UE4.5 released...

That's cool, but I was also wondering about an interior scene, if you have several mesh shaped like lightbulbs, have them emit light in the scene. Would they be able to give you any decent shadow?

And if you need a pointlight for decent shadows, then you end up with the light from the pointlight (for shadows), and the light from the mesh(for GI). So maybe it ends up being too much light and your shadows not blending well with the GI.

Anyone have any experience trying something like that?

 

It's the nature of the Engine. UE4 it's a "clean written from scratch" engine. Also they don't need to be compatible with older versions of the Engine. Unity on the other side still have lot of compatibility stuff inside (even internal project upgrades). Unity is improving his 10 years old engine, UE4 is pretty new (besides they've stated that most their old stuff was again re-implemented in UE4 but still a lot cleaner). And so results in a much tighter/cleaner codebase thus makes future implementations/improvement/features easier and faster to do. Plus the fact that Unity have to deal with more services and platforms. I bet if Unity6 will be entirely written from scratch, the results will be similar to UE4 but you'll lose your ability to re-use your old projects in Unity6.

 

I've been encouraging Unity to make breaking changes for 5 at every turn because I see the wisdom of it - you end up with a much better piece of kit. However (sadly) many users aren't smart enough to see that and urge unity not to change. Frankly, it's illogical of them to act like this because:

1. Unity 5 has different features - you would be changing or adding code regardless
2. Unity4 still exists

Given those simple observations, there's no harm in making breaking changes for major version numbers, and in any case let's be honest - it's not going to take more than a couple of weeks to port even the most demanding title, since the language is the same. For most people it would be a day. That's nothing given a decent title will take a year or more.

But yeah, people fear what they don't understand.

 

Had to go look up the acronym. interesting stuff. This pdf gives some good pictorial comparisons and gives an outline of how it works..
https://www.cg.tuwien.ac.at/courses/ARTR/slides/shadowmapFiltering.pdf

 

There's a few other shadow changes - ie draw call performance increases and you can set the ranges at which things cascade.

 

Enlighten doesn't raytraces any dynamic shadows or light. It doesn't even compute direct lighting, it only does bounce indirect lighting (stored in low frequency lightmaps that you have to pre-compute). Dynamic Shadows and direct lighting are normally computed on the GPU.

 

Yes, the entire 6 months its been alive versus the (round out to a decade) life cycle of Unity.....

I truly feel sad for the people who dislike someone else's koolaid bringing better iteration models and community inclusion. Things are so much better in a black box with no time tables, right?

o.0

 

I'm wondering who those "un-smart users" are?
In case you didn't catch up, I just pointed out why Unity vs UE updates timeframe are the way the are. Let me remind you that I'm an alpha Unity user and I follow the progress of Unity internally very closely and they often share with us internal changes that people "who are not smart enough" don't see.

 

Yup unreal has only been around for 6 months. Better get a stronger drink...

 

Whoever's daft enough to stand in the way of progress

 

Not that I necessarily disagree with you, but Unreal 4 has been in development since 2003. Here is an article from 2005 where Mark Rein states it has been in development for two years.

http://www.computerandvideogames.com/123639/rein-weve-been-working-on-unreal-engine-4-for-two-years/

On that note, it was developed exclusively by Tim Sweeney until mid-2008. So that's when I consider its development to truly have started. Prior to that it would've been essentially the equivalent of a research project.

http://www.tgdaily.com/business-and...unreal-engine-40-aims-at-next-gen-console-war

 

The takeaway from the original comment (to me) was here is a project that is screaming for things that should have been there from the start, things only users can "ask" for, much like with Unity requests. They built it for themselves first, of course.

And it took them all of 6 months to give those things to the community, in iterative and interactive community builds, not behind 2+ years of "its coming" non-transparency *cough* UI *cough* for one salient example.

Haters gonna hate, great, but it is sad they dont seem to get the benefit that Unreal is currently bringing to the table -- adapt or die.

Unity had no clear competition in this market segment i.e. prosumer market, now they do. It is an all around benefit to us as they now should have a proper motivation to make stuff happen sooner, more in line with what users actually want, and probably more transparent.

People bemoaning Unreal for no reason other than its cool to do are really just ignoring *yet another* benefit it has brought to the table.

 

I don't think anyone thinks otherwise. We're all expecting Unity to adapt.

 

Two things smarty pants, did I say dynamic? And another, it does raytrace shadows from directional lights:

"Instant Baking Enlighten’s baking workflow has all the benefits of the real-time authoring process, but with the added functionality of saving the results into high-quality lightmaps. The radiosity is computed dynamically with Enlighten, and the direct lights are ray traced to give the highest quality shadows in the final lightmaps. This is extremely fast and flexible."

