Hi All,

Even though we have a shiny new engine to play with, we are going to struggle to achieve that next level AAA visual quality look if we get our lighting model wrong. To this end I am proposing we have a GameGuru MAX thread, this one, which collects comments strictly related to improving the lighting of a MAX level scene. Don't feel bad if myself or a moderator outright deletes your post if it does not directly relate to solving the issue of creating amazing lighting in GameGuru MAX, but I think it will be worth it having somewhere dedicated to discussing the journey the engine will make from where it is now to where it needs to be.

We start this journey as low to the ground as possible, specifically the ground and the grass growing out of it. I think we can all agree what we have now is pretty 'janky' and one commentator suggested there was no practical difference between what they saw in MAX and what they already had in Classic. Forget the cool feature of spraying multiple grasses for a moment, when we do spray some grass, we want it to look fantastic.



The above is what we get now if a brand new user just painted some grassy floor paint then added some default green grass on top. Of course, we can go in and play with the sun brightness, thickness, etc but I think the goal should be amazing looking right out the gate, no need for some secret knowledge to somehow make it look good. I already have my own suspicions about why the grass looks so bland and dark (and have already started working with our artist to resolve some of the more obvious issues), but I would like to open this up as the first conversation topic of this taskforce.

Finally, as a counterpoint for comparison, my internal team thought this scene demonstrated some pretty good grass:

To start with, as i said before, we need more light types;

Area, Point, Spot, and Directional to start with should be the bare minimum to achieve good lighting in a scene.

And not only that we need more control over those lights, at the moment all we have is the ability to translate them in the world, change the colour, and change the range (which technically isn't range, it's radius). We need more options:

For example the spotlight range should be a distance from the spot, and it should have as well as the current options; Radius and Intensity.

And likewise the other lights should all have intensity.

With regards to ambient lighting, and the sun, the sun should be tied to the ambient lighting, and should be on a slider so if we want to we can reduce the intensity all the way down until there's no light. If we had this in conjunction to a directional light, we could have far more control over lighting, down to even simulating our own sun how and where we want it.

on the subject i'm pretty sure Lee said grass won't cast shows, would you suggest grass should cast shadows to make it look better?

@Lee the 2 comparisons aren't really fair, you haven't got any colour in your grass and it's way darker - a totally different feel
however i will say the biggest change would be the need to remove the gaps from the grass and add back animation/sway... not being able to see the flat textures underneath will already make a big difference - unless the user wants gaps of course in which case the old density slider needs to come back - walking through grass looking ahead is fine, it looks quite dense but if you start looking down it looks very flat and almost disappears.

Maybe shadows might give some sort of 3d looks to the grass, but I afraid it should be too expensive for performance, at least the engine being able to generate those shadows just for the grass being within the player fov.

I think we need the ambience and surface light sliders back, while we have the gamma setting, it really doesn't suite the lighting system if it can't be tweaked for indoors versus outdoors.To be fair the comparison you are using uses most probably HDR while we are using the trusty old bloom. It also has DOF and blur, which max currently lacks.

The way max works now is rather counter intuitive yes it allows you to control the variables of each individual assets light source and marker, with classic you had global scene settings for the light system. While settings each individual settings as needed, for some things like lights there needs to be a global setting that will tweak the light sources in the entire scene or by adding a system checkbox to ignore global settings.

Because we have the ability to change the individual settings of assets, we should also be able to address the individual ambience and surface light settings for a entity .

Adjusting the way the sand pile reacts to the light source be reducing the object ambience and surface lighting one can better blend the entity into the scene, which currently isn't possible and doesn't really blend well with the scene, the artist will have to approach it differently by physically increasing the darkness of the texture used to better blend in with the rest of the scene.

Unfortunately this doesn't look particularly good there is no way to improve this scene without better control over the ambience and surface lighting, it should be a dark scene.

Ability to control the radius of the light would be useful

In this picture the lights circled in red do not spread far enough on the y radius. They should light up more of the area around them.


Currently you can only adjust the radius by making the light range larger, but if you do that it will be too bright on the floor and on the ceiling, because the light acts like a ball. Like this:


Boasting up the light radius in this example gives a better radius but obviously ruins the scene in the process.

Adjusting the light bias (despite it's tooltip description) only seems to affect how shadows are drawn.

