3 ms·
I would like to see more rigorous studies done of refresh rates, especially w/ and w/o (free|g)-sync. 120 was definitely huge for me, but I'm skeptical that 240
by thot_experiment 3y ago
I would like to see more rigorous studies done of refresh rates, especially w/ and w/o (free|g)-sync. 120 was definitely huge for me, but I'm skeptical that 240 vs 120 would make a material difference, especially w/ gsync, which I believe is much more impactful in real terms.
- redox99 3y agoGsync basically doesn't matter if you're fine with disabling vsync, and at high FPS the tearing is almost invisible.
- thot_experiment 3y agoG-sync isn't exactly about tearing, it's also about latency. The thing that matters in game is click-to-photon, which is shortened by syncing screen blank to frame availability.
- redox99 3y agovsync off has better latency than gsync https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/9 https://blurbusters.com/gsync/gsync101-input-lag-tests-and-s...
- thot_experiment 3y agoSure sure, if you don't care about tearing or you're rendering at an absurd framerate. In real world situations where your framerates are not 2x your monitor's refresh rate and you care about tearing g-sync is very helpful.
- mmh0000 3y agoBack when I was cool, I used to love games like Quake 3 and Counter Strike. I could absolutely tell when a game reenabled vsync after an update and I absolutely hated vsync; made everything feel laggy/"smoothed". I don't recall ever noticing tearing.