Setting up your rFactor FOV - Tutorial

strange question.
If you see amount of objects in unnatural proportions, it always be looking strange. on 14", 32" or 100". Because brain compares the same object to how those would look in real world.

But on a huge screen (and I mean really really huge, especially for that GTL shot with its extreme FOV) you are much closer to the centre of the screen than to the edges. Your own perspective cancels out that of the image.

You won't actually see this effect unless you try it physically. So it's not a strange question at all.
 
yes. you will see no big deformations. but small deformations will be still existing, It;s just perspective: lines crossed in vanishing point. You cannot change it ;)

Edit:
Especially on our examples where vanishing point is near main object of the scene. Worse would be on cockpit view where you can see road. angles between edges of road are depending on perspective always. and you cannot cheat it.
 
You might think my 'cropped' example is just hiding the effect you're describing. It's actually showing that a high-FOV image contains the low-FOV image inside it.

If you set up your FOV to 28° and have it looking nice on your single screen, then change to a screen 10x bigger, but adjust your FOV accordingly, you will still see exactly what you did before in the centre of your new huge screen. The extra screen area just lets you see more around you.

Step back from that huge screen, so that you can see it all in front of you like your single screen, and it will appear distorted in exactly the same way the high-FOV images do... because you will have increased the FOV in setting up your new screen!
 
yes. you will see no big deformations. but small deformations will be still existing, It;s just perspective: lines crossed in vanishing point. You cannot change it ;)

Edit:
Especially on our examples where vanishing point is near main object of the scene. Worse would be on cockpit view where you can see road. angles between edges of road are depending on perspective always. and you cannot cheat it.

We are perhaps going in circles here.

Increasing FOV is just the same as 'zooming out'. You're shrinking what you could see before (which is still in the centre of the screen) and bringing more objects into your field of view around it.

The relative position of the objects, and corners of those objects, that you could see originally have not changed. I think that's what you're referring to... the shape of things drawn on screen depends on their size, shape, and position relative to the camera.

So, for a given position in-game, there is only one 'perspective' that truly represents what you would see in real life in that same situation.

To see that correct perspective, you have to match the field of view of the rendered image to suit your screen; the bigger the apparent size of your screen, the larger your field of view would be (if it were a window, for example) and the higher the FOV needs to be set.

If you change the size of your screen, or move closer to it, the 'correct' FOV setting will change. That's the whole point of the first post in this thread! ;)
 
I'm no longer going to argue any of these points as I feel we are all making up terms and the mis-communication of the last 10 posts would show through to someone someday but not me. Instead I will post a quick FOV Deformation video I just made in GMOD showing the extremes of distortion at super high FOV and some scale comparison of 2 cars, a 4 foot cube crate and a 1 foot white cube.


When any tool is selected my H-FOV switches to my customized 32 degrees. When on the camera tool's default it is 90 Degrees.

Music is oddly fitting.
 
Lazza, objects in the center of the screen will always be more or less proper, because deformations (I call it "fish-eye like" because I didn't know how that type of distortion should be called) caused by high FOV are related to sides of the picture, not it's center.
When you generate a picture with high FOV and crop it, leaving only its center, then there won't be much difference to a picture with lower FOV.

With high FOV, objects located on the edges of the picture will aways have wrong proportions, no matter what big screen you use, because that distortions are created by high FOV and screen size won't change anything.

EDIT:
"Dome type" screen won't change anything, because it won't "straighten" all that distorted objects... you would need to squash the picture horizontally, but then objects located in the center of the picture will be narrower than should.
 
Last edited:
"Dome type" screen won't change anything, because it won't "straighten" all that distorted objects... you would need to squash the picture horizontally, but then objects located in the center of the picture will be narrower than should.

It would need to be completely custom rendered to match the hardware. Most likely squashing just the outside bits instead of stretching them and keeping the center intact. I see the Jdome in action and the FOV is close but I still think tweaking could help.