The scene is not getting lit by the lights as you would expect a light source to do in real life.


OldFlak....

Actually, you can tweak that close to perfection by enlarging the radius AND reducing the brightness whiel you do so

Can't see any way to do that: Light Range, and Light Bias are the only tweaks I can see. No brightens control available. Perhaps I am missing something?

EDIT Thanks Ertlov that sorts it nicely

OldFlak....

Brightness is controled (as is color and saturation) with the light color. If you move "down" in the picker, you preserve the color but reduce brightness. Let me grab that for you quickly.

So, problem: Light doesnt lit up enough of the room:



First solution: Increase range! Problem: Far too bright, scene messed up:



So I move down the light color to darken it without loosing the yellow tint:



Much better!

Hi everyone,

I found an old demo version of this software in my archives : https://lumion.com/showcase.html#foobox-1/96/Gu6z6kIukgg

Hello
I think its a real good idea to try and get lighting right ,but this is hard to do ,even in Unreal they have conversations about what is the best light and what looks real, and with Unreal they still have to fake it most of the time. Also everyone sees things differently what looks good to some doesn't to others , so you have a real task on your hands. I have a suggestion take real photos and try to recreate it as best you can notice every little detail and then do what you can with what you have and if that's the best you can do than that's the best you can do. You could also make sure that the textures are the best you can possible make ,Unreal uses something called Quixel and they make the most realistic scenes I have every seen maybe you can use the same process I don't know . But I do like what see so far.

You used to have to do the same in GG Classic. However, Wicked Engine does have separate controls for "energy" (intensity) and "range". That seems more intuitive to me.

For spotlights, it would be great to have some control over beam-spread. Spotlights in Wicked have a "FOV" slider to control just that, independent of any brightness or range controls.

I agree with everything Belidos said. I would also add that we shouldn't focus exclusively on realtime lighting, at least not until ray-tracing is on the agenda (GG Ultimate, coming Dec 2028!). Wicked has a nice lightmapper that can calculate light bounce from certain sources, it's tremendously effective at simulating the indirect lighting that brings ambience to most scenes. AFAIK, lightmapping is still heavily relied upon in the industry and it would certainly save resources to have less dynamic lights in the scene. Please consider porting Wicked's lightmapper over, it's a night-and-day improvement of GG Classic's and would help us create beautiful scenes for a fraction of the resources a realtime scene would.



AE

GG classics lights were good, the scaling was good too, but the only thing that let it down was, lack of realtime shadows and the bleed through entities. Max's light should have the same simplicity of scaling intensity and range, and being able to set shadowing on/off for each light. So far the lights seem overly intense.

I actually think the light control is great .. Just decrease the colour and raise the Radius to control brightness, We did not have that in classic and my tests look much better for it. Better control over shadows like the slider in Classic would be welcome.
Sometimes they are just to dark.

I agree with the others who've said we shouldn't rely completely on real-time lighting. That's going to have a real potential for a big negative hit on frame rates if we need to use a decent number of lights to get an area lit up properly. As was pointed out, Wicked already has a lightmapper that looks really good. Why not include it? A combination of lightmapping AND real-time lights seems to be the way to go. Even the "big boys" use both and not just rely on one or the other. And, as was pointed out, the Wicked lightmapper can calculate light bounces, which produces a more realistic effect.

More light types are important, too. If I have an indoor level where the light sources are long, fluorescent tubes, it's not the best to present the light given off with point lights. An area light, sized to the shape of the fluorescent lights (i.e. a long rectangle) will better simulate the light given off than several point lights would.

Despite the fact that you can technically control a light's intensity via it's color (white vs grey, for example), it still makes better sense, to me, to have an intensity slider. It would make sense to me to also have a FALLOFF slider, too.

As a side note:

Currently we have to adjust the range of a light via a slider. This is okay, but I think it would be great if a selected light had a wireframe sphere around it that is as large as the Light Range. This would serve as a visual indicator for how large the light really is. On top of that, it would be great if this sphere (the range) could be adjusted via the SCALE widget so you wouldn't even have to access the light's properties to adjust this. This is how lights work in most game engines. You select the light, you visually see the range via a simple wireframe representing that range, and you can adjust the range via the SCALE widget. Setting the light position and range is mostly what people are going to do almost every time they place a new light. So, giving them the ability to move the light (the move widget) and adjust the range (via the scale widget) simplifies the process and makes it super fast (no need to have to access the light's properties and then move a slider).

