CaptainLoveline
Member
The next big leap is raytracing in 2025, running at a completely reasonable 15FPS.
Crysis' crowning achievement(s) were physics, which affected gameplay and allowed for non-linearity with regards to approaching combat scenarios.
In addition the title had no real ceiling with regards to scalability.
It was innovative and well ahead of its time, but has since been surpassed in most if not all areas.
Maybe developers can start making better, more realistic AI? I feel that graphics always evolves but braindead AI always remains. I know good AI is pretty hard to make but I think that will be the biggest next leap in games.
I wish we had a game like this
Where do I recall seeing this game?? Was it shown recently at E3?
Remember Brink?For me I want to see developers try to bring unique aesthetics to life rather than obsess over hyper realistic graphics.
Raytracing for sure. It's still a long way off, but it's a total gamechanger compared to traditional rasterized rendering.
You can do some really interesting stuff when you're able to define the behaviour of individual photons in code.
Is real time raytracing in games really the future? We've been able to raytrace in realtime for years now, although always less complex scenes/effects than what could be approximated through rasterisation for the given technology.
Would it be true to to say this would also be true of the future as well? In 2025 we may be able to raytrace GT Sport in it's current form ...but rasterized rendering methods would have moved on significantly too
TBH I think the next big leap will in A.I. You know how you read about neural networks all the time? Well IBM, Nvidia, etc are all developing hardware specifically for that. Those digital assistants are just getting smarter etc.
https://m.youtube.com/watch?v=ZHoNpxUHewQ
Imagine being able to simply develop points of interest for A.I. Characters or to be able to have A.I. With enough understanding of an environment that it could describe how to get to a procedural destination or make up motives on the go or lie to you. The models driving A.I. Are getting better all the time.
https://m.youtube.com/watch?v=ZtP3gl_2kBM
More dialog less mocap. Etc. the biggest issue game devs currently fave is that stastical learning systems require large gpu loads and games already push gpu limits, but with A.I. Specific processors hitting the market and lowering the gap we could see amazing thing.
Imagine neural style transfer inside a game:
https://m.youtube.com/watch?v=Khuj4ASldmU
More extensive voxel renders could also help.
The next big leap is raytracing in 2025, running at a completely reasonable 15FPS.
Boss★Moogle;253060295 said:Don't expect much progress next gen. Next gen will be true 4K/30fps with not that much better graphics than this gen sadly. As long as TVs continue to gain pixel count, extra GPU is always going to go towards rendering those extra pixels rather than more complex geometry, lighting, textures, framerate, effects, etc... I'm hoping that 4K will still be the standard when PS6 comes out so we could get 4K/60fps but I doubt this will be the case.
Is real time raytracing in games really the future? We've been able to raytrace in realtime for years now, although always less complex scenes/effects than what could be approximated through rasterisation for the given technology.
Would it be true to to say this would also be true of the future as well? In 2025 we may be able to raytrace GT Sport in it's current form ...but rasterized rendering methods would have moved on significantly too
Didn't the tomorrow children use ray tracing in its lighting?
Yep, in game animations have been hot garbage, foot planting and basic kinetic movement haven't advanced much since 2005 besides maybe sports games like NBA2K.None of those graphics example impress me.
At the end of the day, next generation graphics have done nothing to elevate graphics beyond adding more post processing, higher resolution textures and applying too many fake filtering techniques to try to hide a basic scene.
Take Witcher 3 for example, if you turn off all the post processing filters, the game's geometry and asset are basic at best.
Games have not made the generation leap as seen from 2d to 3d. There are still terrible animation, clipping, graphics popup in all these examples of next generation graphics. Take away all the facade filters and games today are no more complex than they were 10 years ago.
Raytrace/marching based global illumnation. Lighting plays a major role in how realistic a scene is and raytracing is like the endgame for lighting. To get a taste of how good "traced" GI can look, check out Kingdome Come: Deliverance with it's SVOTI [Voxel based GI, Cryengine].
High quality dynamic global illumination. That's what'll really set next gen apart from this gen. There's some limited solutions already in games like QB, but I think overall lighting will be the biggest difference when you start comparing generations of games.
https://www.youtube.com/watch?v=cH2_RkfStSk
Actually after playing the horizon expansion I noticed something that bugs me a lot but I always forget.
It normally happens in cutscenes.
Objects on clothing with physics don't a sudden jump when the camera focuses as if the character just landed.
Bloodborne did this too when on lifts the coat would spazz out and usually clip through the character.
Id love for things like this to be fixed.
I thought its last of 2 in engine stuff?
So what is next? Will next gen see vastly more sophisticated shaders aswell as an increase in everything else?
What graphical feature will set apart next gen
From the current one? Or will next gen simply be more of what we have had this gen? And PS6 Will be the big leap?
I wish we had a game like this
I wish we had a game like this