 
Lazza, objects in the center of the screen will always be more or less proper, because deformations (I call it "fish-eye like" because I didn't know how that type of distortion should be called) caused by high FOV are related to sides of the picture, not it's center.
When you generate a picture with high FOV and crop it, leaving only its center, then there won't be much difference to a picture with lower FOV.

With high FOV, objects located on the edges of the picture will aways have wrong proportions, no matter what big screen you use, because that distortions are created by high FOV and screen size won't change anything.

I'm not going to continue discussing this, because we're not moving anywhere. Suffice to say I don't agree with your last (bolded) statement :D ('from the user's point of view', I might add - obviously a high-FOV image is going to look distorted when viewing on a large screen from a large distance, just as when viewing on a small screen from a small distance...)

Edit: Ah heck, I'll give it another shot.

Let's say you're sitting outside with a nice scene in view. You set up a pane of glass about arm's length in front of you, and proceed to draw the objects you can see by tracing their outline on the glass. You'll end up with a rather basic looking picture you can take home, pull out later, put back in front of you and you'll see all the same objects with all the correct perspective - obviously, because you drew them from what you were seeing.

This is basically what a game is aiming to do as well, right? All good.

Now, you take that glass, lean it against a wall, and take about 10 steps back. How is it going to look? It will look like a high-FOV image, because the perspective doesn't make sense from that distance. The objects near the edges will appear stretched and warped, because when you traced them those parts of the pane were at a fairly large angle to you, but now they're not.

High-FOV images look distorted at a normal screen viewing distance and size, but with a very large screen you actually need to turn your eyes/head to the side to see it - creating perspective. The far edge of the glass is smaller in your view than the height of the glass in front of you - it's just perspective. Objects at that edge will appear foreshortened, so while they are stretched if you stand in front of that part of the image, from your central position they will look correct.

Again, all this is assuming rFactor renders based on a flat plane for the virtual viewport. It would seem to make sense, with probably 99% of players using a flat screen.
 
Last edited:
OK, I will ask only one more question:

Why do you think, that high-FOV, distorted image diplayed on a big screen, won't be distorted anymore just because you are sitting close to it? :)



@ZeosPantera
Yup, it must be preprocessed to be properly displayed on a dome-like screen.
 
Sorry, didn't expect you to come on in the meantime, and I typed a bit longer than I planned lol... response above.

I remember seeing something done for half-life I think it was, for a projected dome style thing... and it did need a custom dll to warp the image to fit the projection and user viewpoint.
 
OK, so from what distance you recommend watching paintings to have proper perspective? :) I guess, that would be about 1m (distance between a painter and painting when he was painting it). So, when I watch a painting from a distance, it should look warped, distorted etc. ...but it, obviously, is not :) Why? ;-) That's a question for you, to figure that out :) Then, you should know, why you was wrong with your FOV to screen size relation.
Another tip for you. Pictures you took with your camera. Do they look different, when you watching them from a longer or shorter distance? :)

Let's say you're sitting outside with a nice scene in view. You set up a pane of glass about arm's length in front of you, and proceed to draw the objects you can see by tracing their outline on the glass. You'll end up with a rather basic looking picture you can take home, pull out later, put back in front of you and you'll see all the same objects with all the correct perspective - obviously, because you drew them from what you were seeing.

This is basically what a game is aiming to do as well, right? All good.

Now, you take that glass, lean it against a wall, and take about 10 steps back. How is it going to look? It will look like a high-FOV image, because the perspective doesn't make sense from that distance. The objects near the edges will appear stretched and warped, because when you traced them those parts of the pane were at a fairly large angle to you, but now they're not.
No, it won't be stretched, distorted etc., no matter from what point you will be looking at it. Why? Because the world you are looking at, starts at the point where the picture is, not where you are. With bigger screen (but still you want to keep scale of things on the picture), and you taking a few steps back, you don't touch FOV at all. You only move your virtual seat that few steps back.