Would be nice to have all those amazing lights around the game industry, anyway, we end up facking lights as long as possible, in order to take care of the performance, in a real game.
Always there will be a scene that needs special illumination and the other does not (just fake illumination), after all when the enemies are shooting me, I have not time to notice how the illumination of the ceiling bulb is reflecting over the wine bottle, how the ray tracing is working on, and so on. imho.

And yes, I agree we need better control over lights and shadows, in GGMax.

I'm in the middle of a course at the moment on the art of video game, and there's a whole section on lighting in there, and that statement you made is one of the most disproved comments people make about lighting in games, apparently you may think you wouldn't see it, and technically you don't at least not consciously, the thing is you will notice subconsciously. Lighting isn't just about what you see, what you don't see is just as important and if it wasn't there you would have a nagging intuition in the back of the mind that something was missing. There's actually quite an interesting science to game lighting.

That's true.
This is one of the things I like the most about video games, that it puts all your senses to work, even unconsciously.

Uhm, we do. They are standard in most game engines. Point lights, area lights, spot lights, and the sun. It's not "all those amazing lights" it's the standard you work with in most game engines.



This is wrong on so many levels. Lighting is, and has been, the key to creating great visuals and great atmospheres in games. Even back in the early days (using what light you had or faking it via painting lighting on textures). And, frankly, sometimes that's the point (to NOT notice it). When things are lit properly, it looks right to the eye and, as a result, it doesn't pull you out of the game. Bad lighting, on the other hand, can potentially ruin end-user experience.

Using your way of thinking, why bother making nice textures/materials and nice models for floors, walls, doors, and ceilings? After all, I'm not going to notice it when being shot at! Why bother texturing ceilings at all? How often am I looking up? But what about if you do look up? And what about when the firefight is over? Don't you look around and explore a bit? It's about immersion in the game. And lighting is often key to this.

Oh! Don't let Wolf see your comment about lighting. He'll go off on you, for sure. He's got entire posts about the importance of lighting in games (and he's right).

I want one thing to be clear, I am not downplaying the importance of lighting in a video game, what I mean is that right now in the development stage that GGMax is in is not the ideal time to raise the bar so much, a lighting in accordance with the state of development of the game engine, and then improving the lighting (among many other things) in parallel as the development of the engine progresses, would be fine.

The question is not if these things belong to the standard in the video game industry, that is not being questioned in any way, but in improving the lighting as other things are improved, so that everything advances in parallel.

Regarding whether or not to use high-level lighting in a real video game, well, I think there is a lot to talk about here, that will depend on the video game, the people it is aimed at (high requirements), etc. imho

Uhm ... this is a thread, set up by Lee, for the express purpose of talking about lighting and how to improve it. Therefore, it is perfectly acceptable to point out that other types of lights go a long way toward fulfilling what he's looking for (especially since The Wicked Engine currently supports these light types). Secondly, just because there are other things that need to be done in the engine DOES NOT mean that lighting needs to be put to the side. There are more than one person (currently) working on MAX. Someone else is working on terrain, for example. So, just because there are things that need doing in MAX does not mean that nothing gets done on lighting. Lastly, even if lighting needs to be ignored for now does not mean that we cannot offer suggestions IN THE THREAD WHERE THEY ARE ASKING FOR THEM that could get implemented in the future.

It's not necessarily an either-or scenario here, 3com.



That's just a poor way to think about it. If you offer what you call "high-level" lighting (which is really STANDARD lighting everywhere else but MAX and Classic) does not mean that the end-user HAS to use it. They can choose to just use point lights and be done with it. But to NOT include means that the ones who DO want to use it cannot (because it's not there). To include other light types hurts no one and helps those who need it. To NOT include them does not help those who would want to use them. Thus, including them is the better option.

Don't be so short sighted.

I'm not nearsighted, I'm just realistic.
Since this thread was opened to contribute ideas I have given mine, perhaps I do not explain myself well, the fact is that I have nothing against all the improvements and implementations that can be made in lighting, even allowing each user to use The options you want, for that are the sliders and the configurations, and all that is very good for me, even allowing a user to choose whether they want to gasp in the grass or not; All this is fine, but this takes time, what I am trying to say is that this will not be ready for the next release (alpha, beta, omega or whatever it does), but one day it will arrive I am sure.

