That’s like saying that “physics simulation is the only technique that produces accurately shaped water streams” - technically true but generally not a sufficient improvement over the shortcuts currently in use to make up for the downside that the technically most precise method is slow as fuck.
Game making is at all levels finding shortcuts and simplifications (even games about the real world are riddled with simplifications, if only the gameplay rules being a simplified version of real world interactions because otherwise it would be boring as shit) and in the visual side of things those are all over the place even with RT (the damage on the walls, the clouds in the sky, the smoke rising from fires or the running water on the streams aren’t the product of Physics Simulations but, most likely, the use of something like Perkin Noise or even good old particle effects to fake it well enough to deceive human perception).
Yeah, sure RT is, technically speaking in terms of vidual fidelity alone, better than the usual tricks (say, using an extra rendering step for the viewpoint of the main reflective surfaces such as mirrors). Is the higher fidelity (in, remember, a game space which is in many other ways riddled with shortcuts and simplifications) sufficient to overcome its downsides for most people? So far the market seems to be saying that it’s not.
Now, it should be stressed that this is a build of The Witcher 4 specifically designed to show off Unreal Engine’s features. Yes, it’s running on a standard PS5, but it’s not necessarily indicative of the finished product.
So that’s like saying “under laboratory conditions it has been demonstrated to work”.
If you know what to look for you can notice it (mainly light bouncing of objects and tainting shadows with the color of those objects, such as the shadow above the green canvas here), but the difference to the non-RT version when one doesn’t know what to look for is minimal and IMHO not enough to justifying upgrading one’s hardware, especially considering that so much of the rest (the water in the streams, the snow in the mountains, the shape of the mountains themselves, the mud splash when a guy is thrown into the mud, the folliage of the plants and so on) has those visual shortcuts I mentioned.
Yeah, sure, it’s nice than shadows next to strongly lit colored surfaces get tinted with the color of that surface, but is that by itself worth it upgrading one’s hardware?!
When most games with RT in them deliver that performance on one generation old hardware and all environments, then you will have proven the point that for most gamers it has no significant negative impact on performance.
Depends on how you define “accurate”. Even full ray tracing is just an approximation based on relatively few light rays (on an order of magnitude that doesn’t even begin to approach reality) that is deemed to be close enough where increasing the simulation complexity doesn’t meaningfully improve visual fidelity, interpolated and passed through a denoising algorithm. You can do close enough with a clever application of light probes, screenspace effects, or using a second camera to render the scene onto a surface (at an appropriate resolution).
That’s true, but after a few frames RT (especially with nvidia’s ray reconstruction) will usually converge to ‘visually indistinguishable from reference’ while light probes and such will really never converge. I think that’s a pretty significant difference.
Reflection probes are one way. Basically a camera drawing a simpler version of the scene from a point into a cubemap. Decent for oddly shaped objects, although if you want a lot of them then you’d bake them and lose any real time changes. A common optimisation is to update them less than once a frame.
If you have one big flat plane like the sea, you can draw the world from underneath and just use that. GTA V does that (like ten years ago without RT), along with the mirrors inside. You could make that look better by rendering them in higher resolution.
Where RT is visibly better is with large odd shaped objects, or enormous amounts of them. I can’t say it’s worth the framerate hit if it takes you below 60fps though.
I haven’t personally played a game that uses more than one dynamic reflection probe at a time. They are pretty expensive, especially if you want them to look high resolution and want the shading in them to look accurate.
Screenspace isn’t the only way to draw reflections without RT. It’s simply the fastest one.
Most gamers aren’t going to notice, and I can count on one hand the number of games that actually used reflections for anything gameplay related.
What I’m talking about is drawing accurate reflections and I don’t know any other technique that produces the same accuracy as RT
That’s like saying that “physics simulation is the only technique that produces accurately shaped water streams” - technically true but generally not a sufficient improvement over the shortcuts currently in use to make up for the downside that the technically most precise method is slow as fuck.
Game making is at all levels finding shortcuts and simplifications (even games about the real world are riddled with simplifications, if only the gameplay rules being a simplified version of real world interactions because otherwise it would be boring as shit) and in the visual side of things those are all over the place even with RT (the damage on the walls, the clouds in the sky, the smoke rising from fires or the running water on the streams aren’t the product of Physics Simulations but, most likely, the use of something like Perkin Noise or even good old particle effects to fake it well enough to deceive human perception).
Yeah, sure RT is, technically speaking in terms of vidual fidelity alone, better than the usual tricks (say, using an extra rendering step for the viewpoint of the main reflective surfaces such as mirrors). Is the higher fidelity (in, remember, a game space which is in many other ways riddled with shortcuts and simplifications) sufficient to overcome its downsides for most people? So far the market seems to be saying that it’s not.
CDProjektRed just showcased The Witcher 4 running RT with 60 fps on a PS5. Bullshit its too slow to be available for most people.
From an article about it:
So that’s like saying “under laboratory conditions it has been demonstrated to work”.
If you know what to look for you can notice it (mainly light bouncing of objects and tainting shadows with the color of those objects, such as the shadow above the green canvas here), but the difference to the non-RT version when one doesn’t know what to look for is minimal and IMHO not enough to justifying upgrading one’s hardware, especially considering that so much of the rest (the water in the streams, the snow in the mountains, the shape of the mountains themselves, the mud splash when a guy is thrown into the mud, the folliage of the plants and so on) has those visual shortcuts I mentioned.
Yeah, sure, it’s nice than shadows next to strongly lit colored surfaces get tinted with the color of that surface, but is that by itself worth it upgrading one’s hardware?!
When most games with RT in them deliver that performance on one generation old hardware and all environments, then you will have proven the point that for most gamers it has no significant negative impact on performance.
RT was three generations ago, and I don’t think they really vary the number of rays much per environment (and rt itself is an o(log(n)) problem)
If you think that video is representative of the release game’s actual performance and fidelity, I have several bridges to sell you.
I don’t see them lying but that’s on you I guess
They literally did this with The Witcher 3. https://www.youtube.com/watch?v=vZZgL3wnkmE&t=2m56s
Depends on how you define “accurate”. Even full ray tracing is just an approximation based on relatively few light rays (on an order of magnitude that doesn’t even begin to approach reality) that is deemed to be close enough where increasing the simulation complexity doesn’t meaningfully improve visual fidelity, interpolated and passed through a denoising algorithm. You can do close enough with a clever application of light probes, screenspace effects, or using a second camera to render the scene onto a surface (at an appropriate resolution).
That’s true, but after a few frames RT (especially with nvidia’s ray reconstruction) will usually converge to ‘visually indistinguishable from reference’ while light probes and such will really never converge. I think that’s a pretty significant difference.
Reflection probes are one way. Basically a camera drawing a simpler version of the scene from a point into a cubemap. Decent for oddly shaped objects, although if you want a lot of them then you’d bake them and lose any real time changes. A common optimisation is to update them less than once a frame.
If you have one big flat plane like the sea, you can draw the world from underneath and just use that. GTA V does that (like ten years ago without RT), along with the mirrors inside. You could make that look better by rendering them in higher resolution.
https://www.adriancourreges.com/blog/2015/11/02/gta-v-graphics-study-part-2/
Where RT is visibly better is with large odd shaped objects, or enormous amounts of them. I can’t say it’s worth the framerate hit if it takes you below 60fps though.
I haven’t personally played a game that uses more than one dynamic reflection probe at a time. They are pretty expensive, especially if you want them to look high resolution and want the shading in them to look accurate.