficient than ever.My only major complaint is the slow development of paper2d and the fact that I can do amazingly complex stuff that look with almost no effort and on the same time for example I have to use an external program to quickly create my flipbooks . someone will say and what’s the problem? If you have 10flipbooks it is not ,really. But if you have hundreds upon hundreds after a while it takes its toll.And this is exactly what am I talking about. When you create something new you have to test it in hardcore conditions. In my case the first month using paper2d was ok. A lot of missing features but it kinda worked. As I was adding more and more stuff everything came to a screaming hault. No one had tested before tenths of thousands of sprites. The guys stepped in and fixed it but I lost a lot of time, it filled me with rage and anxiety and at times I felt like quitting. I hope my experiences with the engine so far will help the developers to better understand the differences between large and small ones that do not have the recourses, the experience, and the manpower to deal with small issues that could be very easily fixed by you.
Keep up the good work!

Thought I would add my experiences over the last few months. This is only going to be my experiences which has been mostly around visuals. I have yet to really dive deep into matinee, physics, and other elements.

Background:
I have about 4 Year of Unity Experience and almost 1 Year of Unreal 4 Experience.

I am a professional retail and arch viz designer, focused purely on visual aesthetics. I have used Cinema 4D / 3DS / Modo / Zbrush / Solidworks to Unity and Unreal (Just in case someone wanted to discuss specific workflows). I usually focus on interactive and animation content using Unity and Unreal. I am a content artist first with very minor/amateur level of programming knowledge. I create a lot of virtual versions of real world spaces, such as stores and exhibit halls. In addition I have been creating Augmented Reality Apps, Oculus Rift Demos, and just started with Project Tango content.

In addition I am working on two Indie Games with a Fellow 3D Artist. The first game is a web comic translated to a 3D, mobile, cartoon-y, infinite runner game aimed at young teens. Ideally a free game with minor in-game purchases being released on iOS and Android. Due to the age group and pricing structure, we want to release something that will work on older mobile devices. The second game we are working on is a realistic sci-fi game that takes place in a lot of exterior exotic scenery. I would say like the Uncharted series as far as look and feel.

Over the last few months I have actually taken the time to test out my Retail Animations and Interactives in both Unity 4.6, Unity 5 Beta, and Unreal Engine 4.6. Below is my thoughts as an artist on what I do day to day and what I’ve experienced so far.

Hardware Setup:
Mobile - Testing on iPad2, iPhone5, Asus Transformer Prime, Project Tango, and Nvidia Sheild (A strange collection of what I have so far)
VR - Oculus Rift DK2
Work Desktop - Dual Xeon E5 (24 Cores) with Nvidia 980 Graphics Card (We use this for Virtual Walkthroughs)

First my overall impressions

Unity 4.6 Pro:
Basics - Without the asset store add-ons, for me I can’t achieve anything of the visual quality I require. It is very easy to bring in models and setup all the basics, but this is only good enough for amateur looking visuals. As an artist I feel like I am using ā€œWindows Paintā€ when using basic Unity. When I add Pro and Asset Store purchases, I can take the level to something that is acceptable and closer to a ā€œPhotoshopā€ quality program. BUT since 4.3 they’ve added a new shader language system that maxes out on 64 unique shader names (Sorry I don’t know the exact technical specs) which basically only allow you to use 2 shader heavy Assets at the same time, which has crippled a lot of stuff I was doing previously. Also the shader system can be complex to artists looking to just make high end visuals. With no node based shader scripting, you feel very limited. Shaderforge Asset is a nice node based shader editor for artists.
Price - Sucks for a part-time Indie Developer. It takes too long to get a game out to justify $75 a month Pro Fee, on top of an additional $75 per platform you want to export to. They lock you in for a whole year which makes it even more painful. I’d rather be able to pay for pro and do testing with pro on iOS and Android. Then when I am ready to ship pay a one time fee when I am releasing a game to a specific OS. It’s really hard to test an iOS game with just a PC/MAC pro license, and a lot of times I end up breaking my game if I don’t have Unity set to iOS/Android pro from the beginning.
Arch Viz - Works well enough with the right Plugins ($400 worth of asset store purchases to make it Artist Friendly, $75 a month for Pro Unity). The finaly visuals can still feel stiff and lifeless even with purchasing assets for Screen Space Reflections, Skyshop IBL, SSAO, ETC. It takes a lot of work to get it there, but it can be done.
Product Viz - Works excellent with plugins like skyshop IBL setups. A single object in an IBL setup, looks fantastic.
Realistic Games - With plugins you can achieve an acceptable level of visual quality, but it always feels stiff and lifeless. Playmaker was a must for me to do Visual Node Programming. Skyshop a must for visual quality. Candella is the Only SSR option but I am not thrilled with the developers Overpromises / Lack of Hitting Schedules.
3D Mobile - Works fantastic for this, because you can really leverage the Asset Store to get the look you want on all Mobile Devices. Besides the cost for pro versions, if you setup your assets for mobile you can see on your desktop what your going to get on a tablet or phone.
Augmented Reality - It seems everyone uses Unity as the goto for their devices. Open source AR scripts and Project Tango both rely heavily on Unity 4.

