Gunsight imagery distorted

[some1]

Both gunsight reticles are distorted, like they are tilted with the collimator glass.

 

That's not how collimator works, the reticles should appear to the pilot as nice, regular circles not ellipses.

Edited June 25, 2016 by some1

[q800]

Yeah. True. I've noticed and pointed this issue in official thread about new Gazelle features. Obviously no one cared.

[Devrim]

A screenshot at least?

[some1]

 

Do you know where gunsight is, or should I point the issue with arrow?

Edited June 27, 2016 by some1

[RabidRider]

Yup, pointed this out as well, no answer so far.

[Mt5_Roie]

It's noted...trying to get images of how it looks like in a real Gazelle.

[q800]

So it's licencia poetica of yours? Reticle shape is not influenced by tilted transparent medium, as it's image is focused in infinity. Not on the medium. So you can observe reticle, target and surroundings at the same time. That's pretty basically how collimation works in practice. Another thing is specific desing of reticle. I don know how it looks like in Gazelle. What type of gunsight is this? I mean model name, type and such. Probably some standard model used not only in Gazelle.

[Schmidtfire]

+1. I have tried to find some pictures of the sight/reticles in the Gazelle, but no success :(

[Socket7]

So it's licencia poetica of yours? Reticle shape is not influenced by tilted transparent medium, as it's image is focused in infinity.

 

I've seen this before, though not on a HUD (and I don't play with HUD's in real life, so grain of salt here.).

 

Projecting an image onto a medium that is not parallel to the projector results in what is called the Keystone effect. This effect occurs no matter how you focus the projector.

 

https://en.wikipedia.org/wiki/Keystone_effect

 

Keystone is a well known issue, and is pretty easily corrected for by either distorting the image being projected, or by shifting the lenses around to align things. I did a lot of this kinda stuff with boardroom projectors.

 

As for the question "does the Gazelle having no keystone compensation for the hud pipper?" I've no idea if that is the case or not. It could very easily go either way, but my gut says they'd adjust for the keystone instead of leaving it. Guess we'll find out when polychop hears back from people who actually fly em for real.

[some1]

I don't know about any gunsight that wouldn't take this effect into account.

 

Those rings and grids are not there to look pretty, in real life they are used for aiming and ranging. They can't be distored or they would be useless.

[Flagrum]

Just a thought here: could this distortion be related to the overall FOV we are using?

 

That sight glass is quite angled ... 45 degrees or so? And it is very near to our virtual eye. Both factors that are pretty unique to the Gazelle.

 

If the distortion would be caused by that keystone effect, then it should be more pronounced due to the angle of the glas, shouldn't it`? Therefore I was thinking if the distortion has more to do with the difference of the distance of eye <-> top-of-glass and distance eye <-> bottom-of-glass.

[some1]

No, eye position does not matter, you can play with eye position and FOV all day, it won't change how the reticle is displayed.

 

Besides, Mi-8 also has very similar gunsight mount with angled glass, and it works fine there, no distortion.

[Flagrum]

No, eye position does not matter, you can play with eye position and FOV all day, it won't change how the reticle is displayed.

 

Besides, Mi-8 also has very similar gunsight mount with angled glass, and it works fine there, no distortion.

What I meant was, the whole screen is distorted as we look "into" the world through a small window: our monitor. If the monitor were an actual window, limiting our view, it would cover about .. dunno, maybe 60 degrees. But usually we use a wider FOV, like 90 - 120 degrees. So DCS has to squeeze a view covering 120 degrees into a 60 degree wide monitor - which causes distortions.

 

These distortions are more noticable the closer an object is to the (virtual) eye or rather the "window" we look through (i.e. top rim and bottom rim of the sight glass). And this has nothing to do with how the sight works, it is just about how 3D rendering in computers work. So, the precision of the reticle image is not really compromized - the whole picture, including the target you aim at, are equally distorted.

 

Well, that is at least my theory atm. :o)

 

edit:

the Mi-8 gun sight is further away from the eye point and monitor pane and therefore not suffering as much from this effect.

[some1]

Your theory does not make sense. As I said, you can play with eyepoint any way you like (using trackir or keyboard shortcuts), change the fov to extremes, it won't affect the hud image in any dcs aircraft, gazelle included.

 

Hud is focused into infinity, it doesn't appear "close" even if the collimator glass is right next to your face.

[Flagrum]

Depending on the position of the object in relation to the "window"/monitor pane and the FOV, everything, including collimated HUDs, get distorted.

[some1]

Ah, you mean this kind of distortion - yes. The crazier FOV, the more image distortion at the edges due to planar projection. However, that is irrelevant to this thread.

 

If you make a comparison screenshot from the same position, looking straight into the gunsight, Mi-8 gunsight will be ok (maybe minor fisheye effect), while Gazelle gunsight still will be heavily out of shape.

[Flagrum]

Again ... the Mi-8 gunsight is farther away and quite in the center of your view. The distortions are more pronounced on the edges of the monitor and the Gazelle's sight takes a lot of space in your view (as it is very close to your virtual eye) and thus, it's edges are therefore closer to the edges of the monitor pane ... and thus gets distorted more. And as the lower edge is even closer to you than the upper edge, the effect is even more pronounced - resulting in that egg shape of the circle.

 

Compare the size of the upper edge and the lower edge of the gun sight glass - the lower one is much wider. And if you now consider that the distances of the drawn circle must be equal (i.e. the distance of the 10:30 o'clock position of the pipper to the upper left hand corner of the glass vs. 4:30 position to the lower right corner of the glass), it is obvious that the pipper itself has to be distorted as much as the gunsight glass.

[some1]

Sigh. Move your head a little as I said in the previous post and you'll see that the eye position does not matter. The Gazelle sight is distorted the same way, despite being roughly the same angular size that Mi-8.

 

 

 

 

Compare the size of the upper edge and the lower edge of the gun sight glass - the lower one is much wider. And if you now consider that the distances of the drawn circle must be equal (i.e. the distance of the 10:30 o'clock position of the pipper to the upper left hand corner of the glass vs. 4:30 position to the lower right corner of the glass), it is obvious that the pipper itself has to be distorted as much as the gunsight glass.

 

That's exactly what shouldn't happen. The gunsight is projected straight and symmetrical despite the glass appearing wider at the bottom. Game rendering limitations have nothing to do with it.

 

 

 

http://www.ebay.ie/itm/201523447047

[Flagrum]

Pondering over this during the day, I slowly seem to get your point.

 

IF the reticle was projected onto the sight glass (assuming it were opaque / made of cardboard), THEN the distortion like I described would occur. Reason: the image of the reticle would have the same distance to the eye as the sight glass and due to the FOV issues with 3D rendering on FOV-limited 2D panes (monitor).

 

But we have a collimated sight and the reticle image is not at the same position as the sight glass. Instead it's distance to the eye is infinite. And as anything far away from the eye point is distorted less, the reticule image should also be distorted less/unnoticable/not-at-all.

 

Hrm. So, what's then wrong with the reticle? Maybe the origin of the projection is not angled exact perpendicular to the sight line of the pilot trough the sight glass?

[some1]

Yes, it's more like looking at a distant object through an angled mirror. It still has the correct proportions.

 

My guess is that the reticle in the Gazelle is correctly set in the infinity (technically it's a finite distance, but far away), but it is rendered at the same angle as the collimator glass, instead of being perpendicular to the pilot's eye.