Read the .PDF

http://www.google.co.uk/url?sa=t&rc...PH2C6mB52KExai5yw&sig2=HXz9afyvEb92-ru-D16o-g

@Archania

It's been available long before that with NDA in beta. Not that it really matters or anything..

 

You are talking about static lightmapping that can be baked with Enlighten (the same as Beast provides) but the way you stated makes people believe it's dynamic shadowtracing which is not! Smarty pants!
PS: I don't need to read any online manual, I actually work with Unity5, doh!

 

You're not the only one, you still can't admit your ever wrong can you? .. Ahhaha!. You need to show me an update of your game chief.

Also I'm holding all shortcomings of Unity 5.X on your shoulders, because of your close interaction you know it's not going to turn out well..

I'm joking of course ..

 

If I'm wrong, I always admit!
Our soon to be published project on Unity4.X, iOS with a custom renderer (dynamic lighting and shadowing). We got however a secret sauce in the works with Unity5!

 

To be fair, I have been critical of various aspects of Unity as well. I'm definitely not a koolaid drinker. There are things I like and things I dislike about both Unity and Unreal Engine.

I have been running UE4 along side Unity 4 since UE4 was made available to the general public about 7 months back. However, that does not mean UE4 has only been around for 6 months total like you implied. UE4 has been around much longer than that. The hot code reloading should have been implemented from the start.

I am very impressed with UE4 in general, and I am very impressed with Epic's pace of development. I will call it like I see it, though. The automatic C++ hot reload feature should have been implemented from the start. It is fair for Epic to take some flak for not implementing that right away.

 

They must be have a deep pocket.....:/

 

I believe the reason it didn't get higher priority is that Epic appears to have far more interest in Blueprint. Most of the time when they've brought up programming on their Twitch streams it has been related to it.

 

I'm personally confused. Unity seems to be the simplest thing out there that you can pick up and learn to make modern games that can be taken seriously (if you take your work seriously, anyway). Unreal Engine seems to me, to be graphics, graphics, graphics.... etc. So, take me a lone developer or maybe I have 2 people working with me on a project (night owl development, primarily) what exactly would I do with AAA graphics capabilities? I can't create AAA quality graphics.

It just doesn't make sense why anyone who didn't have a pretty significant budget would be working with Unreal Engine. Or, most confusingly, why they've made it so cheap that anyone can use it even though (technically speaking) anyone definitely can't just pick it up and start making Gears of War with it.

What's happening these days?

That's my $0.02 anyway.

 

Have you tried it? It's not that hard to use, sure you might not hold your hand and cuddle your way through, you do have to know at least the basics. But apart from that, it really isn't that hard..

What would you do with AAA graphics? Well you have the lighting and all the tools you could nearly ever need, learn how to make AAA artwork like the other lonely developers on the UE forums are.

In short I swear down a couple of ideas "came from Unity" and a lot of things are far simpler to achieve than in Unity, material editor is boss, so is the cinematics and shiny VFX tools also socket system, want a sword you say? Attach a socket for it. Time of day system you say? People struggle with the plugs to do that sort of thing, I did it in 7 blueprints.. Sure it needs some work, but hell it really didn't take much.

Unity is far simpler for some things, for others it is much much more complicated. It's a matter of getting used to it.

 

It seems like UE is geared toward rapid development of a couple types of games. Mostly though, those types of games need those good graphics for people to enjoy them. Unity is really an engine with an editor. So, in Unity I feel like I'm using tools I would have probably made on my own anyway, if I could have. Unreal Engine seems like a game, you don't program your game you mod an existing game. I think in Unity I could make an RPG, a simulation, a platformer and it's really all up to me how it looks and feels. I see Unreal and I feel like I'm playing Forge on Halo.

Learning how to make AAA 3D models? That's a real job people get paid good money for. Graphic whores are too picky, hard to please. That's why I'll stick with my goal of making my own kind of games, it's achievable. Competing with the titles that are out there made with unreal, that's a pipe dream.

 

Unity is lot more simple to use and learn, in UE4 you have notions to handle like spawn, spawn possessing ,blueprints etc ... i just find they kept their old way of dealing with things from their old baggage programming FPS games in their past. They should have get rid of spwan things and just deal with game objects.
And Unity is the best to target the more large Mobile audience as UE4 has many issues with Android and only supports last generation ones, i find it's too limited, specially for simple 2D or 3D games.


About UE4 the speed of versions is fast and impressive, BUT like some dev director said on twitcher , it will need one or two months of full bug corrections instead.
What i can see is like each update, you find :
- a good amount new bugs because it's stable but not enought tested
- new bugs on previous stuff that wasn't here
- new blueprint changes making your work on a previous version no more valid with last version
I hope they will find the time some day to stop racing for new features , deliver less graphics ones , and instead improve the editor and bring more bug fixes.

