Why is this graphical problem so ignored nowadays when we have consoles capable of reducing them by a big margin??
Pop in has always been one of my biggest pet peeves about graphics. Almost every AAA game out now has them in droves, with only a few devs making any effort to mitigate or hide them from the player's face..
I'd rather see a subpar texture than a whole field of grass popping in out of nowhere in front of the player, is there any reason devs (and players) just don't care about it?
So, I'm curious why you believe these new consoles have unique powers to drastically reduce pop-in that is being ignored?
PS5 and Xbox Series have much faster IO, and that helps a ton, and then virtualized geometry techniques like Nanite can mitigate pop-in visibility. (It actually in a way "increases pop-in" since every object is recalculating its surface constantly based on the master mesh and the visibility of that object in a scene, but the change is on a micro level instead of the defined LODs before and so ideally you will never notice the difference even though it's changing way more often.) UE also has introduced and will continue to create solutions for virtual/dynamic assets in textures and shadows (and audio), same for other engines. The basics are getting better.
However, we're still dealing with multiple layers of quality passes and interdependent textures/shaders for intricately detailed scenes with large numbers of objects in a scene and complex detail (and varied approaches of how to apply that detail to a material) across an often open world, under a limited amount of compute speed... it's a lot. More than before, even with more powerful hardware. (You've even got dynamic shaders and realtime shaders and other elements creating scene-dependent elements on the fly or on-call, and if the timing is off, you'll get a compilation pop.)
Nobody has solved the problem, there's not a one-size-fits-all solution that developers just are too lazy to bother with. Everybody is working to squash issues as well as they can, and the engine manufacturers are coming up with global solutions which knock off the biggest annoyances. Sometimes development schedules affect what they can put in the box or what they can test before launch and so they patch afterwards with greater data reports, but it's not that they didn't have However, there's no magic bullet. In fact, I can't find it but I remember a different tech chat from around the Nanite reveal that was complementary of the geometry solution but still warned that, as games get from 10s of thousands+ of shaders and permutations to millions and beyond, timing issues will get worse not better because it's impossible to actually solve for virtually infinite combinations of variations. And hopefully the answer will be in finding ways of best hiding what cannot technically be eliminated.
aras-p.info
Why did ID Software's Megatextures idea never catch on?
