Home Forums Game Integration ISE can do better along camera pan boundaries

Viewing 3 posts - 1 through 3 (of 3 total)
  • Author
  • #4560
    James Harford

    Hi. There appears to be a boundary condition that Infinite Screen Extension (ISE) is not exploiting that could measurably improve the user experience in games like Elite Dangerous.

    The following is a bit wordy, but I do want to be clear and precise.

    I hope its fair to say that ISE for Elite Dangerous (ED) responds to the (x,y) coordinates of a user’s gaze at some point of interest (POI) on the screen by panning the camera such that the POI moves directly towards the center of the screen at a speed that is the product of a “sensitivity” function, f(x,y), and a maximum speed, S. This function is 0 at the center whose coordinates we define to be (0,0) and 1 at its maximum at points on and near the screen boundaries which we define as x=1, x=-1, y=1, y=-1. The exact form of f(x,y) and the value of S are, of course, determined by the ISE sensitivity settings sliders for ED.

    So far, so good.

    However, games such as Elite dangerous limit the rotation of the camera view to simulate the limited range of head movements of the pilot.

    If I pan the camera view to, say, the rightmost limit of my range, a common occurrence in gameplay, it follows that anywhere I look on the right half of the screen can only move the camera up or down.

    The problem, if I may call it that, is that f(x,y) is no longer correct because it is designed for the case in which every possible POI on the screen can move towards (0,0), an impossibility in this case for any point to the right of the Y axis. What it does do is move a POI towards the X-axis at a speed that is fastest for those that are closest to the right edge of the screen. The result is that these points have a small or non-existent deadzone, and therefore often overshoot the X axis or oscillate about it indefinitely. This loss of control is particularly irksome when the boundary lies near the zenith or nadir, where the oscillations are accompanied by disorienting twists of the camera view. So the user will probably change their ISE settings towards lower performance, even though their settings are perfectly fine except when at a panning boundary. Fixing this could significantly increase the number of high performance settings available to players of Elite Dangerous.

    The fix, if I may call it that, is as follows. When the camera is at the rightmost panning limit, replace the sensitivity function f(x,y) by f(0,y) for all POI to the right of the Y axis. Thus all points in this region have the same deadzone centered on the X-axis. The vertical panning sensitivity of the screen is then the same for all points to the right of the Y-axis, and if the gaze should shift to the left of the Y-axis, function f(x,y) once again applies and panning proceeds normally. Of course each panning limit requires an analogous fix.

    Do let me know if you are thinking about this.



    I have the same issue with DCS World if I understand OP correctly. When looking outside the boundary of the camera limit I’m getting weird camera behavior that can be described as oscillations. This is very troublesome and is for now solved by toggling eyeX support on and off. It’s worst when looking down at either side of an cockpit (trying to get to switches and such).


    Added to the list of suggestions. Thank you!

Viewing 3 posts - 1 through 3 (of 3 total)
  • You must be logged in to reply to this topic.