Unity's new Area Lights etc

Thought these were quite nice. You can try the demos for yourself and see the comparisons between raytraced reference and Unity's up coming lights. I'm not sure how it stacks up against lux or alloy, but it looks quite fast to calculate and quite accurate. Can have it textured or any poly shape apparently?



Here's the link https://labs.unity.com/article/real-time-polygonal-light-shading-linearly-transformed-cosines

Thoughts?

 

Does this mean Unity is better than Unreal now and the debate is over?

 

Dunno about that, I love all engines tbh. Why wouldn't I as a game dev Blitz 4 lyfe.

 

They need to brush up on their marketing. That title is an awful way to say "area lights".

 

Lots of big words....But it looks super cool.

 

I see the quad area lights. What I really want to see is an emissive material actually emit light like area lights do. Is that part of this upcoming tech? It looks pretty, but that request would be top of the line.

 

Like this?

Around 2:35 has it... I guess it would need some fun from enlighten to really take shape.

 

pretty impressive. Not sure how resource hungry that is though...

 

@Ryiah

Yeah, if you want to read something about area lights in English: https://www.unrealengine.com/blog/simulating-area-lights-in-ue4

@hippocoder

It's super cool, I often use diffuser cards.. It really just depends how heavy it is and do we have to do anything special with shaderlab when we create our own shaders? Might be a dumb question but it does mention GGX BRDF approximation with LTC, hopefully we can just define GGX and that's about it..

*Cough or get a material editor..

 

No. But it probably means that mixed mode lighting will remain broken despite all the new features.

 

So this exactly. Why has Enlighten been basically ignored since its debut? At some point we were supposed to get Enlighten reflections and Enlighten destructible objects. And what happened to progressive light map baking? Thought we'd see some more of that at Unite instead of an hour and half of talk about how they're raising the price of the they're products.

I don't want to complain, but I'm getting skeptical of Unity's direction. I'm not sure why they are working on things like these area lights (which already basically exist on the asset store) instead of working on creating things that people developing for the asset store can't make. Like, for instance, improvements to Enlighten or integration with PowerVR. The only obvious answer to me is that their deals with Enlighten / PowerVR don't go quite as deep as they originally believed and now they've got to cover up for their mistakes. And if so, this leaves me questioning Unity's future as the best possible option for game development. Its unbelievable to me the amount of stuff they've presented at some talk labeled as coming soon only to come out with it years later.

I know integrating all this stuff is ridiculously complex and I know they don't want to come out with half backed implementations of stuff, but lets be honest, thats most of what Unity 5 has been. Don't get me wrong, there is tons of stuff I love about what Unity has been doing, their peripheries have been great, integration with analytics was beautiful (aside from the limited information offered online) for instance. But just the meat and potatoes of their engine seem to be going in a different direction than I'd like.

But perhaps my goals have changed over the years and maybe now Unity is pursuing different things than I and that is neither good not bad. Regardless, I'm getting a little worried and more than that a little sad about the possibility of where I and Unity will stand in five years if it continues on this trajectory. I hope I'm wrong though, I really do.

 

Could we have like a mesh sphere that emits light? Like could I model a lightbulb and then make it a light and it all works great?

 

I am more interested if they play well with the enlighten and upcoming powervr solution.

I'm guessing they don't.

 

For the impatient, the WebGL demo is very cool!
http://blog.selfshadow.com/sandbox/ltc.html

 

I dont understand the use of the word impatient, because a unity sample project that allows us to use these area lights in our own unity projects already exists:

https://labs.unity.com/downloads

 

I think this is awesome and the guys working on this have done an amazing job.

I don't agree with unity shouldn't bother because of there's the same thing on the asset store.
NOT AT ALL!!

This is the right direction and I'm grateful that they invested time and money into this. Big huge thumbs up.

Enlighten's issues are far more complex than me or the average users understanding.. so lets not just throw out there they should have fixed enlighten instead, or that they haven't bothered to try and fix it. that's so far from the truth that it's annoying.

 

You could do that in 5.0. Put an emissive material on something, wait for a looong time for lightmaps to compute.

