Surely then that measurement from your eyes becomes the hypotenuse rather than the measurement to the center being the adjacent, so the maths would not alter... the opposite side (from center of screen to bottom (or top) would be constant.... just my 2 penneth.
Well in truth you would need to solve for two triangles. One formed above the eye position and the other inverted below it. Divide up the vertical segments of monitor height and alter the hypotenuse's based on that.
True, but we were talking 'center of the screen', or at least I was. I understand the math if eyelevel becomes very relevant, but when we are talking center of the screen, it is largely irrelevant as eyelevel is not used as a metric.
Poor Nigel, but I found that his comment, in regards to 'the feel of the simulator', is quite useful.
I can understand where Nigel is coming from, especially since the FOV used look more in line with an arcade as the view is higher than what he must have been used to. Also, the feel is totally wrong because there is no feedback for G-forces, among other things. Now if he had stepped into a properly adjusted motion sim, he might have had a slightly different opinion. Not to mention that I am sure he didn't race with all those people standing around watching him beside his cockpit hehe.
This hole fov calculator thing is so &/#%¤ that it couldnt be more wrong. It is based on imagining a window frame to the 3d world. but you dont understand what a 3d world in games is. You think that its the same as holdind the frame closer or further to the viewers face (arms lenght when holding the frame) when looking to real world. If you would be looking to real world these calculations would be right. But 3d world in games is basically a sphere that has the image projected to the inside walls of the sphere and you are in center of the sphere holding the frame. This sphere is needed so that you can pan your 2 dimentional view freely around. The sphere causes what is called a fishbowl effect. Meaning that objects in center of the screen appear further away than they would in real life. Now when changing fov in the 3d world in games you are NOT changing the arms lenght but moving away or towards the center of the sphere while arms lenght stays the same. Q: so what is the correct fov in 3d world then? A: there is no scientifically correct fov, but from gamer perspective there is. Q: so what is the correct fov from gamers perspective then? A: fov that makes objects distances appear correct to your brain Q: how do i find that correct fov then? A: with rFactor2 it is relatively easy: enable extra track markers in game, park your car, say next to 200m marker, then adjust fov so that gap between 150m marker and 100m looks convincing 50m to your brain. then set your seat distance to mach how much you see the cockpit/car so that it is in right relation to surroundings. Q: i find that it looks convincing to my brain on several fovs, is there any other way to find the correct fov? A: if the feeling of speed feels too slow on straights and too fast on corners your fov is too small and vise versa. When object distances and the feeling of speed feels convincing to your brain in all situations youre there.
I have to agree with you. With 3 screens I tried for months to drive at the "correct calculated" fov of 27 deg. Then started experimenting and finally ended up at 42 deg. I can see the corners properly now and have a proper sense of speed.
I'm sorry, but you're wrong. There is no fishbowl effect except when the game FOV is higher than your own FOV. An object 5 metres wide which is 50 metres away from you will occupy 5.72° of your view (horizontally). The same object 100 metres away from you will occupy 2.86° of your view (horizontally). The same object 25 metres away from you will occupy 11.42° of your view (horizontally). I'm talking real life here. Those are absolute values; an object that size, at that distance, looks that size. If you match your game FOV to your screen FOV (in your view, as per the various calculators) then an in-game object of that size, at that distance, will occupy the correct portion of your view. That's the only setting that will produce correct perspective. Your relatively small window (even with 3 screens) will reduce your visibility and your sense of speed; AND, a higher FOV lets you see more and feels faster. And because we're all so used to seeing high-FOV images (every time you watch TV, for example) the more realistic one feels slower. I can't argue with you there. Lack of visibility or a reduced sense of speed does NOT mean the calculation is wrong.
Sorry but youre the one whos wrong. Every 3d world in games is looked from inside the fish bowl. i explained why that is. That's why really high (120 or so) fovs in any game starts to look distorted because you have reversed behind the center of the sphere/fishbowl. With field of view below 30 the gap between the markers seems like 25m or so, that should be hint enough that something is wrong, not to mention the slow straight/fast corner speed sensation. Try fov of 90, and youl see that the opposite is happening. speed in straights feels way too fast when corners feel too slow (too long) and object distances feel higher than what is natural.
My question is: If the FOV is correct for my monitor per the "Calculator" is the depth of field necessarily correct? Changing FOV in rF2 changes the focal length or depth of field. The "Calculator" gives a different FOV depending on viewing distance. Does depth of field change depending on viewing distance? I have not noticed this when I move away from the monitor. Given a 40 inch and 20 inch monitor both set to a FOV of 45 deg will they have different depths of field when viewed from the same distance? I have not noticed this when I got a larger monitor. At the risk of appearing stupid (neither physicist nor optical guru) I can't see where the calculator accounts for depth of field, but only calculates how much of the image to crop. FOV is an area of consternation for me is I have never found a satisfactory FOV whether by the "Calculator" or by trial and error. Here is the link to an IR article on the subject, http://www.silverball-magic.com/my-stuff/Field_of_View_Explained.pdf. SW P.S. I remember for GPL there were two utilities that claimed to change FOV, but I think one actually changed focal length and the other only FOV (ie just changed the way the picture was cropped.)
