[Disco]

https://youtu.be/ycMWmQ-4-ww

Anti lag izgleda nije ni blizu dobar kao SHarpening, tj skoro pa se ni ne kuži...

EDIT: Eh da, i Tim iz HU je objasnio da ovo nije isto što i max pre rendered frames kod nvidie.
Kaže da je ta opcija u catalystima sakrivena i zapravo po defaulutu na 1 odavno.

C/P:

Getting lots of questions asking whether Radeon Anti-Lag is the same as Nvidia's "Max Pre-Rendered Frames" setting. Rather than making a whole new video I'll answer this question for everyone here. They are NOT the same. Since the Radeon Software Crimson Edition driver update in 2015, AMD has automatically been setting max pre-rendered frames (aka flip queue size) to 1 in most instances. AMD does not have an exposed setting to modify this, Nvidia does. What Radeon Anti-Lag does is delay the CPU's processing of a frame until just before the GPU is ready to process it. To explain further, let's imagine we are rendering at 60 FPS with 16.7ms frames. Adjusting pre-rendered frames will mean that with each extra pre-rendered frame you have, you will gain 16.7ms of input latency. However let's say the CPU can do its portion of the work on a frame within 3ms. Normally with 1 pre-rendered frame that would mean the CPU takes an input, it takes 3ms to process it, then the system waits the remaining 13.7ms until the GPU begins processing the frame. The total delay on the input before reaching the GPU is still 16.7ms. Anti-Lag gets the system to wait 13.7ms first, then the CPU spends 3ms processing the frame, then it's sent to the GPU. So while each frame is still being rendered at 16.7ms, the input is 13.7ms newer, so lag is reduced by that amount. This means we get an input lag advantage even when pre-rendered frames is set to 1. This also explains why Anti-Lag only works when GPU limited. Changing pre-rendered frames will always give an input lag improvement whether CPU or GPU limited, it simply adjusts the number of spots in the queue. However when CPU limited, your system cannot delay the CPU processing data to give an input lag advantage, because the CPU is never waiting for the GPU to become ready. The obvious disadvantage here is that if the system doesn't know consistently how long the CPU will take to process its part of the frame, it cannot set an appropriate delay. So in some circumstances this unpredictability will lead to a performance loss. Hope that clears this up for everyone