Unity 5 Pro:
Basics - The Enlighten system adds tons of visual quality to Unity. Also reflection probes are a great addition. But the Beta is super finicky, and if you are doing anything robust like Arch Viz, plan to have your machine calculating over and over. It’s hard for an artist because you focus on how lighting and visuals look. Even though enlighten works in the background, it takes too long on complex scenes to make it useable for setting up lighting. I literally will make a change go to the internet or something else to wait for it to calculate and forget what I had changed due to the time it takes. I wish it had a more instant preview mode and a final bake mode. For small games and mobile, enlighten works quick enough, but once you get complex it is a burden. But that said the results are fantastic and so it’s very tedious trying to work in it because you love and hate it. Lastly for a program that has so many great Asset store Plugins, 80% of my go to ones are unusable on Unity 5 at this time. By the time Unity 5 Releases From Beta, I think it will be better, and they have created a stop gap for shader name limits by doubling it’s capacity to 128 unique shader names(Again this it more technical then what I am saying). Also the new Uber Shader helps fix a lot of shader issues for artists.
Price - Unknown at this time if they will change their pricing, I hope they do, and make Pro costs cover all devices not just Desktop.
Arch Viz - Fantastic, the new enlighten system adds life to visuals. You can achieve great results, and if you are running on a powerhouse PC you can crank up a lot of settings. A lot less asset store purchases required.
Product Viz - Even better then 4, except now you don’t need to buy additional plugins.
Realistic Games - Again, enlighten helps boost this. Although for an artist I still recommend getting Playmaker or another node based programming asset. Unity 5 does not add anything to help the Artist when it comes to this or node base shader creation.
3D Mobile - Is the best solution out there. You can acheive both low end mobile and high end mobile solutions.
Augmented Reality - Having trouble with finding updated scripts for Unity 5. Will need to dive deeper into programming to get this working.

Unreal Engine 4.6:
Basics - Since they’ve added Automatic Lightmap UV creation, this has been my goto for high end visuals. While I unwrap my own stuff as much as I can, when you are doing complex stores and interiors, sometimes you don’t have time to go through the unwrap on every single object (Just the nature of my retail design job). Out of the gate, reflections, lighting, camera , node based shaders, and node based blueprints all give an artist a great start. I haven’t bought a single asset for Unreal 4 yet, and I can blow away anything I do in Unity 4.6, and it’s still has a few steps up on Unity 5. When you go in and enable things like LPV and Distance Field AO you can get amazing results for arch viz. Running on my robust desktop it calculates extremely fast and as an artist I can focus on Mood and lighting versus Debugging/problem solving. As for mobile, it has been a great let down for me.
Price - Hard to say right now. For part time indie game development it feels much better, because I have less investment costs to test out and develop. I don’t have to pay extra to test on iOS/Android and it seems to work much better deploying to my Mobile Devices for testing. I haven’t released a game yet on Unreal so I can’t say how I feel about the Royalty. I am also not investing in additional assets to achieve what I am looking to do. This also means I don’t have as many issues with assets not working together.
Arch Viz - You can rival Vray if you know what you are doing. On a powerhouse machine, you can create inter actives that are super impactful. I have had some issues dealing with Stereoscopic 4k, it feels like with some blueprint command nodes I can get it working, but Unity actually still looks better on the 4K tv then Unreal. I am thinking it’s just me right now, and I should be able to fix this as I get more into blueprints.
Product Viz - Looks Fantastic and is really quick to achieve. Faster than Unity to setup.
Realistic Games - Due to the Shader Node System, handling of large scenes, and visual quality I will focus on Unreal 4 for My Realistic Game. This game is more of a Passion Story Telling game for me, so I don’t plan to release to anything but PC at first. Console port down the road maybe. But as an artist I feel like I can focus on the art and mood Unreal 4. But I do miss some assets that Unity Developers have made, but I don’t see it being a roadblock.
3D Mobile - This has been a huge disappointment for me. Unreal makes my stuff look so great, that I tend to go overboard without realizing I am past what Mobile can handle. After I built my first game to mobile, almost all of the things making it look good disappeared and I was left with a pretty disappointing looking game. Worse than Unity 4.6. So I went back and to the T followed the documentation of what iOS can handle as far as lighting and post process . Better looking than the first attempt, but still underwhelming compared to my Unity 4.6 with all the added assets I purchased. They both ran about the same as far as framerate. I think Unreal is getting closer by introducing the new feature to preview Mobile setting within the editor, but I still feel like it’s hard to work in it for an artist. Even on my Shield Tablet it feels like Unity 4.6 looks better. When you compare to Unity 5 With Enlighten, there is no question that Unity wins when it comes to Mobile quality (But at a steeper Price). So I am still on this tipping point of Cost vs Quality.
Augmented Reality - Almost non existent on Unreal 4. For an artist I have no clue how to get anything working, and there is very little documentation I’ve found and even less developers with scripts. I was really hoping Project Tango and Unreal would have more collaboration, but I can only find one or two articles about it.

I know this was long, but I wanted to share my thoughts because I have been living this the last few months. PM me if you want me to elaborate on anything specific.

I left Cryengine and Unity for UE4 last summer because of the price, and upcoming plugins from Otoy (octane). I can say that UE has to this date UE has jumped in quality, resources, documentation, and updates.

I am so pleased that I made the choice to switch to UE. It has really come along way in a short amount of time and continues…

!!

What do you mean by sprite tearing? It’s not an issue I’m currently aware of, is there an AnswerHub for the issue you are seeing?

Cheers,

https://answers.unrealengine/questions/141162/rendering-jerkiness-in-unreal-paper2d-translucent.html

The only thing that matters is if it’s broken does Epic promise to fix it?

If yes then everything else is about the other dog having a bigger bone.