Taking that into rFactor... let's say that you have screen big enough to have all objects with realistic size with you beeing 1m from the screen. It's like you would sit in a car 1m from a dashboard (on the screen you see the dash directly, think of it like having that pane of glass directly on the dashboard). Now, you move 2m back from the screen.
What you have to change in rfactor?
1. You say - FOV.
2. I say - your virtual seat position (you moved 2m so you should also move seat position 2m back).

But because in most cars, you see about 1m from a dashboard, you actually don't want to move back and out of a car, do you? :)
So if you want move back from whatever reason but still "be" 1m from the dash, then you should make the picture on the screen larger (to have that illusion that you indeed sit 1m from the dash). But FOV is unchanged all the time.
 
Last edited:
A painting is not done from the painter's perspective, as if he's looking 'through' the canvas and tracing objects. That's not a valid example.

If you actually did what I said, you would see what happens. I'm afraid you keep hanging on to a different issue to what I'm talking about, and not understanding what I'm pointing out.

Pictures from my camera are quite small. So is my screen. It is very difficult, if not impossible, to get a high-FOV image to make sense in either case because you can't get close enough and still see the image.

You need a big image to see the true result. I keep saying it, and you keep going to other issues. I will go away for a while, think about how best to illustrate what I'm saying, and come back :p
 
Maybe you want to ilustrate a situation, where you have curved picture that covers 180*, then flatten it and say "hey! objects on left and right are weird!"

Well, that's true... but:
1. You get that, because you have very, very wide screen that covers not only what you see in front of you, but also on both sides, which has been flattened.
2. It has nothing to do with perspective. What you have, is screen made with more than one sort-of POVs, that are merged together. It's like taking 3 photos of your room and merge them together (like you do panoramic pictures).

But to have that, you don't change FOV! You just need very wide screen. Human's horizontal to vertical "screen" proportions are definatelly not 16:9 or 16:10 but a lot more horizontally. That can be twice as much or even more.
That's the whole point of having 3 displays! To keep proper perspective and in the same time, see something more than only dashboard in front of you.
Of course you keep in mind that when adding more displays, you don't need to touch FOV at all!. FOV in rF is vertical, so if you added more horizontal displays, rF will recalculate new horizontal FOV, based on vertical.

You need a big image to see the true result.
Bigger image is just bigger image, nothing more. I saw already racing games (rFactor too) with high FOV on very big screens (2m high from a projector is enough) and they look exactly the same as on a tiny screen.
 
Last edited:
1. You get that, because you have very, very wide screen that covers not only what you see in front of you, but also on both sides, which has been flattened.

I understand what you're saying regarding 180° views (or close to it) because you can't project that onto a flat surface and have it make sense. Hopefully we've not been arguing that point, because we agree... however, when talking about a 'normal' screen, as wide as 16:9, and 'high' FOV up to, say, 80°, you don't get that close to 180° horizontal FOV. That particular example would be around 142°, which would be difficult to manage, but not impossible... and I think it's a little extreme for most people anyway (personally I have my vertical FOV set to 60° on my 16:9 screen).

To help illustrate the 'perspective' correction, and altering, that happens with screen size and/or distance to screen, I've put together a few crudely drawn diagrams. We have slightly hijacked the thread here, but hopefully these will actually help illustrate what ZeosPantera is hoping to achieve (and avoid) with the whole FOV calculation he's detailed.

First, to set the scene:

fov_setup.gif


What we have here is the viewer on the left, looking at a screen which the game will draw the in-game objects on; in this case I've kept things simple and have provided 4 in-game objects, all approximately equal distance from the viewer in the game. You can see the FOV formed by the distance from the viewer to the screen, and the size of the screen. In this case I've assumed the game has been set up correctly as per ZeosPantera's calculations, so the in-game FOV matches the true (real) FOV. (this could be either a top view or a side view - it doesn't really matter as the concept is the same)

Now, let's illustrate the way the visible objects are drawn on screen:

fov_1.gif


Here I've just traced from both sides of the visible objects back to the viewer, showing where the objects extents will appear on screen. It is quite obvious from this diagram that when the virtual and real FOVs match, the viewer will perceive the objects to be in the correct position and of the correct size - just as if they were real objects and sitting the same distance away in the real world. This is the whole objective of this thread.

