Support NeoGAF

[Haggard]

Feel free to tell me I'm wrong but isn't a potential huge advantage of real-time RT that the engines of the future will make it much quicker and easier to get good looking lighting v pre baking? Surely huge for all devs but especially indie devs.

I know a talented artist can fake lighting etc these days so that RT doesn't look much better but that surely takes a massive amount of time and skill compared to what could potentially be a few clicks in a future engine. We just need the hardware to catch up which won't be this gen.

100% correct.
The day after the official development targets reach a level where global RT GI becomes the base standard is the day a lot of level designers and asset creators will wake up with a massive hangover from the party the night before.
The amount of time that goes into faking what RT just "does" is unreal.

 

[Dr.D00p]

I know a talented artist can fake lighting etc these days so that RT doesn't look much better but that surely takes a massive amount of time and skill compared to what could potentially be a few clicks in a future engine. We just need the hardware to catch up which won't be this gen.

Right now we are only seeing very limited RT lighting in games, because that's all the current hardware can handle, that's the only reason why 'smoke & mirrors' trick lighting, meticulously placed by scene artists can still give the real thing a run for its money. Only when games move to a fully path traced render pipeline will we see the full potential of RT in games.

..But don't hold your breath, That's still at least a decade away, if ever.

 

[Shmunter]

Feel free to tell me I'm wrong but isn't a potential huge advantage of real-time RT that the engines of the future will make it much quicker and easier to get good looking lighting v pre baking? Surely huge for all devs but especially indie devs.

I know a talented artist can fake lighting etc these days so that RT doesn't look much better but that surely takes a massive amount of time and skill compared to what could potentially be a few clicks in a future engine. We just need the hardware to catch up which won't be this gen.

You still need to thoughtfully light the scene irrespective. I would wager using RT may make things less controlled for many devs. Add some light and you’re at the mercy of the simulation. Do you call it a day or do you start making guesstimate changes to reach the desired aesthetic?

Currently throwing in some non RT point lights may actually be more predictable in outcome, despite not behaving true to real world physics.