In Unreal PT would get a similar type of indirect area shadows like in the images posted. Lumen amazingly can also get that type of indirect area shadows with the difference that they completely lack the contact part - instead of being sharp at the base, they already start very very soft or blurry.
And about the correct looking DOF, I meant to say I already know how to build that, I already did it back in 2010. I don’t wanna post the link to my website but you can find it if you search “cryengine film tools”.
But maybe someone else already did that, or maybe the engine can achieve that already, in that case I would prefer to not spend weeks trying to build something already possible with Unreal.
This is a great demo someone on my linkedin did. All of you can try it, and check how poor performance and noisy Lumen is for an actual game scenario: Corridor by Dylan Browne
PS: Will DLSS 3.5 save Epic from noise? (So hyped about this)
1 Like
Maybe DLSS 3.5 and it´s ray reconstruction could solve some problems, like noise and ghosting, at least for RTX users 
:
Although I’d like to see it running live in a game (sans video artifacts), I think this sort of intelligent ray denoising could make a very big difference in certain game scenarios.
The bottleneck is, of course, the fact that it’s proprietary. If it can’t run on consoles, Epic’s probably not going to implement it; especially considering that it requires specific hardware acceleration to run. However, if there is some way they could implement a NN to do specific reconstruction work, it could potentially make a big difference in RT lighting quality.
Sure. But if Epic is incapable of achieving it, it will be a solution for, at least, some (a lot) of the users, usually the high-end ones (which, anyway, will usually be the only ones using heavy ray tracing)
I also am not exicted about or interested in the frame generation, but Frame generation and Ray reconstruction are presented as being independent from each other 
But i also don´t have to worry about that, because I cannot use Frame generation, since i only have a RTX 2070, so if anything, i could only get ray reconstruction, and thats the one i am excited about.
And since they used Cyberpunk as one of their testing grounds, i tried to find that place, they showed in the video, so that once Phantom Liberty and DLSS 3.5 gets released, i want make a comparsion to what it looks for my system.
So it will be interesting to see, if Ray Reconstruction really will be a separate option to turn on and off, just like its right now with Frame Generation and f.e. DLSS Super Resolution.
And i found it, its in Heywood - Vista del Rey, close to the Metro Congress and MLK fast travel point, if someone else want test it ^.^
How it currently looks for me ingame:
And a pathtraced screenshot made with the Photo-Mode:
How i read that image, the important part is in the brackets right below those listed categories, and under Frame Generation is still “RTX 40 series GPU” listed.
Thats how it currently looks in my options menu in that section:
As you can see, DLSS Frame Generation is greyed out for me, because i just have a RTX 2070, so no chance to get that, and DLSS 3.5 would not change this. It even says it ingame, that this requires a RTX 40 series card.
But Ray Reconstruction is listed for all RTX cards, and all info i could find for that points into the same direction, that it is independent from Frame Generation, just like f.e. Nvidias DLAA is a separate feature and a separate option in the game.
Take a look at this screenshot from Nvidias video, there Ray Reconstruction is only tied to Super Resolution, and Frame Generation comes way later in the Pipeline as basically the last step in the chain.
Well, we will know for sure in one month
1 Like
In that image you can see that FG is only for 40 Series, and only since DLSS 3 and forward.
Ray Reconstruction, which is what we are talking (and excited) about, will be included for all RTX cards, since DLSS 3.5 and forward.
Probably it will be available as a spare setting to be enabled/disabled, but maybe with the need of using DLSS, or DLAA too.
1 Like
nvidias (and AMD/Intels) “toys” will never be a solution to anything, and in fact they are the very thing causing the issues - because instead of optimizing stuff, people are like: “yeah, whatever”
There were games recently, that had a 2080/5700XT as minimum system requirements, WITH DLSS/FSR einabled - for 1080p! 
This isn’t related to anything currently being discussed, but I did want to share my wonder at something: Upgraded an old UE5EA test level scene to 5.2, and I am blown away by the performance difference. in EA, this scene (HWRT, what counted for hit lighting at the time, ETC) wasn’t getting a consistant 20FPS, but largely CPU bound. Now, 5.2, I’m hitting a stable 60 with much higher settings. Their work was truly astonishing. That scene included emissives, tons of tranlucent and alpha-tested effects, volumetric fog, all of it. Even with SWRT, the performance and visual quality are way better. Between streaming improvements and the lumen scene feedback mechanism, reflections by and large hold up really well.
I guess all I’m saying is that I’m blown away by how much better lumen has gotten in such short time. The technology really has come a long way, and it’ll keep getting better.
Truly a riddle that solves itself
Which devs are you talking about? Game devs, or the engine programmers?
