XIM Community

General Category => Game Support => Topic started by: full58 on 04:45 AM - 12/17/16

Title: coefficient variables, bf1
Post by: full58 on 04:45 AM - 12/17/16
 just the other day I came across some information I've never heard before. I don't know how valid the claim is but the guy seemed reasonably knowledgeable on the subject.

 thread here:

https://forums.battlefield.com/en-us/discussion/72159/psa-universal-soldier-aiming-youre-using-it-wrong

basically he's saying the universal soldier aiming (which I use)  is not actually set correctly for our 1080p  screens. the coefficient for a 4:3 aspect ratio screen should be 133% but we use 1080p with an aspect ratio of 16:9, so ours should be set to 178%

he goes more into detail about how exactly he comes up with the number. it makes sense. the coefficient for universal soldier aiming is only in effect for ads.

I just wanted to bring it up. anything to create a more perfect aiming situation. I just assumed the ads was trained at 133% and guessed that maybe training at 178% might improve the ads a little   
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: full58 on 04:51 AM - 12/17/16
the link didn't look right when I clicked on it. here's the body of the link

those not in the know/not already using this feature, first a little introduction:

Universal soldier aiming was a feature added to the control menu of (I think it was) BF4 the first time. It is aimed, no pun intended, at unifying the perceived sensitivity across different zoom levels. In earlier titles, or when this is not enabled, you'd often notice a big jump in sensitivity when you started using scopes and it was almost impossible to get the same muscle memory when switching from a 4x to an 8x scope, to 2x ironsights et cetera. You'd have to learn everything seperately and nothing would be cohesive.

Now, universal soldier aiming does some calculations based on your aspect ratio + FOV and changes your base sensitivity on the fly to give you the same feeling with whatever zoom level you are using. It can help you become more consistent and build up a crazy muscle memory.

But, there is a problem with universal soldier aiming.

The feature is built around a coefficient which you can set in the control menu, just below where you enable universal soldier aiming. That coefficient is 133% by default. Coefficient is not a very good term for it though, because it stands for your monitor's aspect ratio.

133% => 1.33 => 4:3

Unfortunately, for most people, this standard setting will not be right as there's not many people left playing on 4:3 screens.
That means the default coefficient is wrong! Most people are using either 16:9 (1920 x 1080 and similar) or 21:9 (2560 x 1080 and similar).

To correct this do the following:

16:9
16:9 => 1.777 => 178%

21:9
21:9 => 2.333 => 233%

Dependent on what aspect ratio you have, set the coefficient to either one of those, or calculate it yourself if you have a very special screen.

One more thing: if you have changed any individual advanced sensitivity setting, I'd recommend returning everyone of them back to 100% to make sure this doesn't throw off the system. Just mark down whatever you had set it to, so you can return to it if needed :wink:

My aspect ratio is 21:9 so I put it on 233%. It took me about 20 minutes of burn in time to get use to the new feeling - after that everything was feeling so snappy and accurate, I've never experienced that before.

It might be scary to change something in sensitivity when you're already reasonably used to your settings but trust me, it's worth it! I really don't know why DICE never bothered to explain this feature correctly.

Greetings,
Foodie
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: ViolaMB on 06:34 AM - 12/17/16
Any captains here to approve or disapprove this theory?
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: KraftyMattKraft on 07:17 PM - 12/17/16
The logic here is sound, and the math "adds" up.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: Breezer on 12:39 AM - 12/18/16
Please have a look at this thread : https://forums.battlefield.com/en-us/discussion/73116/uniform-soldier-aiming-on-or-off/p2 (https://forums.battlefield.com/en-us/discussion/73116/uniform-soldier-aiming-on-or-off/p2)

The coefficient should be kept at 1.33 if you are playing on a 16:9 screen.

The reasoning is well explained in the thread I shared by soneone who apparently worked on uniform soldier aiming.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: TSARGA on 02:53 AM - 12/18/16
There's already a thread about that, but I can't find it.

We came down to the conclusion that the USA factor is a constant multiplier.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: ForZaKRC on 04:12 AM - 12/18/16
Well, this actually helped me since the patch. Setting it to 178 (aim assist off).
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: KraftyMattKraft on 07:13 AM - 12/18/16
Please have a look at this thread : https://forums.battlefield.com/en-us/discussion/73116/uniform-soldier-aiming-on-or-off/p2 (https://forums.battlefield.com/en-us/discussion/73116/uniform-soldier-aiming-on-or-off/p2)

The coefficient should be kept at 1.33 if you are playing on a 16:9 screen.