Lazza is absolute right, if we want to simulate our eye field of view using the proper focal lenght. Of course the human eye is not a lens with hard cuts on the edges, but the area where the vision is sharp and still inside the stereopsis, could be compared to a 50mm lens for the standard 35mm camera format. This means you can get the same result looking inside a viewfinder of a 35mm reflex when using a 50mm lens (or focal lenght when using a zoom lens). What you'll see inside the viewfinder will be the same (or very very similar) to what you'll see without the camera. You could, in theory, walk and make other activities, just watching through that viewfinder as your brain is still getting the same focal lenght (same perspective, same parallax, same distances ratios etc...). Something you can't do using different focals, without starting wobbling or getting false informations about distances. So, what we've to do, when searching for a proper FOV - is to simulate, as much as possible, the human eye focal lenght/field of view to give our brain a similar perspective ratio as when we're seeing the real world. The sense of speed has nothing to do with the realism aspect. This is what happens when you're faking your real field of view ;
Oh you people. Even the fact that allmost everyone adds a little to your "correct calculated fov" to make it feel comfortable, or that people suffer motion sickness when the fov feels too low doesnt ring any bells??? Well i have given my 2 cents about the issue, who will take my advise and who will not doesnt really matter to me. The way i see it, is that now the info is out there, i've done my part. Cheers!
No, the objects drawn near the side of the screen are bigger because of the high FOV. The virtual screen in 3d games is flat, just like your screen. The way objects project onto it is reliant on how far the eye is from the virtual screen; how does the game know what distance this is? The FOV you set. Let's look at an example of high FOV and why the fishbowl effect happens. If I have my rFactor2 FOV set to 90° on my 16:9 screen, the horizontal FOV will be 121.3°. In a practical example, if my screen was 60cm away from me and filling this much of my vision, it would need to be 120cm high and 213.4cm wide - a 96" screen. Just picture that for a second - sitting 60cm away from a 96" screen, with the top 60cm above your eyes and the bottom 60cm below. The sides are 107cm to your left and right. Now, imagine you want to represent a real life situation where you have an object directly in front of you, and a same-size object at the same distance from you but 60° to your left/right. Obviously in real life those 2 objects are going to look the same size. They're the same physical size and the same distance from you. So here you are sitting in front of your giant screen. You draw the object in front of you, at the right size, all good. Now you want to draw the 2nd object right near the edge of your screen so it looks the same size. Here's the problem: The screen is 60cm away from you directly in front of you. The left and right edges of the screen, because they're 107cm away from the centre, are 122cm away from you. This means you have to draw the 2nd object bigger, in order for it to look the same size from your viewing point. And if you do that on your giant screen and sit back in the middle, guess what? Both objects look the same size. Now step back a few metres, or draw that same image on a smaller screen, and suddenly it goes fishbowl. Nothing to do with spheres or anything else... just a matter of perspective. *For those metrically challenged: My hypothetical situation is sitting 2 feet away from a 96" screen; that's 4 feet high and 7 feet wide.
Yeah, I understand this! I can drive something around 5 to 10% faster leaving the cockpit cam as the sense of speed and the vision quality is going to be FAR higher due the absence of the static element inside the vision (the cockpit) which is going to be connected (perception) with the real world static element (monitors). In real life the sense of speed is everywhere; you car is shaking, rolling, vibrating, moving all the time. It is communicating with you. In a static simulator this is not going to happen, for obvious reasons and what you're doing here is just compensating the lack of sense of speed we have in real life, thanks to the 3D world, the stereopsis, and the real car feedback, pushing on the FOV. This doesn't mean the FOV theory is wrong, just mean you need to fake this aspect to drive better. Nothing wrong with that, but the realistic FOV still remains a good point if we want to simulate the reality. If you were a real racing driver, the faked FOV could be a total fail if your goal was to train a real life track, as everything you're seeing in your monitors is faked in terms of field of view. This way you're not going to learn brake points, reference objects positions in relationship with corners etc... It's exactly the same as the binocular soccer...
Laws of perspective have been discovered in Italy by Filippo Brunelleschi circa 1413. Time to change, maybe a new theory, a new "general relativity theory" for geometric representation of 3D space in two dimension is due.
"just mean you need to fake this aspect to drive better". Yes! You have to "fake" it (or make it really) so it gives the brain everything it needs to make you fast. My point exactly! thats why there is no need to calculate the fov. just make it right (convincing) for you/your brain.
I'm not buying that that's all you were trying to say. You tried to pseudo-scientifically prove that a calculated FOV is wrong because it "feels" wrong. That's fine, nobody can argue how something feels to you or anyone else. However, your original post actually said "there is no scientifically correct fov" which couldn't be further from the truth, and shows how little you know about the subject. My argument against this: math--learn it, it's a wonderful thing--you really should know it before you try to persuade people that something isn't scientifically possible, lazza already did a great job explaining the math part. Not to mention all that B.S. about projecting onto a fishbowl. Games actually use planar projection, which is VERY much different from curvilinear projection, which is what you're suggesting they use with all that talk of fishbowls. All of that crap you said about fishbowls is horribly inaccurate. IMO, the reason that most people bump up their FOV a little bit from what's calculated is because using a relatively typical situation of a single monitor approx. 23" in diagonal sitting .5-1m away from their face yields an extremely low FOV which is nearly impossible to see your surroundings out of. Try driving around the real world using only 20 degrees of your vertical field of view; you'll be bumping into everything, stopping short, speeding, etc. Not to mention, it's all the stuff in your periphery that actually gives you the sense of speed. Without a triple screen setup (at least), you lose nearly all of your visual cues, so people like to add a little bit back in in order to gain back that sense of speed---nothing to do with correct geometry. Just because something is difficult or uncomfortable, which is all you're really saying, doesn't make it inaccurate, which is what you were trying to convince people of.
Depth of field doesn't matter in the case of rF2 at the moment (and probably forever) because it isn't simulated at all. DoF refers to the blurring that occurs because of the way lens optics works; i.e. it only matters if something is in/out of focus. Everything in rF2 is rendered with perfect focus. Also, because the game uses planar projections and does not account for lens distortions whatsoever, changing the FOV is exactly the same as cropping the image.