Heh. Okay.

Lee didn't start this thread in order to see what could be done to improve lighting by the next alpha/beta release, sir. He wanted opinions on how lighting could be improved in GameGuru MAX. No date or release was mentioned. Concerning what Lee is looking for, TAKE HIS OWN WORDS (in the very first post):



So, Lee, himself, is after AAA quality in the looks of the game engine. To this end, he wants to know how do this via lighting. Thus, recommendations have been made, such as adding in some more controls on the current lights and adding new light types (light types that are COMMON to other game engines). This is NOT an unreasonable request. It is REALISTIC (especially since these extra lights already are a part of Wicked). It IS SHORT SIGHTED to think that it CAN'T be done in MAX until some unforeseen future. I would even venture to guess that it's a bit easier to implement these things NOW, from The Wicked Engine, then to add them LATER, though I may well be incorrect.

It's a pity when someone asks you to dream (Lee asking us to speak our minds about how to improve lighting) and others try to say you shouldn't dream ... at least not too big. Let's dream of big things for MAX, especially for lighting, and let Lee and the team decide what's to be implemented and what to not implement. It's not your place, 3com, or mine to make those decisions. However, Lee did give us permission to make requests here as it pertains to light. You say you are not short sided, but realistic. I beg to differ.

@3com
@Argent_Arts
Hey Guys c'mon
This is turning into your own little debate and not really helping the situation.

You can do what you want mate, me too.

I'm already done with this debate.

Subsurface scattering is a must for a AAA look in my opinion, especially on the first person arms. It would also look amazing on vegetation/leaves!

What I really dont like is that strange shining effect that is on the terrain.




Couldn't agree more!

In my opinion, I think that the textures of the terrain palette could be one of the problems causing GG Max Beta.x crashes.
Indeed the size of the textured terrain is growing fast as shown in this explained screen capture relating to this folder : "..\Documents\GameGuruApps\GameGuruMAX\Files\levelbank\testmap\TTR0XR0"

It's the little things with lighting that will make a big difference, accuracy of shadows, I have noticed that some shadows do not seem to be correct in MAX, under a barrel for instance the shadows appear away from it, making the barrel seem to float. Stopping any light bleed through entities like we have in classic. I'm not sure if we can change light colour in lua within classic, (never tried it) but that feature would be great in Max I think.

Hello
I m no expert but seems to me what Lee should do is take a look at Unreal 5, after all it's the engine that is and will be used more than any other. He could also check out the games like a plagues tale, metro 2033, Dishonored, these in my opinion have some of the best use of lighting ,and think the Wicked engine can come pretty close from what I have seen.

Going back to the original question about the grass:- using no other light sources, just sunlight (because its outside) the grass should look better. And while others have mentioned subsurface scattering and diffuse reflected light from the grass onto the grass, I dont think replicating these mechanisms is required to improve its look to an adequate level for a good FPS. I think a better route may be the quixel one, take some good hires photos of grass, produce from them hi quality seamless textures, select textures based on ground slope and height - then compress. Next tweak the fill routine to randomly rotate the texture tile on usage. Have a single alpha overlay (animated for wind effects) then give the user gui control over all factors of the sunlight and the overlay. Hopefully the end result would be so good that actually painting 3d grass wasnt needed. I dont think 3d grass is needed because while a plant/shrub/tree casts a shadow (sometimes) grass doesnt cast a noticable one anytime.
Please dont bother flaming me for the above - it is just my thoughts relating to the question posed, and I completely agree that looking at the visuals from other games is a good guide to the achievable end result, but NOT a guide to how to achieve it in Max.
Thanks Lee and the team looking forward to the next Beta

I've noticed the detailed terrain lighting/atmosphere (etc) sliders are absent in GG Max, which were there in Classic. I used to be able to make wild changes to how the terrain was rendered (I could give it coloured glows and such) with the sliders, meanwhile Wicked's sliders seem more tame and just includes sun settings. The colour grading is a nice addition, but not the same.

Is there any way in lua to compress textures and light settings more than usual and load them like winRAR?(i think there is a way to load them faster and add true color for darker scene.)

@Supe. No, unless Max radically alters it's lua system, which I highly doubt. You can load in images for sprites but not textures or objects.