To resume, you wil need to subscribe more than one time to benefit editor and engine improvments and avoid having to change too much things picking up some update after many months of non subscription.
But their clear open roadmap discussions and more frequent public updates instead of one per year is what we will need in Unity

 

I'm making an RPG in Unreal, it works like other engines and it has handy tools and I wouldn't really say it bends specific ways, sure the super shiny corridor scenes run better due to lightmass but that's about it. I'll agree with zenGarden, Android is still firmly in the realm on Unity. Sorry but you saying modding is a load of crap , there is nothing you couldn't do in Unreal if you're a game developer with some experience, also you have the option to get your hands dirty if the something doesn't suit your particular palate. You can start off with a blank scene and work your way through, it's your option..

If you're brave you can even smash CryEngine round the back of the head and force it to do whatever you want, but that's not for the faint of heart, Unreal is no problem.

UE4 is a great alternative, it has it's quirks (mainly bugs that really are pissing me off lately)..

 

You know what over the span of years I used Unity, I rarely hit that many bugs. I think this has put things into perspective, ok ready?

Adding some statements in Mat editor like constants (Crash to desktop)
Terrain in general is a mare to work with, Can't blend colour maps seams due to vertex calculation issues, removing trees until very recently = CTD, foliage rendering on there shader model is still horrible with overdraw as a whole performance is crap. Removing / adding trees until recently = crash to desktop.

You can only really get away with weightmaps, SM-Blend overlays colour maps no go, Splat maps no go, you have to convert splat to weightmaps to even stand a chance. Just found that one out lately, UMG plenty of crashes, particle system is good YAY! Lightmass, plenty of crashes and odd issues when upgrading (E.g. shadow maps look like a rock from a game in 1996).

Importing .FBX files, which I now believe has mainly been resolved caused a lot of crashes. Shaders constantly calc and are extremely slow, but even when they calc sometimes they don't update so you have to re-calc.

Is this a good time to stop? It is still rough around the edges I won't lie, very powerful though and looks amazing gotta give it that.

 

They have some really weak areas that should not be as UE4 should no more be FPS corridor games only, i hope they will improve that soon.

I just hope they will deliver less things and test lot more, also breaking less existing things like blueprints for example.These bugs are because they just do basic testing and deliver to people the last tech they worked on without making you wait some months to try it.
So the users participate in that way pointing the bugs or issue they didn't found.

But at least , it's better finally, as you can use latest tech they just crafted

 

As you said though, they really need to just cool off a bit and spend a couple of months fixing bugs.. It's not big or glamorous but it keeps customers.

 

It never crashed since i started using it (early beta 2013). As of todays version I've never encountered something that will stop me working with, it's more than rock solid and fully stable (PC). You got a project upgrade in between releases, API wise, you get more stuff to play with, if you don't get into modifying engine sources your gamecode won't break at all.

 

You mean UDK (UE3)? If you ever tried UE4 you'll see it's totally different from modding a game. It's a general purpose engine, decoupled of any game specific type/behavior.

 

Not really. There is a lot of useless stuff geared towards the general purpose multiplayer fps game, yes.

 

There's nothing related to FPS in the engine, either TPS. You got templates kits or game examples but they are independent from the engine and tools.The networking system in UE4 is an authoritative client-server architecture where you mostly define which property, function or actor will get replicated on the network. It have nothing to do with any game specific code or behavior.

 

Guys. I think what is best is for all of us to broaden and expand our knowledge. Limiting it and locking it down has no benefit to yourself.

How about carrying on with your project using whatever you started it with and then as a hobby go use everything, all of it. Click it, apply it, change it, remove it............why lock yourself down? Why use just one tech and defy the universe and anyone in it for using something else.

Be the guy/gal that uses all of them; with prowess and proficiency. When you look at it the only person you hurt when you lock yourself down is you.

 

@tatofever :
Calm down , no one bached your UE4 tool ,some of us are "multi engine" , not one engine fanatic only.
You know from the "fifth element" movie : MultiPass



A simple example , UE4 is today very far from Crysis performance an open world features or Unity is lot better for mobile or more easy to program and UT5 will be a serious alternative for advanced graphics.
I use UE4 on a team project and Unity for two personnal games, each as it's own strong points for different project configuration and needs.


Their spawn and spawn possession among other are somthing i don't like , it's herited from Fps days, they should use some sort of game objects instead.

All points that Shadowk pointed are real, and i encountered many of them, Epic should fix some lightening issues, performance, try keep stable Blueprints instead of delivering things no more compatible some times.
I had some ten crashes since trying different things :
- crash on lightening intense calculation
- crash trying packaging (for example their FPS futuristdemo)
- crash with Android packaging
- a project no more usable and lost after updating it to a new version
I don't say that happens in a reugular way, but that happened , it depends on what you do and your project.

Yes UE4 bring on regular updates, but please do more testing, try to keepin compatibility as much as possible between versions.

