LabRat 891Correct me if I'm wrong but,
couldn't malware that's intercepting, inspecting, and altering TCP/IP packets
or
bad hardware that's 'inconsistently' dropping packets
also 'trigger' an actual VAC Ban?

I'm asking/mentioning as I recall devices for the Xbox 360 that did similar, and were a PITA for msft to detect and ban.
IIRC, the eventually-found un-bannable 'trick' was using a physical switch on one of the Ethernet conductors/pairs, to simulate (on-demand) an 'ISP/Network problem' while not-entirely disconnecting the 'hacker/exploiter'.
If done quick-enough, an exploiter could 'freeze' everyone else in the game (they hosted),
frag them, then re-enable proper LAN/WAN function.

I wonder if VAC has a detection mechanism for this kind of dumb hardware-level TCP/IP exploit?

I don't believe so, but I can't say for sure. Potentially, a VAC ban could be issued if Overwatch (player feedback against demos from reported players) found the player guilty through inconsistent or impossible movement.

Valve exclusively using dedicated servers with a server authoritative model for ranked matches really helps mitigate that.

AsRockWOW, i see Valve being a dick once again what a surprise. Sure AMD cross the line so steam\valve bans people for some thing that a user might not of understood.

Sounds like a lawsuit.


All sounds like a good way for Valve to make a buck.

AMD is exclusively at fault here for their poor and unsafe software implementation, as well as their lack of communication with game developers.

1d10tNothing new for online players, considering these anti cheat already doing false detection even with obnoxious and hostile services like Riot's Vanguard.
Personally I once got a VAC ban because the ping was too slow, yeah you read that right. Considering I left CS:GO because aimbots were rampant, I'm kinda surprised they are starting to get serious about fighting cheaters again.

Sure. And we are supposed to believe you.

Dr. DroAMD is exclusively at fault here for their poor and unsafe software implementation, as well as their lack of communication with game developers.

People here suggested that they should have an SDK instead, yet if they did that they'd leave it all 100% up to the developers and then people would complain why more games don't include it.

I'd rather have as many features as possible on the driver side than have to wait on developer support, we all know how that goes, there will always be problems with shitty anticheat software and false positives.

R-T-BI mean if you want to make that argument then AMD needed to program this differently.

No, valve should change their crappy anti cheat, there are ways to detect cheaters without having to play this stupid game of cat and mouse to detect which problematic software runs locally which simply just doesn't work that well, valve games have all been plagued by hackers at all times and I am sure this one isn't going to be any different.

How can something like that pass the software control and quality ?

No one said that dll injection COULD be bad in competitive online shooters ? Seriously...

A big mistake, the issue is that it will require them to change the whole source code of the anti-lag and make it implementable by the developers rather than the driver enforcing it.

Vya DomusNo, valve should change their crappy anti cheat, there are ways to detect cheaters without having to play this stupid game of cat and mouse to detect which problematic software runs locally which simply just doesn't work that well, valve games have all been plagued by hackers at all times and I am sure this one isn't going to be any different.

I mean I agree but the reality is that ship sailed decades ago and that fight is lost. Thus this is on AMD for not reading the room. In the present market, you whitelist or get banned. Period.

Dr. DroI don't believe so, but I can't say for sure. Potentially, a VAC ban could be issued if Overwatch (player feedback against demos from reported players) found the player guilty through inconsistent or impossible movement.

Valve exclusively using dedicated servers with a server authoritative model for ranked matches really helps mitigate that.



AMD is exclusively at fault here for their poor and unsafe software implementation, as well as their lack of communication with game developers.

Point is it's not the users fault, why is the user getting banned just using new option(s) in drivers.

It's pretty sad that Valve just don't kick people for having the option on with a pop up or what ever error message why. Just out right banning people is bollocks.

AsRockPoint is it's not the users fault, why is the user getting banned just using new option(s) in drivers.

It's pretty sad that Valve just don't kick people for having the option on with a pop up or what ever error message why. Just out right banning people is bollocks.