 

Now all we need is a volumetric fog that works with those area lights, like we saw in Unity's last cinematic demo... only available for all mortals out the box, and Unity's ready for CGI film production!

 

But i want to produce game

 

Too late!

 

Wait. Are you suggesting that the company who designed and developed the engine, and then integrated Enlighten into it should not focus on fixing the problems in it? Or are you suggesting that because you don't understand the intricacies of the problem that no one else understands them?

Just as food for thought, some of the users who are most vocal about Enlighten are known to have built their own engines from scratch. It's not that much of a stretch at all to believe they know the subject well enough to comment on it.

 

Nope. They should've fixed enlighten instead.

Here's a problem:
http://forum.unity3d.com/threads/when-is-mixed-mode-lighting-going-to-be-fixed.332250/
Hasn't been fixed for one year or more.

The feature (seamlessly blending dynamic and static/baked shadows) is among of things a basic 3d engine is expected to be able to do.

 

Yeah although I have time of day so that wouldn't help me any, so I kind of want Unity to just improve shadow performance or store realtime shadows a bit better, with an approx for distance. This can be similar performance to baked under the right circumstance (but not so useful for mobile).

 

i know that unity have a dynamic lightmap etc, but still they have a static lightmap and until now i still don't understand why they not have a static/baked shadow system?

anyway it seems this thread starting to derail. . .
or not?

 

It's a general lighting/graphics thread when it comes to the etc part

 

so are you saying that you understand the issues on why enlighten is not working the way it should? could you explain in some detail why these problems exist? what are the solutions Ryiah? maybe you or the vocal people could contact Aras and give him a few tips too... it should be easy righ
Ryiah, if i learn to make a game engine from youtube tutorial maybe then ill be able to fix these enlighten issues, what do you think?

do you think unity just dropped working on enlighten and shadows ect and decided we will focus instead post effects?

do you think that by making a simple game engine that you can now make a game engine like unity? no you or the vocal people you speak of cant.. if you could have you would have.

its not as easy or simple as focus on this problem now its fixed, and focus on another problem now thats fixed...

 

i checked a while ago 5,4 beta and that problem was fixed. did you check it yourself?

 

I have no idea why enlighten is a problem for Unity. It would be enlightening if Unity staff could explain why.

 

no not quite fixed technically, shadow blending still off, and static lightmapped object still rendered in forward

 

One of the points about this area lights stuff is that they've used a technique/implementation novel enough to present to the world at siggraph in a month or so.

People are quite right to have gripes about other areas of unity but its a huge mistake to blur the issues on that front with the work that unity labs and unity demo scene makers do. Unity the company like everything else has finite resources, but lets not pretend they are a single team sitting around in a room neglecting your priorities so they could make this stuff instead. The reasons enlighten doesn't do what people want are to be found elsewhere - its frustrating not to know the exact reasons why, but jumping on what other unity stuff goes public, just to fill the void and reach a conclusion at the expense of accuracy is a mistake. Better to accept that we don't know the reasons why some parts of Unity are allowed to languish in poor states, and that we should appeal for clarity on those fronts rather than join the wrong dots and start moaning about other positive stuff that does get done. Personally I'm very unhappy about the state of the speed tree shader and the lack of communication about instancing being built into standard shaders during unity 5.4 beta, but whats that got to do with the people who have done this area light work? Should I resent everything good that Unity does from now on until they happen to do exactly the good thing I've been waiting for? No.

 

I'm going to hazard a guess right now that Unity would like to build this new sort of area lighting into Unity 6 in a fully featured & well integrated manner.

 

It is marked as fixed while still being broken. Check the thread, it discusses this.

Yes, it would be easy. If source code was available.
Problems you mentioned do not fall into "extremely difficult" category. Programmer of average skill will be able to deal with those.

Your offended tone is not useful. There are loads of problems in unity that persists because nobody is workign on them (for years), and some other things were given higher priority. However, assignment of those priorities is often questionable.

Problem with mixed mode lighting existed for for at least one year. Slow baking for enlighten existed since the version 5.0. A lot of people asked for beast or similar lightmapper. Those people were ignored.

So one would have to assume that unity is either unaware of those issues or not interested in fixing them.

Also, all this was discussed one thousand times already.