Now, the reason games don't have this setting by default is that, as you can tell, the FOV isn't very big.. which limits how much stuff you can see in the game. In this case we can't see the red objects at all - in rFactor they might be cars that are actually slightly ahead of us, but we don't see them at all because the FOV is so limited.

I believe this was what Tim was referring to when he said he wouldn't go below 60° FOV - not because it's not realistic (perspective-wise) but because you severely limit your visibility when you cut down your angles that much on a normal screen.

So, like every other mainstream game, rFactor defaults to a larger FOV, allowing you to see more of the game-world... like so:

fov_2.gif


Ah, that's better. Now we can see all the objects... but wait. Look at where the object extents hit the screen, and the size of them from the viewer's position. The red objects are much wider than the green - the change in FOV has warped the image, giving the 'stretching' near the edges. It's an unavoidable side effect of viewing an image with a lower FOV than that which it captures - and it does happen with photos you have in front of you, and most TV you watch, which is perhaps part of the reason we don't really notice it so much when we're playing games. Mind you, some people do get nauseous when watching/playing games, and I think this is a large part of the reason.

So, this is a high-FOV image viewed on a small screen from a 'small' distance, or on a large screen viewed from a large distance (obviously you could scale this diagram to any size you want, the same effect will be apparent).

Finally, we ask: what happens if we keep the same screen size, and same image, but move closer to it?

fov_3.gif


Now, perhaps, it is clear why I keep saying you need a big image/screen. Getting a high-FOV image 'correct' with a normal size screen will not end well - you'll be so close you can't focus on the image at all, plus the fact you have 2 eyes an appreciable distance apart (in the context) will makes things even more confusing.

However, given a large enough screen to make getting this close comfortable, the perspective will make sense.

There is an important distinction here: if you look at the screen, the objects are drawn on it at different sizes, just like they were before. The red ones definitely use up more screen area, more pixels, than the green ones.

The key is that from the viewer's position, and perspective, they appear the same size.

One final point: I'll happily admit this is an extreme example. But given the limited scale of my poor little drawings it helps to show the difference. When you're viewing the screen, especially moving around nearby objects, even a fairly small discrepancy in FOV is quite apparent.

So, this is what I've been trying to explain without a lot of success. The stretching that occurs at the screen edges (well, across the whole image really, but it gets worse and much more obvious at the edges) can be offset by the foreshortening that occurs from the viewer's perspective.

You can see in the diagram above that the central part of the screen, where the green objects are drawn, is not at much of an angle to the viewer. Conversely, out near the edges the angle is quite acute - which will appear to shrink the stretched image at that point, so it comes out the correct size. (if green and red were the same size on the screen, the red ones would look smaller because of the screen angle)

I sincerely hope this all makes sense now ;)
 
Yes, I know that all the time :) Thanks for explanation. Should be useful for other guys, but... :)

I'm talking about that "fish-eye" perspective you get at high FOV. It won't be normal, no matter how close you put nose to your screen and how big the screen is! That's the whole point in our discussion. You keep talking about matching rF FOV with real world FOV and nothing more - and that's wrong. Because with high FOV and distorted perspective, you won't "straighten" those stretched objects when beeing closer to the screen. They will be unproportional (stretched horizontally) all the time and THAT is the issue with rFactor that prevents you to setup your POV with both FOVs matching each other.

In real life you see all objects in their proper dimensions (even though our horizontal FOV is about 108* or even slightly more). We are not aliens to see the world like that (that's of course very extreme FOV :D Picture taken from some other game. Could find anything good from Aliens games):

sro2008040908191579cc1.jpg


Will you agree with me? :)

EDIT: Of course, for most cases proportion between viewers POV and screen size is more or less constant and aims at lower FOV range. That means, most of people using ZeosPantera and your guide, will be setting up low FOV non the less.
 
