The Witcher 3 - (23.9.2) 7900 XTX, Anti-Lag+ vs. Anti-Lag vs. Off

Sdílet
Vložit

Komentáře • 13

  • @galandilvogler8577
    @galandilvogler8577 Před 8 měsíci

    Out of curiosity, did you disable both Boost and Super Resolution (2/3 of HYPR-RX) in Adrenalin, so that the only thing you're activating in the video is Anti-Lag Normal/+?

    • @AMD718
      @AMD718  Před 8 měsíci +1

      I don't use hyper-rx. So, no boost or rsr enabled. Only Anti-Lag and Anti-Lag+ ... and RIS at 50% in this case. I almost always use RIS unless the game has excellent in-game sharpening, which isn't very common in my experience.

    • @galandilvogler8577
      @galandilvogler8577 Před 8 měsíci

      @@AMD718 Good to know, I think I'll upgrade to 23.9.2. soon and test Anti-Lag+ on Bugfield, just to see how it works there, given how slow the engine is on that game even on my XTX at 5120x1440. Thanks for the quick demonstration though, it's sad that Anti-Lag+ is still going very under the radar so far, it's the thing I missed the most compared to nVidia Reflex, I hate upscalers in general, so I look forward to use FSR3 NAA + AntiLag+ as soon as both will be available.

    • @AMD718
      @AMD718  Před 8 měsíci

      @@galandilvogler8577 I'm on the same page with you. A lower frame rate is much more tolerable when the latency is minimized. Conversely, 120 fps with high latency will feel soft and rubberbandy. So far Anti-Lag+ seems a decent match for reflex and I'm also looking forward to trying fg with FSR3, at least out of pure curiosity. Let me know how al+ works in starfield when you try it. I don't own the game or I would test it myself.

    • @galandilvogler8577
      @galandilvogler8577 Před 8 měsíci

      @@AMD718 Ok, tested quickly in Akila city, where the avg framerate is around 65 (5120x1440, everything maxed out except shadows at High instead of Ultra, no Motion Blur/DoF, no VSR, no FSR, only CAS for sharpening).
      Without any AL: 52-53ms
      With standard AL: 43-44ms
      With AL+: 33-34ms
      Results seems pretty similar with yours on TW3, except the AL+ which is around 10ms higher (probably because SF is much more CPU intensive than TW3, even though I'm on a 7950X3D, with the game using a CPU Set on cores 0-15 - vcache die - via Process Lasso).

    • @AMD718
      @AMD718  Před 8 měsíci

      @@galandilvogler8577 Results do sound similar in pattern to what I have seen now in TW3, Ghostwire Tokyo, and (most significantly to date) in Tomb Raider SOTR. Regular Anti-Lag is usually around half-way between no anti-lag and Anti-lag+. Of course, this assumes you're GPU-limited as when I'm not GPU-limited, latency is the same between all three - that is to say very low. For why your SF AL+ was a little higher at 33-34 I'm guessing your base frame time was a little higher (frame rate below 90 fps)?

  • @user-ub5ys5dy4w
    @user-ub5ys5dy4w Před 8 měsíci

    how we get 50ms with 100fps ? 1000ms(1s) \ 100fps = 10ms is 1fps

    • @AMD718
      @AMD718  Před 8 měsíci +3

      You're thinking of frame time which is a straight forward calculation. The ms value in this case is latency which is the total sum of all latencies and frame time which occurs in-between the mouse click and the result on the screen.
      BTW - frame time that you are thinking of is in upper right.

    • @user-ub5ys5dy4w
      @user-ub5ys5dy4w Před 8 měsíci

      @@AMD718
      It’s still strange that AMD at the driver level can directly influence other components, significantly improving their performance. I would believe in a small value, but not x2

  • @br4k_ob4ma
    @br4k_ob4ma Před 8 měsíci

    Nice ms drop.

    • @AMD718
      @AMD718  Před 8 měsíci +1

      Yes I've been fairly impressed with Anti-Lag+ so far. AMD just needs to roll it out to more titles (which they have been doing) and (hopefully) make it work with non-rdna3 hardware.

    • @br4k_ob4ma
      @br4k_ob4ma Před 8 měsíci +1

      I'm with rx 6800 hope so. Let's wait for FSR 3.@@AMD718