The reasoning is well explained in the thread I shared by soneone who apparently worked on uniform soldier aiming.
What I read in that thread is that it is that the math is sound, but DICE kept the 133% because players generally aim with a box in the center of their monitor that is 4:3. This is widely over simplifying the thread, but the TL:DR answer is: "It's personal preference. MOST players prefer 133%, because they are snapping to target in an area of their screen equal to 4:3. While other players would still prefer the higher %. Play what's comfortable for your eyes."
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: full58 on 02:12 AM - 12/19/16
sluice. that was an excellent thread. it sure makes the case that USA should definitely be On.

i see where their coming from with it should be at 133 but they admit it's not perfect. after reading all that and the important guide to USA thread  it still makes me think a ST trained at 178% could benefit xim users if not non xim users. we do after all have the ability to adjust zoom and ads sensitivity separately and with the hip profile for ads if we so choose.   we can also change the xy ratio too.

there's a lot talked about in that thread and I was gonna try and mention some of it,  but I don't know what half of it is,  so I'm not. everyone else should read it though. it's a good thread.

i set mine to 178% (and 18 dead zone) it's not hurting anything for sure, and I'd say it felt better but I can't because of the Xbox one's recent look mechanic change. and for me at least,  the micro movement when ads is inconsistent.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: TSARGA on 05:04 AM - 12/19/16
Make some measurements with a constant input, with and without USA. Vary the coefficient. Vary the sights. Measure the rotational speed and the FOV, then check if there's proportionality.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: Breezer on 06:30 AM - 12/19/16
sluice. that was an excellent thread. it sure makes the case that USA should definitely be On.

i see where their coming from with it should be at 133 but they admit it's not perfect. after reading all that and the important guide to USA thread  it still makes me think a ST trained at 178% could benefit xim users if not non xim users. we do after all have the ability to adjust zoom and ads sensitivity separately and with the hip profile for ads if we so choose.   we can also change the xy ratio too.

there's a lot talked about in that thread and I was gonna try and mention some of it,  but I don't know what half of it is,  so I'm not. everyone else should read it though. it's a good thread.

i set mine to 178% (and 18 dead zone) it's not hurting anything for sure, and I'd say it felt better but I can't because of the Xbox one's recent look mechanic change. and for me at least,  the micro movement when ads is inconsistent.
Very important : All zoom sensitivity have to be at 100% for using the usa correctly.

Also, just don't roll with 178 because you think : 16/9 = 178% because that logic is flawed.

A quote from TheNoobPolice:
Quote
The coefficient isn't a setting of what aspect ratio you have, it's a mathematical modifier for mouse velocity synchronisation to a specific distance of screen space as a percentage of the vertical FOV.
...
TL:DR Some very smart people have created the USA forumla, the default of 4/3 (133%) is there for a reason and we were all aware that most people are playing on 16:9 screen. The only argument for setting the coefficient distance higher would be if you're playing on a 21:9 or wider screen setup.

But by all means roll with 178 if it feels right for you.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: KraftyMattKraft on 08:44 AM - 12/19/16
sluice. that was an excellent thread. it sure makes the case that USA should definitely be On.

i see where their coming from with it should be at 133 but they admit it's not perfect. after reading all that and the important guide to USA thread  it still makes me think a ST trained at 178% could benefit xim users if not non xim users. we do after all have the ability to adjust zoom and ads sensitivity separately and with the hip profile for ads if we so choose.   we can also change the xy ratio too.

there's a lot talked about in that thread and I was gonna try and mention some of it,  but I don't know what half of it is,  so I'm not. everyone else should read it though. it's a good thread.

i set mine to 178% (and 18 dead zone) it's not hurting anything for sure, and I'd say it felt better but I can't because of the Xbox one's recent look mechanic change. and for me at least,  the micro movement when ads is inconsistent.
Very important : All zoom sensitivity have to be at 100% for using the usa correctly.

Also, just don't roll with 178 because you think : 16/9 = 178% because that logic is flawed.

A quote from TheNoobPolice:
Quote
The coefficient isn't a setting of what aspect ratio you have, it's a mathematical modifier for mouse velocity synchronisation to a specific distance of screen space as a percentage of the vertical FOV.
...
TL:DR Some very smart people have created the USA forumla, the default of 4/3 (133%) is there for a reason and we were all aware that most people are playing on 16:9 screen. The only argument for setting the coefficient distance higher would be if you're playing on a 21:9 or wider screen setup.