Last edited:
Lazza, your examples looks fine, but only from 3rd person perspective. You forgot that 3d scene is projected on 2d screen turns into 2D which produces distortions
Once distorted, it cannot be corrected by moving closer to the screen. Brain is not able to correct it from 2D projection.

Moreover you just proved by those pictures, it is not true that the central part of screen has correct FOV but border parts are distorted. Everything is distorted except of one pixel at the centre of 2d screen. Any other objects are more or less distorted depending on distance from the scene centre.
 
Well Lazza you have done an excellent job of explaining how and why adjusting FOV is important to the sighting of objects and the distortion that occurs at high FOV. I'd like to combine your images with text labels explaining the situation in each for other threads if you don't mind.

Now Lesiu that screen you posted is VERY fish-eye. BUT I still think it would seem correct to you if you were looking at the screen with only one eye and were literally 1 inch from the screen or closer. Any further then that and you would see the terrible fisheye.

Maxym the image distortion is only present at high FOV, not all the time. If you set the FOV low enough all distortion will be removed. Just look at the Garrysmod demo I recorded using the round steering wheel to show distortion. http://www.youtube.com/watch?v=SecYuFVVrY4&#t=0m33s&hd=1
 
@ZeosPantera
No, it won't and (it looks like I have to say it again) no matter how big screen will be there and how close you will be towards the screen, you won't remove that distortions until you set lower FOV. That's the whole point of my discussion here.
 
Yes, I know that all the time :) Thanks for explanation. Should be useful for other guys, but... :)

I'm talking about that "fish-eye" perspective you get at high FOV. It won't be normal, no matter how close you put nose to your screen and how big the screen is!

Ok, I'll try to keep this brief. I've just said, in my last post, that 180° FOV or close to it won't work on a flat screen. Why do you keep pointing out things I've already said, or disagreeing with arguments I haven't stated, and then asking at the end if I agree?

Now you say you're only talking about these extreme examples... well here's a few things you've said (I beg you to go and view your own posts, because I haven't sourced about half of them here and have cut a lot of stuff out):

Here are two videos...

FOV 28:
FOV 60:

First thing to notice is, how perspective deformes items in screen's edges (for example, how ADL get stretched when I'm turning my head). You don't have that with low FOV.

Lazza, rF perspective has nothing to do with size of your display, and how far you sit in front of it.

So you tries to tell, that rFactor magically will change perspective when you move further back from the monitor, without touching rF settings? Please, be serious :)

I am calm :) I'm just trying to explain, that position in front of the screen has nothing to do with how you should set FOV in rFactor

As I said earlier, having even 150" and sitting 1m in front of it does not mean, that you have to set high FOV in rFactor, because you will have wrong perspective in rFactor

Everything, and I mean everything, you said above contradicts the post I just made - which you now say you already knew.

Lazza, your examples looks fine, but only from 3rd person perspective. You forgot that 3d scene is projected on 2d screen turns into 2D which produces distortions
Once distorted, it cannot be corrected by moving closer to the screen. Brain is not able to correct it from 2D projection.

Moreover you just proved by those pictures, it is not true that the central part of screen has correct FOV but border parts are distorted. Everything is distorted except of one pixel at the centre of 2d screen. Any other objects are more or less distorted depending on distance from the scene centre.

You've rather missed the point. Look at the top and bottom diagrams, where the 'true' FOV matches the virtual FOV - what the viewer sees on-screen is exactly what they'd see if there were no screen and the objects were real. Perspective cannot be warped or distorted - how can a real situation be wrong? Is the world warped when you look out of your window at the street? That's what we're simulating here... and there's no reason it will be incorrect. FOV is all that matters.

I'd like to combine your images with text labels explaining the situation in each for other threads if you don't mind.

You can do whatever you want with them :)

@ZeosPantera
No, it won't and (it looks like I have to say it again) no matter how big screen will be there and how close you will be towards the screen, you won't remove that distortions until you set lower FOV. That's the whole point of my discussion here.

