Viewspeed v2

[potato psoas]

1 hour ago, seventhfrost said:

Yeah, I agree on both points. I just wanted to clarify, since the mentality that the best solution should feel best most immediately is likely to just lead people to assume ones closer to what they're used to are better.

That aside, is there any way to try out the method you've been working on where the monitor match % changes with the FOV in whatever game? I saw you posted something on using it in csgo, but I'm interested in trying it in siege and maybe pubg along with this newest formula from Drimzi.

Yeah that is definitely a problem with testing. I actually put some graphs up a while back explaining that you should really test out which method is best at the higher FOVs because the 360 results differ so much. Whereas they all converge at lower FOVs, so can't really tell the difference there. Then the best way to test is people should check how the sensitivity feels at the center, the entire screen, and then compare both to the desktop to get a better idea of what feels off. Most people are testing around 90 FOV but it honestly will be too hard to tell this way. They need to be testing at like 170 FOV. It's very easy to tell what feels off and what feels acceptable.

Well the method is just using the monitor match formula but where you input monitor distance you have a separate formula that scales the monitor match according to your FOV. I thought the best way to scale was to use the unit circle (makes sense to me, though I doubt it matters) and therefore monitor match is just COS(FOV/2). I also added the ability to set a limit since using the COS(FOV/2) causes the extreme FOVs to feel too fast. So instead of scaling from 0% at 180 FOV to 100% at 0 FOV you can scale, for example, from 50% at 180 FOV to 100% at 0 FOV.

I might link an excel sheet so you can play around with the limits and stuff.

I can get pretty much the same results as Drimzi's formulas if I just set the right limit.

Which limit I like best though is undecided.

Edited December 22, 2017 by potato psoas

[seventhfrost]

Ah okay, that's pretty neat. I'll definitely try some of it out if/when there are links or if you find limits you think give it some edge over other solutions.

[potato psoas]

1 hour ago, seventhfrost said:

Ah okay, that's pretty neat. I'll definitely try some of it out if/when there are links or if you find limits you think give it some edge over other solutions.

Okay so I got the Excel sheet almost done, only problem is my FOV formulas aren't working. They aren't calculating properly for different aspect ratios

 

I guess you can pull the FOVs from the calculator (needs more decimals) so I'll just post it while I figure out how to fix them.

Edited December 22, 2017 by potato psoas

[seventhfrost]

I've been using this experimental monitor match formula and I'm really digging it still. The sensitivities for it are certainly fast but they make sense. I have no complaints about it at high/low fovs personally, but I'm not too used to testing it outside of a smallish range of fovs. 

I'm pretty sold on it. Definitely took some time to adjust to how fast it was, but it fits the desktop better than the rest to me so far. Probably need more opinions though.

Also for some reason CSGO doesn't like me trying to use high dpi with a low wps value. Ended up having to go back to 6 so I could try things out there.

[Cocyx Skeleton]

At 250 DPI 6/11, CSGO feels very obviously faster than the windows mouse speed to me using the experimental monitor match -- it might be because I play at a considerably lower sensitivity/dpi than most people, and therefore I am more sensitive to faster speeds (although the 2.5 viewspeed sensitivity was far too slow for me, and clearly slower than the desktop mouse cursor).

Personally, the best formula is still 100% Monitor Match - 1:1 Diagonal Desktop, 1:1 Diagonal FOV when it comes down to matching the windows mouse speed to me.

I'm going to tinker with the experimental formula some more though and see if I can train myself to get used to it.

Edited December 24, 2017 by Cocyx Skeleton

[iBerggman]

7 hours ago, Cocyx Skeleton said:

At 250 DPI 6/11, CSGO feels very obviously faster than the windows mouse speed to me using the experimental monitor match -- it might be because I play at a considerably lower sensitivity/dpi than most people, and therefore I am more sensitive to faster speeds (although the 2.5 viewspeed sensitivity was far too slow for me, and clearly slower than the desktop mouse cursor).

Personally, the best formula is still 100% Monitor Match - 1:1 Diagonal Desktop, 1:1 Diagonal FOV when it comes down to matching the windows mouse speed to me.