But by all means roll with 178 if it feels right for you.
You are reading half the thread. IT IS based on a 16:9 resolution. 133% was chosen, because it matches CSGO and because CSGO uses 4:3 or 4/3 the lowest resolution you can play on. According the the thread that you are quoting, 16:9 or 16:9 was taken into consideration when creating UsA, but ultimately they decided on the default value of 133% due to the fact that players are generally not going to be snapping to targets on the very outside edges of their 16:9 monitors, but instead snapping to targets within a 4:3 fov or the center box of their monitors or tv's.

TheNoobPolice even mentions this is the article that you continue to quote here:

Quote
If you set the coefficient at 178%, then the mouse movement distance will be synchonised to a screen space distance which is right at the horizontal edge of a 16:9 monitor at ANY zoom level BUT aiming at less than that distance (e.g aiming to 2/3rds width of your screen) will actually be less uniform as you change through different scopes.

Since you are very rarely snapping to target right at the edge of your monitor, a 178% coefficient will likely make high zoom scopes feel too fast during the majority of your aiming relative to hipfire on a 16:9 screen.

Since most "snapping to target" happens in the centre 2/3rds of the screen, this is why the default of 4/3 was used. In extensive testing with multiple players, it made all different zoom levels from hipfire right up to 10x zoom etc feel the most uniform. The bottom line though, is that different peoples brains create the "average sensitivity sensation" differently and this is impossible to measure.

Right at the horizontal edge of the monitor is really the key sentence in that entire few paragraphs there. TL:DR version: 178% is the MOST CORRECT, but does not feel "right" at all zoom levels, because players play with 4:3 of their screen and not the full 16:9 of their screen.

One more thing. Let's go back to CSGO for a minute, while you claim it is a stale argument, it is the the reason why UsA is the way it is. Many players play on 4:3 stretched on their larger resolution monitors. Why? Several reasons. 1. The myth that the hitbox is slightly bigger. 2. More importantly, mouse input is uniform across zoom levels.
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: full58 on 02:33 PM - 12/19/16
that's what I read into it too krafty: yes the sensitivity may feel higher using the higher zoom optics such as x8 or x10, but it was "most correct"

but again that's why I think it could benefit xim users over non xim users.  we can adjust ads separate from the hip sensitivity, globally thru the xim or individually with the in game zoom adjustments. I know for sure we all plan to sink in at least a few hundred more hours into bf1 right?  I'm definitely gonna hit at least 800 hours with it.  and tweaking the xim4 to perfection seems to be a thing amongst its users

and i do keep all my settings at 100 for the various zoom levels.  except for the horse. for some reason it just feels right at ~65 for me
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: TSARGA on 06:52 AM - 12/23/16
the coefficient for a 4:3 aspect ratio screen should be 133% but we use 1080p with an aspect ratio of 16:9, so ours should be set to 178%
Prove it.

The coefficient should be kept at 1.33 if you are playing on a 16:9 screen.
Prove it.

Any captains here to approve or disapprove this theory?
I guess BF1 would behave similarly to this:
https://s30.postimg.org/wunyfoo0x/bf4_2.png
We are looking for the point where normalised omega/fov is 1:1:1:1:etc. As you can see, it's somewhere between 100% and 300%. Might be 133%. Might be 178%. Might be anything else. Of course, we are obliged to assume it exists.

We just need to take the same measurements with the right values to see who's wrong.
Title: Re: coefficient variables, bf1
Post by: TSARGA on 04:20 AM - 01/23/17
(https://s29.postimg.org/3z90eiu6f/bf4usa2.png)
*corrected the FOVs
Title: Re: is the coefficient taken into consideration when training the ST for bf1?
Post by: full58 on 12:16 AM - 01/24/17
the coefficient for a 4:3 aspect ratio screen should be 133% but we use 1080p with an aspect ratio of 16:9, so ours should be set to 178%
Prove it.

The coefficient should be kept at 1.33 if you are playing on a 16:9 screen.
Prove it.

Any captains here to approve or disapprove this theory?
I guess BF1 would behave similarly to this:
https://s30.postimg.org/wunyfoo0x/bf4_2.png
We are looking for the point where normalised omega/fov is 1:1:1:1:etc. As you can see, it's somewhere between 100% and 300%. Might be 133%. Might be 178%. Might be anything else. Of course, we are obliged to assume it exists.

We just need to take the same measurements with the right values to see who's wrong.

I'm posting what someone else said. I'm asking for somone to prove it. your quote of me is extremely out of context
Title: Re: coefficient variables, bf1
Post by: grandgt on 04:32 PM - 01/26/17
So, what does everyone run with, 133 or 178?


