6 ms·
sli and crossfire. Double the cost and higher fps for equal or negative latency gains. at least in talos 2 and highest quality fsr, 120 fps feels like 70, mayb
by beebeepka 3y ago
sli and crossfire. Double the cost and higher fps for equal or negative latency gains.
at least in talos 2 and highest quality fsr, 120 fps feels like 70, maybe 80.
- HumanOstrich 3y agoThere's something else going on then such as your settings or that particular game. FSR2 and DLSS work within the same frame cycle after the frame has been rendered (much faster at a lower resolution). So latency is the same at the same FPS, and latency is lower at higher FPS, all other things equal.
- beebeepka 3y agoI think your conclusion could only make sense with vsync on. Otherwise, doing something could only be slower than not doing something.
- HumanOstrich 3y agoYou're oversimplifying things to the point of being willfully ignorant. Yes, of course "doing something" could only be slower than "not doing something". But we're talking about 2 somethings involved in generating individual frames. 1) Generating a big slow frame or a small fast frame. And 2) Not doing upscaling or doing upscaling. Generating a small fast frame + upscaling CAN be faster than generating a large slow frame without upscaling, so it is possible to get the same perceptual visual quality at a higher FPS. Again, this is an intra-frame operation. Latency is introduced independently of frame generation rate by frames being queued or delayed in the pipeline. For example, the new DLSS 3 frame generation feature needs 2 complete frames to generate the one in between, so the first one is queued, and the remaining 2 are delayed. Delays in processing inputs can also be a big factor in latency. Since vsync happens outside of individual frame generation, having it on or off doesn't matter for our argument. I found plenty of results in Google demonstrating that upscaling doesn't affect latency, or at least not enough to be perceived by a human.