LesiU... imagine a 2 metre high image. Now imagine putting your eye 2cm away from it, at the centre, and looking over towards the edge. Do you realise the sort of view you're going to have? Don't you think the extremely stretched image you've posted might actually be necessary in that scenario? Of course it's not practical... but that doesn't make it incorrect. (you can use a low FOV image in the same situation, and it's still impractical - and it'll have incorrect perspective)

180° is impossible to get right. It would require your eye to be on the screen's surface. Not a great position to see the image from. I've said "close to 180°" but I guess I'm really limiting myself to practical examples. You could view a 179° horizontal FOV image with no problems if you had a screen 500 metres wide and sat 2.18 metres away from it. It will look perfect from that position.
 
Lazza, in rFactor you define vertical FOV, not horizontal... I already told you but it looks like you have no idea about that... With 3 16:9 screens, with 28* vertical FOV, you get horizontal 150* FOV.... but that is WITH LOW VERTICAL FOV, THAT WON'T GENERATE DISTORTIONS! It is NOT the same with one 16:9 with high vertical FOV (83.5*, for which you get 150* horizontal in that case).


I also have to quote myself again... Look at those 2 screens:
https://community.racesimcentral.ne...actor-FOV-Tutorial?p=4617&viewfull=1#post4617

They both have 109* horizontal FOV (vertical 25* with screen proportions ~4.37:1). DO YOU SEE ANY DISTORTION?!?! There is no damn distortion at all, because of low vertical FOV! Why? Because the proportions are wide enough to provide natural horizontal FOV!
 
Last edited:
Did you really?

Well, it's true I was focused on what LesiU was saying, and passed by your objections quite quickly. Still, there's not really a lot more I can add without simply repeating myself.

It seems to me you have a problem with a 2D image being able to portray a 3D scene correctly. I would refer you to my example about drawing a scene in front of you on a pane of glass; the image you draw is certainly 2D, but does absolutely reflect the real world (because you are tracing the outline of real objects in a real scene on the glass). Perspective is correct, despite it being a 2D image. You can take your drawing with you, look at it later from the same distance, and everything is still correctly proportioned and all parallel lines (in the real world) will point towards their respective vanishing points, just as they did when you drew it, just as they did when you were actually looking at the real-world scene.

The problem comes when you hang the picture on a wall, and view it from a greater distance. Now it becomes a representation of the scene, rather than being directly viewable as what you saw when you were there - because your FOV no longer matches that of the picture (refer FOV diagrams on the last page).

You have said a few times that you can't 'fix' incorrect perspective, but I suspect you're just thinking about it too much. As long as the image is drawn correctly, based on a specific distance from the eye (real or virtual) to the screen (again, real or virtual), you can make the perspective appear completely correct by just matching that distance when you view it. (but, again, change the screen size and you need to alter the distance... refer ZeosPantera's original post)

You are perhaps picturing things like the edges of a straight road receding into the distance, and saying that if they are drawn at the 'wrong angle' they will never look right - but when you get closer to the image their apparent angle (to you) will change - because of perspective. They're still straight lines, they're still drawn in the same position on screen, but because your position changes their apparent direction (and position) will also change.

I really am in great danger of just repeating myself now, so if you have a reason this isn't making sense I'd appreciate the input.
 
Recently I sat down and rethink some things about setting FOV and it looks like you were right, guys :)
That distorted picture - it will indeed work when display surface is flat (like you have one big screen).

I apologize for the whole fuzz :)
 
I apologize for the whole fuzz :)

No Harm No Foul.

I have a good iRacing triplehead video showing proper FOV. Most I find look off.


I'd love to find a similarly good rFactor video with it set up. No luck yet.
 