Good point, they are tracking people getting flagged tho which is already more than what most devs would ever do.

AsRockPoint is it's not the users fault, why is the user getting banned just using new option(s) in drivers.

It's pretty sad that Valve just don't kick people for having the option on with a pop up or what ever error message why. Just out right banning people is bollocks.

Unfortunately, due to the myriad ways this could be exploited, Valve cannot be faulted for doing this and VAC is doing its job as intended. Any anti-cheat worth its salt would raise red flags everywhere if there's any live code injection taking place and actively altering game code.

Sure, I acknowledge that there is no bad faith from either party involved, but accountability for what software is executed on each client's computer is still exclusively pertaining to the account holder, which is why IMHO - anyone who used Anti-Lag+ with their AMD graphics card should get banned, with a reversal in place once Valve and AMD come to an agreement and an official implementation is vetted, done in a manner that both sanitary and safe. Live code injection is a huge no-no, and I fully expect Valve to tell AMD to go pound sand with this approach as it can and will result in crashes, security holes or game exploits over time. Sure it sucks to be unfairly banned, but since this is a competitive eSport, I do not think it's too unreasonable to ask players to educate themselves about the technical aspects of the software, and the title's viability as an eSport relies on an absolute zero tolerance towards any sort of unsanctioned modification. I can say that I personally trust CS2 more as a competitive title over such a decision.

The true problem here is the blatant technical incompetence from AMD developers implementing this feature in the shoddiest way possible (this is indefensible, they are taking every single shortcut they can, there is a reason Nvidia developed an SDK and documentation, as well as lots of engine- and driver-side maintenance to support Reflex), their horrible to non-existent communication... some brief, unhelpful tweet on Radeon account is all that they have done.

Vya DomusPeople here suggested that they should have an SDK instead, yet if they did that they'd leave it all 100% up to the developers and then people would complain why more games don't include it.

I'd rather have as many features as possible on the driver side than have to wait on developer support, we all know how that goes, there will always be problems with shitty anticheat software and false positives.

No, valve should change their crappy anti cheat, there are ways to detect cheaters without having to play this stupid game of cat and mouse to detect which problematic software runs locally which simply just doesn't work that well, valve games have all been plagued by hackers at all times and I am sure this one isn't going to be any different.

FSR wasn't ignored. Integrate them both in a single easy-to-deploy package, as RTX-enabled games often do. Reflex is actually toted as the antidote to the latency overhead caused by DLSS-G frame generation, and AL+ is similarly the companion to AFMF/FSR 3.0 frame generation.

Dr. DroAMD is exclusively at fault here for their poor and unsafe software implementation, as well as their lack of communication with game developers.

While I agree that AMD should have thoroughly tested before rolling this feature out, but I think they did this to try and improve user experience and with good intentions. So credit should be given where its due, instead of slamming them. It is a misstep, yes. But this does not negate the improvement from this feature.

watzupkenWhile I agree that AMD should have thoroughly tested before rolling this feature out, but I think they did this to try and improve user experience and with good intentions. So credit should be given where its due, instead of slamming them. It is a misstep, yes. But this does not negate the improvement from this feature.

I never questioned their intentions, I questioned their aptitude, or lack thereof. As it stands, this feature is a nuisance at best and a threat at worst; and they've made no effort whatsoever to mitigate any impact beyond some very hushed acknowledgement on their X account. This driver should be pulled simply because of the risk of getting people VAC banned in the world's largest game.

Dr. DroHigh ping has never ever resulted in a VAC ban. Match abandonment in ranked games can result in game bans, however. These are not VAC bans but bans issued directly by the developer for unacceptable conduct.

In some games, said bans will display alongside VAC bans on a Steam profile, but the profile will remain in good standing with VAC, and game bans apply on a per-title basis.

