I’m a Twinmotion/TM user, I mainly use unreal engine for the same reason I use TM, for Archviz.
So i’ve watched a couple of videos, and i’ve followed them properly to a T. But my scene just doesnt seem to show any realism when rendered. I thought this could be because of my laptop, but when I tried to render Megascans Godess Temple, it looked great, i was even surprised that my laptop could render that. (I’ll upload the two still images down below)
I mainly use dynamic light I believe (if what dynamic light means is “Movable” and not “Static”) and I dont bake light because I assume that I could just render the scene using MovieRenderQueue. (I’m a beginner)
Also, i dont use ray tracing because my laptop cant handle it. I opened the megascans goddes temple after tweaking with it a lil bit and change from DX12 to DX11 so I could open it.
So, what I want to know is, what’s the problem here? Why is it that I could get a decent result with megascans Godess Temple but couldnt with my own scene? Any info on this would be appreciated.
Btw here is my laptop spec:
Windows 8.1 Pro 64-bit
Intel(R) Core™ i5-5200 CPU @2.20GHz (4 CPUs), ~2.2GHZ
first of all, realism is not always dependent on the engine. Sure, it’s very helpful to have all the tools for making a great rendered image at hand, but for amazing results, there is so much more to watch out for. One very important question you have to ask yourself is, if you want to create a realistic or photorealistic image. There is a huge difference between the two. A photo does not always have to mirror reality. Take Photoshop for example. Many artists create awesome photorealistic stuff with it. But it is manipulated and not necessarily realistic. If you want to try to replicate reality - then it is realism you’re after. In your shot, a lot of things don’t work out the way you want it, because you try to recreate something an AAA-team put together without having the required knowledge. I suggest, not only to watch tutorials, but to attend classes for 3D art first. It will give you a lot more understanding of what is happening in the background and how you can control more precisely how you can to achieve your visual quality. As I said, there is a lot more to it than tech only (composition, asset quality, post processing, lighting, …), but I would advise you to start with the basics first. If you’re rocksolid with those, all the rest will come together like a puzzle.
By the way, I’m actually going for photorealism actually, because the render from Megascans Goddess Temple look Photorealistic, and I want to reach that level.
No problem
If you target Quixel quality: You should really take a look into photogrammetry then. Scans can help achieving realism drastically, because the data was taken from the real world itself.
in your architecture scene you could use an hdri (download one from polyhaven.com) in the skylight for lighting, combined with a sunsky (ue4 plugin) system that gives you even more control. this alone should make your render look much more interesting. then its about adding detail to your surfaces…if you want to have a clean look you can at least create bevels on your geometry on your sharp corners…the little bit of light they capture makes your scene look much more convincing. then add depth of field (camera aperture) to give it a cinematic look and add a colorgrading LUT on top in your postprocess and you will be already quite close to an impressive shot.