I'm going to tinker with the experimental formula some more though and see if I can train myself to get used to it.

If you didn't do it already, try lowering your dpi so your csgo 360 distance from the new formula matches the old one if you feel the higher sensitivity throws you off. I used 1200 - WPS 3 before and lowered it to 1050 to get a more similar in game sensitivity with the new formula and now I'm liking the new one a lot more than I did initially. 

[Cocyx Skeleton]

38 minutes ago, iBerggman said:

If you didn't do it already, try lowering your dpi so your csgo 360 distance from the new formula matches the old one if you feel the higher sensitivity throws you off. I used 1200 - WPS 3 before and lowered it to 1050 to get a more similar in game sensitivity with the new formula and now I'm liking the new one a lot more than I did initially. 

Yeah I lowered my dpi down to 200 (the lowest it can possibly go) and it feels much better atm, but it's slower than what I'm used to so I'm still adjusting to it.

[potato psoas]

So I was playing around with the monitor match percentage. MM(FOV) = COS(FOV/2) actually makes the different aspect ratios give similar 360 distances so I'm going to see if I can figure out a better formula for the monitor match percentage so that the whole monitor matching formula is aspect ratio independent, so we don't have to use VFOV or DFOV.

 

Our aim is for this equation (monitor match formula) to be true:

HCL((π HMM)/(ATAN(HMM TAN((π HFOV)/360)))) = VCL((π VMM)/(ATAN(VMM TAN((π VFOV)/360))))

 

Since CL (chord length) and FOV are known, the only unknown is MM (monitor match percentage).

Edited December 26, 2017 by potato psoas

[Drimzi]

Edited January 28, 2018 by Drimzi

[iBerggman]

2 hours ago, Drimzi said:

A diagonal monitor match formula for those who want to try without aspect ratio independence.

.....

Ooh, definitely liking the fov - fov scaling on this one after testing it quickly with the logitech script. The previous experimental monitor match felt like it was too slow at 30-50 and too fast at 100-120 fov, especially while trying to stay on target while strafing around a bot or when using one of those maps with the moving platform and static target. But with this one I can comfortably track and shoot moving bots and keep crosshair placement while strafing, even with the fov moving in and out between 30 and 120. I'll play some games tomorrow and try it some more but for now this one is looking really promising 

[Drimzi]

Edited January 28, 2018 by Drimzi

[Drimzi]

Edited January 28, 2018 by Drimzi

[fenriquez]

Testing out Full Dynamic Monitor Match formula, found the lower FOVs too slow so when ADSing from hipfire ingame did not feel natural to me. Personally I'm aiming better with the experimental formula you had earlier that matched desktop to 180 degree turns. I don't i know if its just me having to adapt between formulas but thats what finding right now. But i find the fov scaling with the experimental formula to me really good, it may feel a bit fast at first but once you get use to it, feels really natural to me. Thanks drimzi for all the great work!

Edited December 26, 2017 by fenriquez

[Drimzi]

Edited January 28, 2018 by Drimzi

[Drimzi]

Edited January 28, 2018 by Drimzi

[Drimzi]

Edited January 28, 2018 by Drimzi

[potato psoas]

This is actually just 0% monitor match - who knew!

 

Okay, so I have figured out how to make the scaling monitor match formula aspect ratio independent. @seventhfrost it's done now

Here is an Excel link for you to play around with:

https://1drv.ms/x/s!AgoMjDNmWWpe6DLsZ8sBaZmddWHI

You can tell it's aspect ratio independent because if you make the resolution width the same as the height for Hdeg 4:3 games the 360 distance won't change even though the FOV and aspect ratio is changing - it's truly adding and cropping FOV.

It has all the formulas you will probably need. All the highlighted areas are the things you need to customise/worry about. Keep in mind that the results are in cm.

The only way to make it aspect ratio independent was to use MM(FOV) = COS(FOV/2). I tried including a limit but it didn't work. I don't think this is a flaw, just the way circles work. If you imagine the distortion is corrected and we used 100% then there would be different 360 distances but it would still be right - hard to wrap your head around (or maybe I don't know what I'm talking about).