I apologize for the whole fuzz :)

I think half the battle with forum discussions is being able to communicate what we're actually trying to say, and obviously also to comprehend what's coming from the other person :) The posts as they are might answer some questions for others, so I don't see it as a bad thing, but I must also apologise for being a bit abrupt in the last couple :eek:

Anyway, onwards and upwards :cool:
 
A few pictures I edited up for the next argument to help show how proper fov works.

rFactor2011-03-3001-28-03-2.jpg


rFactor2011-03-3001-28-21-81a.jpg


rFactor2011-03-3001-26-18-91.jpg


Think they pretty much speak for themselves.
 
Not had my pc setup for a while and now I've also got a 26" screen to play with :D

Calculations said I need 20 as my fov.
loaded it up,adjusted seat and it looked great :)
then I got to a few corners and they seemed like hairpins rather than the nice flowing corners I was used to :(

just can't win :(
 
They seem like hairpins because they are. All turns should seem more severe since they aren't being squeezed onto your screen artificially. Best example of that is probably the downhill at Bathurst. Seems like a gentle chicane with a default setup.
 
Jumping in here with some feedback of my first experiences ....

holy cow .... Basically with low 20ies FOV, and tweaked down motion, it was all over place and nowhere right!

then I got to a few corners and they seemed like hairpins "Yesterday 10:28 PMMystaMagoo "

A perfect way to say it.

I've setup my TrackIR to better look into the corner and this helps with the 'preparation / anticipation' side of my brains input and now it's better but there's alot of getting used to.

I'm building a rig [pet project] and experimenting with FOV alot so I think this is worth the time and effort [I'm trying to be as realistic as possible] but for the more casual desk pilot [as I still am by the way], it'll be horses for courses!
 
It's amazing, really, how our brains can see high-FOV on small screens (as most of us have) and make sense of them - whether it's computer games or TV shows - and even get to the point where seeing 'true' perspective actually looks wrong. Or feels wrong.

Anyway, for most people I don't think true perspective is practical, because your average screen and viewing distance gives you a very small window into the game world. I wouldn't go and drive my (real) car with such a small amount visible (imagine blacking-out all the windows and just leaving a screen-sized hole on the windscreen), and I don't choose to emulate that situation by giving myself true perspective in rFactor.

Still, for those spending the money to get themselves set up nicely, I think having the screen real estate (and position) to be able to run proper perspective is something well worth considering. But, while you can work out the maths and set the FOV to give you true perspective, and it's exactly the same (and only) perspective you get in real life, there will certainly be a period of adjustment as you have to re-learn how much to turn the wheel for corners because they will seem to 'bend' more as you approach them - the same thing happens even with relatively minor FOV adjustments such as when buying a bigger screen.

My advice would be, if you can't drive with proper perspective because your screen isn't big enough or close enough to let you see a decent amount of track, don't even bother trying to get close. In my opinion ZeosPantera's videos, although showing correct perspective on his setup (and potentially the viewer's if they show it fullscreen and have an equivalent screen size / viewing distance ratio), are not practical for racing. You need to be able to see more around you, both for seeing other cars and for keeping sharp corners in sight as you approach them, and you won't get that with proper perspective unless you have a big screen suitably positioned.
 
And despite all that, it is still do-able.


My brother, who doesn't sim-race ever, finally sitting down on my setup and taking the HistorX cobra around Bathurst.
 
They seem like hairpins because they are. All turns should seem more severe since they aren't being squeezed onto your screen artificially. Best example of that is probably the downhill at Bathurst. Seems like a gentle chicane with a default setup.

I kinda agree,we definitely need better fov.
But take for example the 1st corner at silverstone.
Virtually flat out in an F1 car but with tweaked 'correct' fov it's not.

So it's a compromise of realities.
Do we want it to 'look' right or 'drive' right?
 
Virtually flat out in an F1 car but with tweaked 'correct' fov it's not.

So it's a compromise of realities.
Do we want it to 'look' right or 'drive' right?

Its a matter of knowing the track well enough to be confident in it's layout. If you go full tilt into turn one of Silverstone in a high FOV you should do the same in correct FOV. What you do with the car doesn't change. Clipping the Apex and braking markers don't move.