 

actually that is their plan

 

Those are hardly the only two possibilities.

I'll throw just a few more into the mix:

Over a particular time period they have problems getting exactly the right member of staff with just the perfect skillset to nail a tricky problem.

They end up with any one of umpteen possible issues with a third party tech provider. Some technical, some about the relationship or terms.

A problem needs fixing and they care about it, but the correct solution relies on something else needing to be put in place first, and there is a problem getting that prerequisite sorted.

I am aware that many of my points may cast me in an 'over-generous to unity' light considering that peoples gripes are often well founded. I'll make another post shortly that should rectify this imbalance.

 

The time period is "one year". And unity have 900 employees.

Look, at this point I have no reasons to assume that any of existing long term issues will be ever fixed.

I'll just use whatever is available now. However I'm not going to waste time expecting a bugfix to anything to miraculously materialize in foreseeable future, and will just assume that all things that are broken will remain broken.

Also, It's been discussed too many times already, and I'm not in the mood for chatting about it again.

 

You negative tone is not helping anyone at all.
lol all you need is the source code and you could fix it hahahahah now thats positivity you.

 

So, many of us probably know that when Unity go on about democratisation, well thats hyperbole, and a more accurate label for what they describe is simply 'lowering the barriers to entry'.

But since they are fond of using the democratic language, I suggest that you cannot have democracy without accountability. Where is the accountability on the forums for specific unity systems? Why don't multiple key people at unity have a small part of their official job include responsibility to actually keep an eye on posts about particular areas of the system? I'm not talking about the equivalent of bug reports or paid support, but the more general flows of valuable conversation about the general direction a feature is going in, when it might improve, etc? We know there are a bunch of people at Unity who are more than capable of doing this on unity forums at times, and its worth its weight in gold and goodwill. But it happens far too sporadically and whole areas of Unity end up far away from this picture of healthy knowledge-sharing. Some specific technologies in Unity have featured sustained conversations on these forums that feature truly embarrassing absences of voices from anyone at Unity. It speaks volumes and it really is toxic when it ends up leading to lack of goodwill seeping into many unrelated threads.

 

I'm skilled. What about you? I do not exactly see a shining programmer's wisdom radiating from your statements.

 

Also note that my suggestion about forums does take into account the reality of what message boards are like.

It's not a suggestion that loads of people at Unity spend way too much time monitoring forums and getting dragged into infinite conversation loops or issues of the human ego.

Rather, it is the suggestion that large elephants in the room not be tolerated, and that the 'big issues' which clearly emerge all be dealt with in time by the right people and the right amount of information. In terms of scale I'm not talking about dozens of posts per day needing to be addressed, more likely a dozen or less issues/areas of technology need more upfront and ongoing discussion per year, or even less.

 

lol no you wont see shining programmers wisdom coming from me.. hahahahah... no i leave that to you.. you da man!! give da man da source code so da man can fix it yo!!!

 

i think unity's new progressive baking tech will solve these issues. looking forward to it

 

Why are you acting as if that is the most outrageous thing you've heard all year? It's not that insane that if given source code, a decently skilled coder can improve *something* in Unity's code. I'm not saying it's trivial, but it's not ROFLMOMGWTFHAHA either.

 

A problem is that source code still doesn't necessarily get you everything. If Unity don't have permission to distribute source code for all 3rd party systems that are used by unity, there will still be black boxes that we cannot peak into and tinker with.

 

well because he hasnt even seen the engine source, how can you assume that an average programmer can fix unity's issues?

 

The same way you assume the notion that one might fix *something* is OuTrAgeoUS.

 

Now I'm going to have to do something cool with these area lights in the weeks ahead so that I can post screenshots to this thread that take up at least as much space as all my waffle about the other subject thats emerged here.

 

nope... i assume that an average one that hasnt got the experienced required would be able to fix unity's issues.

 

So much for the moderator's own thread being on topic lol

 

It can be saved. Have you tried the sample unity project yet? I got a brief go on it when it first came out yesterday and am just about to dive in for a longer session, yay

 

So uhh... Area lights, huh!

I guess they are a supercharged version of what Alloy does, with alloy being limited to circles and tubes.