I know it is faster than what everyone is used to but it really does maintain the consistency with all points on the monitor, so it won't feel so slow at the center but it will feel faster at the edges as your FOV increases - which was the goal, because of distortion. At 90 FOV it's not too bad, very consistent.

Edited December 27, 2017 by potato psoas
This is just 0% monitor match

[potato psoas]

14 hours ago, Drimzi said:

This is using sin(fov/2) instead of potato psoas' idea of cos(fov/2), so that it becomes reversed for 100% towards 180 fov. 0% is stupid fast for high FOVs. All methods converge towards 0 FOV anyway. This will scale a bit differently than the standard diagonal monitor match, since it's only 100% at 180 FOV, and gradually gets faster as the FOV decreases, but still converges to the same 360 at low FOVs.

Scaling to 100% as you approach 180 FOV and scaling to 0% as you approach 0 FOV doesn't make any sense. The only reason high FOVs get stupidly fast is because they are stupidly distorted. The virtual window of useable screen space gets smaller and smaller as you approach 180 FOV. If you scale to 100% then you are making the screen space useable but it won't feel like it is converted correctly.

It's why I think we shouldn't use them at all - maximum FOV should be 120 or even less. For the next release there should be a warning in the calculator telling you not to use them haha.

Edited December 27, 2017 by potato psoas

[potato psoas]

8 hours ago, Drimzi said:

...Actually, using the same logic as above, but shrinking it down to the original 1:1 box:

 

It's fast, but preserves the scaling behaviour that I have already tested and loved, and has the same fov/360 distance combo used in the matching warp stabilizer test from a few pages back, and actually feels really good at 100 vertical fov in Shoot the beat!. It's going to be stupid fast though, I need to test McOsu with the firstperson mod to get a better perspective of the 2D speed, since the whole playfield is going to be moving rather than just my cursor.

This one is even faster than my one

[potato psoas]

1 hour ago, potato psoas said:

Okay, so I have figured out how to make the scaling monitor match formula aspect ratio independent. @seventhfrost it's done now

Here is an Excel link for you to play around with:

https://1drv.ms/x/s!AgoMjDNmWWpe6DLsZ8sBaZmddWHI

You can tell it's aspect ratio independent because if you make the resolution width the same as the height for Hdeg 4:3 games the 360 distance won't change even though the FOV and aspect ratio is changing - it's truly adding and cropping FOV.

It has all the formulas you will probably need. All the highlighted areas are the things you need to customise/worry about. Keep in mind that the results are in cm.

The only way to make it aspect ratio independent was to use MM(FOV) = COS(FOV/2). I tried including a limit but it didn't work. I don't think this is a flaw, just the way circles work. If you imagine the distortion is corrected and we used 100% then there would be different 360 distances but it would still be right - hard to wrap your head around (or maybe I don't know what I'm talking about).

I know it is faster than what everyone is used to but it really does maintain the consistency with all points on the monitor, so it won't feel so slow at the center but it will feel faster at the edges as your FOV increases - which was the goal, because of distortion. At 90 FOV it's not too bad, very consistent.

Lol this is just 0% monitor match... but the interesting thing is that 0% is not aspect ratio independent, so this could still prove useful.

Edited December 27, 2017 by potato psoas

[Drimzi]

Edited January 28, 2018 by Drimzi

[ScottyFPS]

3 hours ago, Drimzi said:

At 400 DPI, 1920x1080, a 2:1 ratio is pretty close to 800 eDPI in CSGO, which is the most recommended starting point that most people can adjust to straight away, probably because it is almost 2:1 to their desktop.

I think a 2:1 ratio would be more useful than a 1:1 ratio.  Personally, matching desktop speed on a 1:1 ratio to 3D based games feels too fast to aim with accurately.

[Drimzi]

Edited January 28, 2018 by Drimzi

[Cocyx Skeleton]

I actually quite like the 1:1 formula with a lower DPI, personally. I recommend people just lower their DPI and use the 1:1 ratio.

[Drimzi]

Edited January 28, 2018 by Drimzi