It's ironic that this was the last video I uploaded, 1992 F1's around Mid-Ohio. Keeping in mind my V-Fov is still 20.5 and this is my first outing in both the car and on the track AND my paddle shifts are broken so I had to use my Gated shifter for these laps.


Don't lose that sense of speed in an F1 car.
 
It's not about sense of speed. It's about not hitting a car that's right in front of your right front corner, like you did with your friend in that other video (Falcon. GT40, Nurburgring) :D If you have a very unrealistically narrow window in front of you you'll have to make a compromise between the realistic undistorted vision and realistic field of view (i.e. being able to see thing around you).

You try really hard you can learn to run hotlaps round the track looking through a keyhole, if you want. But you won't be allowed to race like that with other cars in RL and you shouldn't be in simracing, IMO.
 
Last edited:
Filling your actual field if view with screen(s) is definitely the solution... a tiny percentage of simracers can afford (and/or accommodate). Most have to look for a compromise.
 
I think most simracers:
- use default settings, and doesnt think about what is right or not
- opts for greater visibility at one look rather than real view
- opts for greater speed feelings, rather then real one
that's all.
 
Filling your actual field if view with screen(s) is definitely the solution... a tiny percentage of simracers can afford (and/or accommodate).

Building a huge multi-monitor setup may be expensive but moving your current, single monitor closer to you costs nothing. I have seen many setups where 22" monitors sit over three feet away. If you can comfortably move the monitor to say 24-26 inches away using proper FOV is much more manageable.

And 23" monitors now cost as little as $160 and all modern graphics cards will support multiple monitors. Another route would be grabbing a single 32-40" 1080P television for around $400. Anyone keeping track of sim-racing product pricing will realize modern hardware like wheels cost from $250-600. Then there are shifters, external gauges, pre-made cockpits, pedal mods, racing seats, etc. This can be an expensive hobby and the monitor IS the most important part or on par with the wheel and pedals you use.
 
I kinda agree,we definitely need better fov.
But take for example the 1st corner at silverstone.
Virtually flat out in an F1 car but with tweaked 'correct' fov it's not.

Suddenly you realise just how much grip those cars have... because that's how the corner actually looks, and they'll still go around it pretty much flat out... the problem is you're not used to it looking right :D

The other problem, obviously, is that when it looks right you do need big/multi screens or you're literally looking off-screen to work out where your apex is on anything but the fastest corners - which I agree isn't much good.

Building a huge multi-monitor setup may be expensive but moving your current, single monitor closer to you costs nothing. I have seen many setups where 22" monitors sit over three feet away. If you can comfortably move the monitor to say 24-26 inches away using proper FOV is much more manageable.

As it happens I've recently moved my 22" screen closer, and calculating the VFOV gives me 21° - a pretty close match for your vid. So I opened it up in youtube, buffered it, put it on fullscreen, sat at the proper distance... and it did look pretty cool down the straights. Corners, though, had me feeling nauseous pretty quickly, and just constantly wanting to look farther (off-screen) than I was able to, so I could see corner exits etc.

To me it's just not practical, though I'm sure with time you could overcome the nausea and learn the track well enough to be able to hit apexes and maximise exit speed without actually being able to see what you're going to do before you get there - but wheel to wheel racing is surely compromised. (it's obvious from that Mid-Ohio vid - not being able to see cars you're liable to run into just isn't a way to race)

I have recently had the pleasure of trying a 42" screen at close quarters, albeit not with correct FOV, and even having bridges actually being above your head adds a lot to the immersion - and I was surprised the larger pixels weren't too much of a problem down the straights. I could easily imagine playing with three such screens, with correct FOV, and being very happy with it.

Now... just need more money and more space... lol

(and just to add one more component to the mix... 3D with proper perspective would be very very nice)

For actual racing, I would maintain the best approach is to run as little FOV as you're comfortable with (as in, how much you can actually see around you, be that corners or cars) and do whatever you can to get your actual FOV (your eyes to your screens) as high as possible through screen positioning, size, and number. If you can get the two to match, enjoy :p
 

Back
Top