Anti-Lag Thread

Radeon Anti Lag

This is not about smooth game play. Nor is it about monitor refresh rates.

This is about reducing input latency which can, depending on the game, help foster game play and immersion that was intended by the developer.

No, it didn't change flip queue size in the registry. I checked the registry on that one.

Requirements:
It does work in DX11

If you up the IQ settings a bit you should be set. Or increase in game resolution scaling options. Or use a low level Super Sample AA you should still see a benefit of using RAL.


Best way to enable it so far is through the profile of the game, not the global profile.






We Need Cliff Notes here:
@5:50 Anti Lag
@6:21 YOU MUST BE GPU LIMITED Not necessary (by HiAlgoBoost)
@6:59 Esports Situations: If you enable more IQ settings were you strain the GPU or if the GPU is already strain...this is where Anti Lag really kicks in.
@7:46 Frames are Metered. It reduces any queued requests from the CPU to GPU.
@9:46 Explains how it's done. Hmm, there is no OSD way to show it at this time CTRL-ATL-B (Or any other button you've assigned to enable Anti Lag) will bring up a green OSD showing you the frames you need to have in order to get that level of latency without Anti-Lag. (by HiAlgoBoost)


Project Cars 2
In single player once I enter a race where the weather was going to change from sunny to rain I would lose control of the car in the corners. The car would do a complete 180 with no control of the sliding.
I was never consistent with hitting my braking points. Steering into the apex was never a good option (dry tracks).

With Anti-Lag and this driver:
Now I hit all my braking points with absolutely no issues now.
I can plant my steering and hit the apex every time now.
I can be the very last person out of the garage to qualify and the AI no longer cause issues with controlling the car (odd issue discussed on their forum).
Change of weather from sunny to rain doesn't make the car feel like it's in a ice rank (another issue discussed on their forums but never fixed).
I can play several races with AI now.


Racing in the rain was nearly impossible before. Now it's FUN with Anti-LAG. Again, I use to have major problems racing in this game. Specially in the rain.

I am speechless as this is the 1st game where I noticed such an obvious difference.
The frame rates are about the same in replay mode. And the frame rates appear to be similar while racing. However, the inputs using the gamepad are simply 100% better.

It's hard to tell if frame rates are lower. It looks the same to me.


Injustice 2
I've not log as many hours in this game as I've had in PC2 but I do notice how crisp my button mashing is when I play this game. I'm better able to hit more combos consistently now without the frustration of wondering if I need a new gamepad. Apparently, I don't need a new one at all now.




Blackops 4
This game...this game...oh well, if I can get a few good matches in without cheaters then I'm really having a blast dominating the match. One match that I've joined in progress allowed me to carry the team to a win! I've was never able to be such a factor before. I seem to get the jump on people by being able to juke around fire my weapon with greater pin point accuracy. I now set in game mouse sensitive to around 3 and put my mouse's dpi to around 1500-1700 range. I couldn't do that before and now I have more mobility then before.


This is another match I've played.



Edit:
In game settings are still the same.
However, to make the game more GPU dependent I enabled Super Sample AA and disabled Post AA.
1080p @ 120 Hertz all other settings is the same.
I've tried FlipqueueSize 1 before and it didn't make the difference I noticed with this driver 19.7.1
with Anti-Lag turned on.

 

What is your screen refresh rate? 60/120/144Hz?

 

I've only tried measuring latency so-far (high speed camera + LED clicker) on my 144hz displays, in Battlefield and SOTTR. The impact was within the margin of error / statistically not significant compared to flip-queue size of 1, yet there may very well be games that benefit {e.g, ones I've not tried}.

None the less, having the option as a toggle in Radeon Settings while ingame and on a game-by-game basis is still pretty great. I also can't comment if this works differently on newer hardware, having only tried my Vega FE.

 

watch dogs 2
horrible -20% fps
rainbow six no difference
assasins creed odyssey -5%fps and effect just like 1 prerender

 

Keep in mind you want to be more gpu dependent when using AL

 

Did you just enable anti-lag in global, or did you make a game profile for PC2 and do it that way?

 

It's a toggle through an hotkey.

 

With ALT-R (in game) you can enable it and for me it stays enabled after (after reboot etc).

 

I gotta laugh here.
I wasn't understanding the prior questions. But apparently, Nvidia chimed in and thinks AMD is using flip queue size of 1 or 0 as Anti Lag.
So now you got some parroting the same from Nvidia. This is about the dumbest thing I've ever read.

And yes, I checked the registry settings and they haven't changed. It's the same.

 

Quick question...no DX12+anti lag, right?

 

No DX12 at this time.

And to further update I'm still getting repeatable, good, results from Project Cars 2. I'm simply amazed that it took this long to solve the nagging problems with PC2. Yeah, I still liked this game regardless but to see the issues I had fix is simply mind blowing. You have to put yourself in my shoes by having to deal with the issues I've experienced and simply worked around for months. If you play PC2 for the 1st time with this driver there is no way you will be able to relate to the improvements I experienced.

 

Oh, my apologies then.