Sent from my iPhone using Tapatalk
Title: Re: coefficient variables, bf1
Post by: TSARGA on 04:47 PM - 01/26/17
So, what does everyone run with, 133 or 178?
178 but all my measurements were taken in BF4.

I'm posting what someone else said. I'm asking for somone to prove it.
Best thing to do would be to prove it yourself and post the results lol
Title: Re: coefficient variables, bf1
Post by: grandgt on 04:48 PM - 01/26/17
Lol, will do.


Sent from my iPhone using Tapatalk
Title: Re: coefficient variables, bf1
Post by: Zombieguy on 04:59 PM - 01/26/17
I did a long quest on what exactly Coefficient does and there really is no straight answer. Even the answer from Dice was vague. You can find conflicting posts on what it does but again there is no real answer.

I call 133% the default and you adjust it to how you like it.

From what I can have experienced is if you have it at 133% you maintain the transition HIP sensitivity into ADS but after 1.3 seconds while in ADS the sensitivity will slow down and default to the ADS sensitivity. The higher the number the longer it stays. This is how I have experienced it and drove myself crazy at times figuring it out. You might experience it differently than I do.
Title: Re: coefficient variables, bf1
Post by: TSARGA on 06:39 PM - 01/26/17
I did a long quest on what exactly Coefficient does and there really is no straight answer. Even the answer from Dice was vague. You can find conflicting posts on what it does but again there is no real answer.
It makes the rotational speed proportional to the FOV.

How is there no real answer?

From what I can have experienced is if you have it at 133% you maintain the transition HIP sensitivity into ADS but after 1.3 seconds while in ADS the sensitivity will slow down and default to the ADS sensitivity. The higher the number the longer it stays. This is how I have experienced it and drove myself crazy at times figuring it out. You might experience it differently than I do.
What??
Title: Re: coefficient variables, bf1
Post by: manaus on 06:48 AM - 01/27/17



I call 133% the default and you adjust it to how you like it.

Exactly...with R6 is precise the same, ppl come with numbers, figures, tables...oooh come on!

Verstuurd vanaf mijn LG-D805 met Tapatalk

Title: Re: coefficient variables, bf1
Post by: TSARGA on 12:18 PM - 03/21/17
Look what I just found: http://forum.symthic.com/battlefield-4-technical-discussion/7836-cte-uniform-soldier-aiming-info/

Quote
When the coefficient C = 0, then the formula used to calculate a sight's sensitivity factor is: SensFactor = tan(ZoomFov/2)/tan(HipFov/2)
At this setting, the formula acts to make the "mouse velocity to linear screen center velocity ratio" constant. What that means is that if you could measure the velocity of the image where it passes the screen center, when moving the mouse at a constant speed, between two different FOVs, there would be no change.

When the coefficient is C > 0, then the formula is SensFactor = atan(tan(ZoomFov/2)*C) / atan(tan(HipFov/2)*C)
To understand where this formula comes from, the original idea was to use the formula ZoomFov/HipFov to synchronize the "mouse distance to half FOV angle turn" number to make it constant for all scopes. What that would do would be to make it so that it would take the same mouse movement to move something at the top edge center of your screen to your cross hairs, no matter what FOV you were at.
When this formula was converted to work on synchronizing the horizontal mouse distance to half FOV angle turn number, the following equation was obtained:
SensFactor = atan(tan(ZoomFov/2)*W/H) / atan(tan(HipFov/2)*W/H)
Where W is screen Width
And H is Screen Height
Which means the current formula replaces W/H with C.
So if C = W/H, the horisontal half FOV turn distance will be constant.

Of course, my measurements aren't extremely precise, but...

(https://s29.postimg.org/3z90eiu6f/bf4usa2.png)

tan(0.522/2)/tan(1.492/2) = 0.289 and I got 0.291
tan(1.381/2)/tan(1.492/2) = 0.894 and I got 0.896

That guy might be right, but he also says:

Quote
C = 4/3 (New default)
This is the best working value we have found so far. Incidentally, it's the same value used in counter-strike games!
C = 4/3 makes all the scopes feel the same speed, and should provide a good sensitivity relationship to maintain drag shooting muscle memory across sights too.

Then why is 178% closer to 1:1:1 than 133%? Why do we get the same results for c = 0 and not for c > 0??
Title: Re: coefficient variables, bf1
Post by: dave079 on 03:03 PM - 03/21/17
I play at 1920 x 1080, that means 178% right? I am not that good with such things...
Title: Re: coefficient variables, bf1
Post by: TSARGA on 01:04 AM - 03/22/17
I play at 1920 x 1080, that means 178% right? I am not that good with such things...

That's what my measurements say.