Game devs, or the engine programmers?
Reading the documentation. Most game devs that are releasing UE5 games with trash performance in low 40s without lumen.
They should optimize their games, the same way you should optimize yours too, not only asking the engine programmers to do all the optimization work.
2 Likes
Yeah, I wanted to see “the scene”, if it had anything special in it - but it doesnt, its really just a simple scene where the trees dont even have leaves etc.
Also, please use cooked shipping builds for performance comparisons, in editor and standalone is meaningless.
Addition to this
I was interested in this because your and my scenes overall partition in terms of “milisecond distribution” is as different as I expected. (I pay a lot for translucency etc. that you dont have there at all, so my performance numbers are “fine” if that is taken into account.)
All of the debug-stuff running in background is completely screwing performance to an extent where the data shown in stat unit etc. are worthless.
If I do this in Editor, I gets “20ms” (for gpu), which would equal 50 fps, but in reality, I do get 60-61 (1440p that is) - which is a big difference, just from making a build.
My game runs with 20-25 fps in Editor and standalone (while showing weird numbers in stats), while the shipping build runs as fast as your hardware allows it to
Also, you already noticed: Engine was downsampling stuff in viewport, so I dont need to point at it.
I also noticed large visual differences between the two versions, as if they are not running on the same settings at all. (PS: Anti-Aliasing wouldnt hurt 
)
I would go and take a look at the city sample, but its meaningless because of my 6900 XT. (It doesnt really seem to care much from what I gathered.)
I also have been fighting with UE5s new physics the entire day… lets say, it was exhausting. 
This heavily depends on the scene.
yes, in a simple scene with just a few buildings etc. it will work - but as soon as you introduce a decent amount of foliage and other things (which games usually have), you are paying a lot more for translucency than in the city sample etc.
While I agree, that the recently released UE5-games are jokes in terms of performance, especially since they dont even allow to disable Lumen and just say “get a 2080 and turn DLSS/FSR on, you consoomer”… the city sample really is “simple”.
I have no idea how they manage to pull this off.
Even I managed to make the game run “just fine” on 10 years old graphics cards as long as you leave lumen disabled, with nanite enabled…
In fact, someone recently made it work on a GT 1030, allthough the 2GB VRAM couldnt handle the Trees and the fps went trash when he got close, but otherwise were fine. (I think he also used TSR, but then again, a 1030 is really potato tier - a GTX 960 is like 4x as fast.)
It is certainly possible to get “the best of both worlds” with small tradeoffs on both ends, but some effort has to be put in.
I see, I’m with you now. I want to avoid bad-talking the engine devs, I think with all the considerations of the next-gen, including Unreal’s newer usage as a film production tool, Moore’s law making performance gains gen-on-gen smaller, and the staggering tech base they support, they’ve done a pretty phenomenal job. That’s not to say there aren’t legitimate ways the engine could do better, but in a lot of the analyses I’ve seen of UE games with bad perf, it’s way more often how the devs optimized their work than any sort of fundamental engine limitation (with shader compilation stutter being a partial exception).
1 Like
One of the key reasons that lumen runs as well as it does in the city sample is that all the meshes are modular for culling and watertight against eachother-almost no overlap, which keeps BVH traversal low. Never mind the far-field letting them cull thousands of instances while still keeping distant GI. If anyone’s kitbashing large, complex meshes together without proper HLOD, they should expect HWRT performance to suck.
Perhaps this is unnecessary, but I think developers and the public are getting the wrong message if they look at these these tech demos and only take away 'Unreal can do this. What I wish Epic did a better job illustrating is ‘Unreal can do this, only if you tune your content, optimize your scene, and set your visual bar accordingly’.
1 Like
What is it, specifically, that you think you need in order to better optimize your project?
Lumen has probably over a hundred cvars exposed that allow you to control everything from the downsampling factor of screen probes, the placement and resolution of the radiance cache, the resolution of the surface cache cards, the density of the translucent lighting volume probes, whether to do spatial filtering between radiance probes… the list goes on and on… Individual components of lumen can be enabled/disabled wholesale even.
1 Like
Wait, you said you had two 3060s. Are they connected via SLI?
One thing I will never say about UE5 is that it lacks optimizationability.
Every now and again I run through the CVar list for lumen, just to see if there’s any way to claw performance back or enable new features. You can absolutely make it run better and look better if you can dedicate the time to it. I’ve gotten to speak with a handful of AAA devs who were in the early stages of getting their upcoming titles working in UE5: they were trying out every CVar they could to claw back performance at an acceptable overhead. Optimization doesn’t look like your PPV settings as much as tuning the engine to have exactly what you need, and carry the cost of nothing else.