All people should use at least three engines , they would feel less provoked when someone brings bad points

 

If you didn't read my previous post, UE4 isn't our main tool (at least not right now), besides being an early UE4 user, we are using Unity4 and 5 doh! Take your fanboyism insults back please.

 

Nope i don't remeber all posts i read, i just didn't understood such defensive position stating all is great in UE4 ,as i had good crashes and such fast updates bring lot of bugs and like they said on Twitch they should do only bug fixes during two months if they could.
(it wasn't insults also , only some humour, relax )

 

well, maybe its just me but when I went to un real to try something different from unity, I could not figure out how to do their building area, could not figure out its 3D grid if it has one, could not do anything really except copy and paste what ever some body made for me to copy and paste but when I first came to unity I had no problems with anything and I knew that this is the way a semi profesional and profesional 3D game builder from SCRATCH is suppose to be like. create cube or sphere create terrain, CREATE every thing else my self and if i need something created thats more complicated i go to asset store and buy it.

i think unity wants people to make and create and LEARN from scratch as much as buy their assets, unreal, i did not like it and i do not know who could unless they r only in it for the money and allure of its powerful engine but not learning about computer science or computer codes and scripts.

i know what i like and dislike and why i don't like it, i would chose unity over unreal for reason i just described.

 

@Amon

Agreed, I heavily used CryEngine for personal projects.. I've used Torque, built my own worked on some AAA in-house engines it's fun. But even though I love my tech, focus shifts when you're trying to make a game. Now I just want it to work ..

But you can't really get a feel for pro's and cons of a certain engine until you see how the competition rolls.

@aaronhm77

Yeah, it's a bit like switching from Blender to Maya, it takes a bit of getting used to definitely..

At least you tried it and appreciate what you have with Unity, good going.

 

well its probably a little of both but some people are only in it for the money and do not even know what real love is; UNREAL does not care if I learn and create and own but they only care that they make money and they own, they do not give a damn about me, unity does care thats the difference between love over gold and gold over love,

unreal=gold over love

unity=love over gold i take love over gold any day

 

Unreal is quite a different beast from Unity. In the case of both engines I watched Zak Parish's video tutorials ahead of time. Epic's YouTube channel for UE4 and 3D Buzz for Unity. Both did a good job touching on the user interfaces.

 

I'll say this the reason Epic do care is if you do well they get a chunk of your earnings, they don't care about $20.00 a month, that's just to tide the staff over. Whether that's right or wrong, they have a reason for you to do well..

 

Well, not quite right. Blueprints changed it a lot. Our art guys are loving blueprints, they were not able to code simple stuff in C# (this did not feel right for them), but now they bring levels to life without any help from programmers. I'm not even start talking about how much more user friendly is matinee, cascade, material editor, persona, blueprints, vehicles, etc.

There are quite a good amount of issues, UE4 needs 1-2 versions to iron these out (two months ).
But if you have time, produce good repro. project and you can see your bug fixed in 48h and you can merge it into stable branch and recompile engine with this fix and you are good to go (no need to use some beta / preview build).

Two weeks ago we were messing with UE4.5 preview version and ios8 metal, and it had good amount of issues. We decided to left it alone until official release and revisited old Unity project after 6 months of UE4. We wanted back to UE4 after getting out of memory crash when opening first scene...worked well in older version of unity, but it looks like it was really at the limit back then.

 

right;

and to do well all any one has to do is copy paste some prefabed "pipes" with over 1000 to chose from and all ready made avatar and BAM! UE made $1000 dollars.

UE is an engine with copy and paste prefabs, not a game builder if you ask me.

I wont be going back

 

I have decided I hate UE with a passion, certain aspects about certain things can turn me off completely,

I cant stand UE

 

I don't even have a clue what you're talking about here? Prefab "Pipes" like prefabs in Unity?

 

I like the rapid iteration time with less QA testing because most of the engine's bugs are hardware specific. Epic could spend a lot of time testing every feature on every PC setup but it's faster just to get a version out the door and have the users test it for you. A couple thousand developers can find bugs a lot quicker than a dozen QA employees and they generally release a hotfix within a couple weeks of the initial version.

That said, anyone working on a serious project should wait till the end of the version's life cycle to upgrade. 4.4.3 seems to be the most stable release so far.

 

Not so helpful if there's a shipping showstopper you find and a deadline to meet.

 

I wouldn't mind at all. One of the main advantages is they quick iterate features and instead of you guys worrying if it's good or not and spending years tweaking, we can let you know if it sucks straight out the gate, or if it's great but needs tweaking .

The problem with UE4 is some of it seems more like it's in Alpha, a bit of a compromise is the best bet. I don't mind bugs in Beta, constant wide issues across the board is a bit much. Some features they hide so you can enable and test, that's one way of doing things which is cool.