NVIDIA Reflex Low Latency Mode causing inconsistency/discrepancy in mouse movement?

[ruleset]

I was watching a video by Battle(non)sense titled "NVIDIA Reflex Low Latency - How It Works & Why You Want To Use It", I decided to dive into it since I noticed some of the best CS:GO pros have the mode set to "off" (m0nesy and s1mple, two of the best aggressive AWP players) and there is this logic in my head that if NVIDIA has it off by default why would we turn it on? I always felt that this setting makes the mouse movement smooth or too reactive, might be placebo, but it's just what I feel. Why would a gaming company want you to find some youtuber guy with a "fortnite performance boost fps fix 100% epic nvidia settings MAXIMUM FPS" in order to have the optimal gaming performance? You know?

I stumbled upon this comment posted by @codcouch1 in the video:



I think consistency is way more important than low-latency. While I agree with your methodology and you guys did a great job with Reflex Mode, you overlooked consistency of mouse movements. When I turn on Reflex Mode, I can notice a huge difference when I try to flick to a target. My flicks never go where they're supposed to. They always end up slightly too short or too far. I played for a couple of months with reflex mode on before I realized this inconsistency was happening. Once I turned Reflex Mode off, I instantly felt the difference. My flicks were way more consistent. Now if I miss, it's because I messed up or the enemy suddenly strafed the other direction, not because my cursor ended up in a different spot than it should have. This is ultimately the problem with Reflex Mode and why you shouldn't use it. I would suggest putting the mouse on a track or something and measuring mouse movement versus character rotation. I predict that you will find a huge discrepancy when Reflex Mode is enabled. Here you would get a much more important metric than just latency. Character rotation is way more important than low-latency clicks. Not knocking your findings so far, you did a great job, keep it up.

 

Now obviously this is a random comment on the internet, but can someone good with things like these do some testing? I always hear about discrepancy on this site so I thought it's worth posting.

 

 

[DPI Wizard]

38 minutes ago, ruleset said:

I stumbled upon this comment posted by @codcouch1 in the video:

Do you know what game he is referring to?

I measure mouse movement all the time, and have not found any issues with this lately. But I did have some issues with it a few years ago when it was new, perhaps the same issues this comment is referring to which now seems mostly fixed from what I can tell.

[ruleset]

1 minute ago, DPI Wizard said:

Do you know what game he is referring to?

I measure mouse movement all the time, and have not found any issues with this lately. But I did have some issues with it a few years ago when it was new, perhaps the same issues this comment is referring to which now seems mostly fixed from what I can tell.

Unfortunately I don't. I understand some games have mouse smoothing etc so that could also be his issue.

If it's fixed than great!

[RedX]

Low latency mode can make the CPU a bottleneck, so if your computer is having heavy CPU , it might effect performance in some situations.

Having poor FPS can effect aim. 

8k on a 1k game can cause lag 

 

Testing

This seems like something that could easily be tested with Sensitivity Matcher, or program a teensy to act as a mouse. Like configure Sensitivity Matcher to do a 360 in CS:GO, and other games. And then test with with Reflex Mode off, on, low latency. And see if it makes any difference. 

Not sure how to test this in a safe way in csgo, sense they have anti cheat. And it might see this as cheating. 

 

FPS
Some games are effected by FPS. q3a bridge jump, MH jump required 125 FPS. Apex Legnds tricks like Supper Jumps, seem to work better at 180 FPS. Early versions of q3a used object collection which was effected by FPS. They later switched to a server side math based approach.

Some games might not consider what's happening between frames. So having a bunch of frames, even if you can't see them, might be useful in some circumstances. 

I can't find the reference to it, but I seem to remember overwatch making a big deal about how their game considered mouse position and clicks between frames.

[Vaccaria]

On 3/28/2023 at 7:50 PM, ruleset said:

I was watching a video by Battle(non)sense titled "NVIDIA Reflex Low Latency - How It Works & Why You Want To Use It", I decided to dive into it since I noticed some of the best CS:GO pros have the mode set to "off" (m0nesy and s1mple, two of the best aggressive AWP players) and there is this logic in my head that if NVIDIA has it off by default why would we turn it on? I always felt that this setting makes the mouse movement smooth or too reactive, might be placebo, but it's just what I feel. Why would a gaming company want you to find some youtuber guy with a "fortnite performance boost fps fix 100% epic nvidia settings MAXIMUM FPS" in order to have the optimal gaming performance? You know?

I stumbled upon this comment posted by @codcouch1 in the video:



I think consistency is way more important than low-latency. While I agree with your methodology and you guys did a great job with Reflex Mode, you overlooked consistency of mouse movements. When I turn on Reflex Mode, I can notice a huge difference when I try to flick to a target. My flicks never go where they're supposed to. They always end up slightly too short or too far. I played for a couple of months with reflex mode on before I realized this inconsistency was happening. Once I turned Reflex Mode off, I instantly felt the difference. My flicks were way more consistent. Now if I miss, it's because I messed up or the enemy suddenly strafed the other direction, not because my cursor ended up in a different spot than it should have. This is ultimately the problem with Reflex Mode and why you shouldn't use it. I would suggest putting the mouse on a track or something and measuring mouse movement versus character rotation. I predict that you will find a huge discrepancy when Reflex Mode is enabled. Here you would get a much more important metric than just latency. Character rotation is way more important than low-latency clicks. Not knocking your findings so far, you did a great job, keep it up.

 

Now obviously this is a random comment on the internet, but can someone good with things like these do some testing? I always hear about discrepancy on this site so I thought it's worth posting.

 

 

You ask a technical question about technology, but at the same time you equate this technology with some players. You have a contradiction here.
What were these technologies created for?
How do these technologies work?
You have shared your experience (feeling), but what next? You decided to refer to a ready-made answer, but how did that answer come to be with another person? 

NVIDIA Low latency - it works with the buffer for rendering frames, for example in the game OW there are dots after FPS, this is the load of the buffer, if you turn on Low Latency, there will be 1 point, if you turn off, the buffer may fill more than 1 point, which will cause a queue for rendering frames, that is waiting, which means already delayed for rendering, not for mouse input. (The video shows this process)
REFLEX is a different technology, the video explains how it works.

But how is the mouse involved? How does the mouse work? Where does the video card output the "frames"? Does that mean there is a monitor? What is Ghosting, Overdrive, Overshoot, Black Smearing and why Response Time is not Input Lag?
Let's not open a can of worms...

Logic is the science of the laws and forms of the correct construction of thoughts. Two have been formulated so far: "formal" and "dialectical".
Mathematics(a tool created by humans) is formal logic. 
The world around us is dialectical logic.
For some reason you decided to point to logic in your reasoning, but there is a contradiction in your reasoning, formal logic is not concerned with resolving contradictions. And you have not engaged in the resolution of contradictions, referring to "why should we include? "We" is to whom? You asked a question, and therefore you have to answer it, to look for an answer, otherwise you will not learn how to think, and questions will continue to appear and remain unanswered.

A person who can conduct tests(A man who can, is a man with an understanding of what he is looking for...), will not waste time on tests, in his head are already theory and practice. Since the request in the tests does not refer to theory, the request refers to the feeling of the person. And a person is not a constant creature, today he may have a stomachache, and tomorrow his headache.

Also, I do not deny your observations or the observations of another person, but your question and your request are not correctly formulated, and that is how the world works, because you know how to design a rocket, and I do not know, but I know something else, so for you the same language is differently understood. 
For example. I don't understand your example of players playing games on the professional stage. Do they have a technical background? Did their parents or society teach them how to think? They are people just like anyone else.