Matching first and third person sensitivities

[Kilroy]

Is it not possible to match the sensitivities of first and third person games / game modes? They don't feel the same like how matching sens of different fovs through the viewspeed formula does. I might I have ideas but I wonder if there's a solution already.

[DPI Wizard]

It is possible, 3rd person have a FOV and a 360 distance just like 1st person. The issue is that camera distance and how far away objects are affects how you perceive the sensitivity. Since you can't control these two factors, the only sensible match is using the FOV and 360 distance.

[Kilroy]

Yeah, the camera distance affecting the perception of speed is what I'm talking about. Just like how there was an attempt to match sensitivities across fovs through monitor distance and viewspeed, I feel that there could also be a way to factor in the distance of the camera from the point the camera pivots around, which is usually the character.

I'm writing this from my phone so I'll see if I can make pictures when I get home. Let's say you're playing a third person game. Although third person games usually have the camera offset to the side, and also it gets closer to the character, the pivot point, whenever a wall gets in the way, I'm not going to worry about those things for now. Let's just say the camera is directly behind the pivot point and pointing in the same direction, forever keeping the same distance. In this game, the horizontal fov is 60. I figure out through the viewspeed formula that 60 hfov is 50 cm/360 and I put it in the game. However, the sensitivity now feels too slow compared to what I'm used to.

So my idea starts from an obvious place: that the sensitivity ought to be faster. This makes sense because the third person camera is situated "further down the line". In a first person game, move mouse = rotate camera. In this game, move mouse = move camera along the path of a circle. I realize it's futile to compare the two speeds since you can't compare rpm with m/s, but I believe the attempt will lead to better performance in game. Anyway, if it is further down the line, it ought to feel faster, since that point in the line normally moves at much higher speeds and it gets slower as the point moves towards the center. If I put the the sensitivity for 60 fov, this would be right for the point at, or infinitely close to the centre. Therefore, it ought to be faster as the camera moves off the center. So imagine both the camera and person were pointed the same way, I'll draw out the cones for their sightlines, which show the area that the vision of the camera and the character cover. My idea is I would measure the distance from pivot point to camera, make this the radius for a circle, situated at the pivot point. I would then draw 2 lines starting from the character, which intersect with the circle and the lines from the camera. I'll make these two lines my new fov, which is now 90 for example, and put it into the viewspeed formula. Since this fov is higher, the resulting sensitivity would be higher, which is what it ought to be. Even if the solution is arbitrary, it is most likely closer to ideal. Just don't know I can prove that or if there even is an ideal.

Edited September 16, 2020 by Kilroy

[DPI Wizard]

The camera distance is only half the issue though. There's a huge difference if you aim at something right in front of you or 200 meters away. In fact at distance the sensitivity is pretty much the same as in first person. So no matter how you do it you will only make the sensitivity match for objects at one specific distance.

[TheNoobPolice]

Most 3rd person games also use a dynamic camera position depending on where you character moves around the terrain of the map, e.g open area, camera moves back and broadens FOV, narrow area or indoors, camera moves in close to character etc.

Because of this, the only thing that makes sens really is to match either 360 distance just so your 180s are familiar, or focal length so your tracking is familiar. Anything else feels totally different anyway and becomes completely arbitrary