Yes, by leaps and bounds it does...in Project Cars 2.

BO4...still testing...but so far I'm getting the jump on other players when I run and gun w/sub machine weapons now. Even when I use a slower firing weapon like the Cordite when someone is using the Spitefire. It seems that I'm noticing the effect of this anti lag most when I run and gun. If I use other weapons that require a more tactical approach it's very subtle. I can snap on target better without the over aim/under aim I had before. But it's very subtle.

Battlefield V... I do snap on target better though just like in BO4 with Assault Rifles. On thing I've notice is that I am more able to switch from healing a down player to shooting an enemy player who is standing 10 feet a lot quicker now (and down him). Before, I simply didn't have any chance. I thought that was how the game was played.

So far it seems to me that you need to be "doing something" in order for me to see this Anti lag in game. If you are walking around stealthy and using a tactical approach I'm not finding much. This seem to coincide with what I've found in PC2. I have to "always be in action/motion" to see the best results.

However, I'm still testing this. I've been doing this for a few days now and so far this is my impression. This is doing much, much more then what flip queue size did on it's own.

What I have yet to do is test the difference between flipqueuesize at 1 vs 3 with Anti Lag. That's going to take some time though.
From initial tests with PC2 I'm not seeing any difference.

Now I want to say that I've been playing these games for months. So I know what I was experiencing before these drivers came out. If you test them I suggest that you do so with games you are very familiar with. A game that you know you've had some sort of issue with.

 

I think NVidia is just referring to that AMD had not (untill now) had an option within their Radeon Settings for users to make a choice (with preference to rendering latency or performance). It's getting a bit taken out of context by several reviewers already though, as we have no idea how this actually functions (under the hood).


Assumptions do us no good, honestly what we need at this point is to run trials across various games comparing the latency impact of flip-queue set at 1 [via registry] vs AMD's anti-lag feature. To take this beyond anecdotal claims and observations to something data backed. [not doubting your case with Project Cars btw]

Perhaps somebody (a hardware reviewer) with ownership of a compendium of demanding, GPU-choked games (namely shooters), can turn this into a great article. It'd make a nice addition to just raw benchmarks that the 5700's are getting at most sites. (*hint hint*)

^ perhaps they can also run a comparison against NVidia at a flip-queue of 1 to see if this is a selling point to go team red {as a competitive gamer}.

 

How do you know it works? Are there any motion to photon latency tests FlipQueueSize 0x3500 vs 0x3300 vs 0x3100 vs 0x3000 with Radeon Anti-Lag on and off? I saw some bullshit "Driver lag tests" and they're not what we're interested in as gamers, we're interested in motion to photon latency.

I asked it in the other thread, but I didn't think it would actually change registry settings. Registry settings are just flags for drivers, drivers can use them, or they can ignore them. What I meant, is if RAL settings will have exactly the same effect as setting FlipQueueSize to 0x3100, and driver would ignore FlipQueueSize setting in registry and use RAL On setting instead(which would effectively mean 0 or 1 prerendered frames ahead by CPU).
So the fact that RAL setting doesn't change registry doesn't mean driver doesn't use that setting to set amount of prerendered frames to 1.
And I'm not saying that it is.
But if it's a real deal(and it doesn't just change amount of prerendered frames), there should be difference in motion to photon latency, not just "feel in the game", but you should be able to capture it with latency tester.

 

at some of these responses.

Anyway,

I tried HItman 2 and it I'm not noticing much difference between keyboard/mouse and controller. They still react the same. But I wasn't expecting much as you really don't do much anyway when compared to a FPS game.


I found another video about Anti Lag. And no it's not fast sync.


Skip to 22:55
AMD Drivers prevent the CPU from being ahead of the GPU. It will delay the work the CPU has to do so the CPU and GPU align with one another.

 

Update,

I've did some more test between FQS 1 and 3 in BO4. I'm not noticing any difference that would give me an edge. Nor am I noticing anything that would indicate to me that I should re-enable FQS 1 over 3. Run and Gun remains the best way to notice the difference between drivers IMO.
1080p
Render Resolution was increased from 75 to 90.

I'll increase resolution from 1080p (competitive play for me) to something higher. I was at a lower res. then that before this driver. Will report back.


Edit
Ok, I upped resolution to 1440p and kept RR 90 and was shocked to see that I was still getting over 100 fps. That certain wasn't the case before I updated drivers (I skipped a few drivers updates so I think I had May 2019 or so). So far at default FQS I don't notice anything that would indicate that I need to get it to FQS of 1. I had the highest KD on my team after a few matches.

When I get more time I will may test FQS of 1.

 

We need to test Tekken 7 on Joypad
IMhO Anti-Lag will shine there.....

 

This isn't showing a whole lot but I wanted to provide a video none the less. For me, I'm noticing a pretty big difference. Wish there was a way to demonstrate it better.
This is at 1440p @ 120Hertz same settings as before.
https://dai.ly/x7d88aj


This is PC 2 at 1080p 120Hz, Super Sampling. I tested to see if I got FPS loss with AL on. I didn't notice any difference.
https://dai.ly/x7d890f