I'll give you benefit of the doubt blaming Logitech Gaming Software or Comodo Firewall, the only software I run in the background. That being said, I never leave a competitive match intentionally nor use any kind of illegal software, but that's what happened.
Don't need to argue anymore, I've created another Steam account and quit Counter Strike for good, there are many other games with more modern graphics, responsive movement and more challenging modes.

PapaTaipeiSure. And we are supposed to believe you.

You don't have to read or reply my comment though.

Tried to research this anti lag actually does, if I am right it removes intermediate queued CPU buffered frames that the GPU cant keep up with, so that the GPU gets to the most recent frame faster?

So basically anti lag is at its best when pushing GPU to max (uncapped FPS), kind of trying to resolve a problem gamers pushed on to themselves with this uncapped frame rate nonsense. :)

Just cap your games to 60 fps or anything significantly below what the GPU can handle, and be done with it. :)

chrcolukTried to research this anti lag actually does, if I am right it removes intermediate queued CPU buffered frames that the GPU cant keep up with, so that the GPU gets to the most recent frame faster?

So basically anti lag is at its best when pushing GPU to max (uncapped FPS), kind of trying to resolve a problem gamers pushed on to themselves with this uncapped frame rate nonsense. :)

Just cap your games to 60 fps or anything significantly below what the GPU can handle, and be done with it. :)

Just capping your fps to 60 is terrible advice, because it greatly increases system latency.

The optimum setting, accounting for the highest degree of frametime stability and maximal reduction of latency is to use just in time rendering (without a CPU frame queue) at 6.8% below target refresh rate and keeping GPU load below 95%.


As Radeon Antilag Plus offers no SDK, receives no hardware feedback or otherwise, and was implemented haphazardly, and since AMD has not publicly released any document that describes how it works, I just assume the driver injects its code in the game, starts reordering frames at the engine level while processing the entire input data stack on its own, introducing start of frame latency markers and this is what caught VAC's attention.

This is pertinent reading:

www.nvidia.com/en-us/geforce/news/reflex-latency-analyzer-360hz-g-sync-monitors/

developer.nvidia.com/blog/understanding-and-measuring-pc-latency/

chrcolukTried to research this anti lag actually does, if I am right it removes intermediate queued CPU buffered frames that the GPU cant keep up with, so that the GPU gets to the most recent frame faster?

Pretty much and as far as I remember there is a similar option for this in the Nvidia control panel as well, wonder if that gets you banned.

Vya DomusPretty much and as far as I remember there is a similar option for this in the Nvidia control panel as well, wonder if that gets you banned.

No, because Nvidia got game developers to support their implementation and gave them complete control over it.

Dr. DroJust capping your fps to 60 is terrible advice, because it greatly increases system latency. You also got AL confused with AL+, which is seems to be a ripoff of Nvidia's Reflex Low Latency mode.

Well I think thats down to whether you can perceive this extra latency, personally my brain isnt quick enough to notice 10-20ms latency saving or whatever it is, So I have never been convinced that things like 300fps gaming, nvidia reflex and so on are nothing more than pseudo benefits. Still find it funny that for that new 500fps monitor Nvidia had to slow down the footage 30x or whatever it was to show the advantage of it. :)

I apologise I am not convinced by this stuff, I am just a simple gamer playing games I enjoy, things like stutters bother me a lot, but not if I see something happen 10ms later than it happens.

But I did also say for those who believe they need these super high frame rates, just set it to something significantly below what the GPU can handle, so if its maxing out at 300fps, set it to say 200fps, you still have these super fast frames, but now your GPU actually has a buffer to handle stuff, depending how the cap is set it should also cap the CPU frame processing as well so its in sync with the GPU, simple solution that doesnt get you banned or cause other oddities. Ultimately the problem is a side effect of allowing all your components independently to process everything as fast as possible and the GPU running at maximum load.

Dr. DroNo

Lol how do you know that ? Common sense says a similar driver feature should also get you banned.

Vya DomusLol how do you know that ? Common sense says a similar driver feature should also get you banned.

Because Nvidia is not recklessly hot patching in-memory user mode code to achieve the desired result? Which is what we've been talking about for a while now.

I'm actually trying to understand more about the in depths of this but AMD as usual refers to it as a special algorithm and refused to elaborate further, haven't found any public documentation on it.

chrcolukWell I think thats down to whether you can perceive this extra latency, personally my brain isnt quick enough to notice 10-20ms latency saving or whatever it is

I actually agree with your practical take on it, that's why I consider myself too old for competitive gaming, someone is always quicker than you and now has faster hardware too :P

Dr. DroBecause Nvidia is not recklessly hot patching in-memory user mode code to achieve the desired result? Which is what we've been talking about for a while now.

No, I am pretty sure we're talking about different things here, Nvidia has an option in their driver to the eliminate queued frames to reduce input lag, my question is does that also get you banned.

Vya DomusNo, I am pretty sure we're talking about different things here, Nvidia has an option in their driver to the eliminate queued frames to reduce input lag, my question is does that also get you banned.

It doesn't. The thing that gets you banned is code injection (to place latency markers and prevent excess frames from being generated, perhaps?)



It behaves similarly to the render queue size control back in DX9 days.

Vya DomusNo, I am pretty sure we're talking about different things here, Nvidia has an option in their driver to the eliminate queued frames to reduce input lag, my question is does that also get you banned.

That feature culls frames after they've been rendered, thus it is incredibly dumb and mostly ineffective, but is encapsulated entirely in the graphics driver. Whereas Reflex/Anti-lag prevent unnecessary frames from ever being rendered - but in order to do this they need both knowledge of how the relevant game engine renders frames, and they need to hook into that engine to mark frames it creates with metadata. The driver then uses that metadata to determine whether to actually render them or not.

AssimilatorThat feature culls frames after they've been rendered

No, it drops them before they are rendered, I am pretty sure it even says that explicitly in the drop down when you select the maximum number of frames to be allowed in the queue, it would make no sense to drop frames after they've been rendered as that would achieve nothing. The point of the feature is to make sure that whatever frame ends up being rendered was the most recent one in the queue to reflect player input, that's fundamentally what all of these "latency reduction" features do and while I don't know how that one works I am pretty sure it must inject it self in the same way anti lag does.

Vya DomusNo, it drops them before they are rendered, I am pretty sure it even says that explicitly in the drop down when you select the maximum number of frames to be allowed in the queue, it would make no sense to drop frames after they've been rendered as that would achieve nothing. The point of the feature is to make sure that whatever frame ends up being rendered was the most recent one in the queue to reflect player input, that's fundamentally what all of these "latency reduction" features do and while I don't know how that one works I am pretty sure it must inject it self in the same way anti lag does.

Quite contrary. Since Nvidia provides an SDK, game developers are responsible for implementing it onto their engines and all major engines already have ready to ship Nvidia Reflex support. Implementing the plugin rests on the game developer's choice. No code injection takes place, as it's not necessary. AMD is taking a shortcut here and instead of documenting their features, writing a proper SDK, pitching in their work for game developers, doing proper support, they just chose the lazy way out.

Special K is capable of injecting Nvidia Reflex support onto any game in a similar manner that AMD's been doing with Antilag+, but even then, SK is to be considered unsafe as far as anti-cheat software goes, and has very strict injection requirements for it to work correctly, it must be loaded before the D3D context is created for its graphics toolkit to work.

Update is there:

www.amd.com/en/support/kb/release-notes/rn-rad-win-23-10-2

AMD has received reports of some games triggering anti-cheat bans on gamers when AMD Anti-Lag+ technology is enabled on Radeon graphics. To address this, we have released the AMD Software: Adrenalin Edition 23.10.2 driver that disables Anti-Lag+ technology in all supported games, and we recommend gamers use the new driver. AMD is actively working with game developers on a resolution to re-enable Anti-Lag+ and to help reinstate gamers who were impacted by the anti-cheat bans. We will provide more information when available.

store.steampowered.com/news/app/730/view/3717217413231124446?l=english