Setting up your rFactor FOV - Tutorial

[ZeosPantera]

Yeah I too think 60 is too close - unless you have more than one screen. 60 feels perfect when you have triple screens setup.

? really ? 60 VERTICAL FOV is too close? If your working on a SIMULATOR, a real racing sim, then you should be working to get as close to the real thing as you can.

Honda seems to understand FOV and triplehead.

Look at the center screen. Is that not the view I am preaching here? I am pretty confident honda know how to create a vehicle sim to realistic spec.

Now I seem to have to whip this example out in every forum at some point. I will show you the worst most incorrect FOV I have ever seen.

He is sitting in an F1 car with visible REAL front wheels and look at where her virtual ones are. This is a great example of "Doing it wrong"

Now for a shining example of doing it right. Same basic cockpit. Same style of f1 car. Note where the in-game wheels appear to be.

The virtual wheels are where the real wheels SHOULD BE! He is using a simulator. The first pic is just someone sitting in an overpriced uncomfortable seat playing a racing game.

UPDATE: This image was posted along with the new interview with simHQ. It is a correct FOV as you can see the tires almost line up with the real ones and the virtual wheel is turned off (like it should be) but the actual view is obviously too high. A bit of tweaking could get that close to perfect. Better still cut a hole and lower all the monitors down 2-3 inches.

 

[LesiU]

Here are two videos showing differences between low and high FOV. In both, I modified camera position (my reference was visual distance to ADL display - in both is the same) but only in horizontal plane (fore/aft). Vertical seat position is exactly the same in both videos.

FOV 28: http://www.youtube.com/watch?v=HoARQ1P_nFY
FOV 60: http://www.youtube.com/watch?v=0c-S0F0Sj78

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.
Second thing - relative distance between ADL and steering wheel. At low FOV it just feels right. ADL is about 17-18cm wide. OMP steering wheel is 32cm in diameter and is about 20-25cm from ADL.
At high FOV the ADL looks like it is very small when looking at it straight. When I turn my head, it get zoomed in towards camera (and got stretched). That's because with high FOV you get artifically zoomed out picture (in center of the screen that effect is more visible). The same is for example with right external mirror, how far it seems to be when you look at it with high FOV... but in fact it is only about 1,2-1,3m from the driver

 

[fobban]

I too like lower FOV, but I must say that it makes it a bit harder since it's nearly impossible (at least for me with my single monitor) to look at the mirrors during race (one would have to look right and left). That makes it really hard, especially during F1 starts when it's crowded.

Easiest when only having one display (this is my personal opinion) is using the TV cockpit view and the virtual mirrors on top. But the TV cockpit isn't exactly realistic, so I don't use that view any more.

Why can't we get the virtual mirrors when using the cockpit mode? Sure it's not realistic considering the small monitor and the FOV you've got, but at least it tries to mimic real life where you don't really need to turn your head that much to look at the mirrors.

 

[ZeosPantera]

Why can't we get the virtual mirrors when using the cockpit mode? Sure it's not realistic considering the small monitor and the FOV you've got, but at least it tries to mimic real life where you don't really need to turn your head that much to look at the mirrors.

Happy Birthday.. YOU CAN! Instructions are at the bottom of the first post.

 

[fobban]

Happy Birthday.. YOU CAN! Instructions are at the bottom of the first post.

Awwh, awesome! Thanks ZeosPantera!

 

[Ferrino]

Now for a shining example of doing it right. Same basic cockpit. Same style of f1 car. Note where the in-game wheels appear to be.

The virtual wheels are where the real wheels SHOULD BE! He is using a simulator. The first pic is just someone sitting in an overpriced uncomfortable seat playing a racing game.

This is a great (and rare) example where the virtual and real cockpits align well (i.e. where you don't feel like you're driving a remote-control car).

How do you manipulate the position of the camera to get a similar view in rFactor?

 

[ZeosPantera]

How do you manipulate the position of the camera to get a similar view in rFactor?

See first post.

 

[Jouni]

I´d like to thank all of you contributing to this thread. This has risen the enjoyment of mine in rFactor greatly. Ive been driving with 60-65 FOV since rF release in all mods i race with.
Im now at default 35 FOV, although not still probably low enough, but is much more fun to drive. At first, I thought i wouldnt manage with it very well, but after a few hours of practice and testing different mods, I settled in and it felt very good and right. Last night broke my pb in an online race that i managed to also win to my suprise.
Today have hade awesome fun playing with the old Historix cars with the new FOV.
I highly recomend to give it a go. Hey! if an old 55y old Grandpa can learn new things, im sure lots of you will be suprised how well it reals-in after some practice.

edit: the difference in 65 vers 35, moving seat back to still see the dash.

 

[LesiU]

Great!

 

[ZeosPantera]

Some new footage of the FOV from a real world perspective.

 

[pcline25]

AWESOME Tutorial! As one who runs a lot of iRacing I found it hard to go back and run ISI type sims and mods but this guide changed all of that. Can't thank you enough for this.

 

[Lazza]

As usual in these discussions most of the 'arguments' are just people referring to different things.

And the problem with videos like the one above is that we're viewing the screen as a relatively small image - so the perspecitive mimics high-FOV values in a way, but with a much smaller viewport. So of course people will think it's unrealistic, because to them you're just limiting your view.

The point with using lower FOV is to match it to your screen, and its distance away from you, so that what you're seeing is 'correct'. As far as realistic goes, that's the only way to get it right.

However, this requires a good amount of screen real estate to be able to see anywhere but directly in front (unless you want to use that terrible 'look-ahead' feature [ok, I'm sure people get used to it and like it - congratulations ]) - so if you want to be able to see a bit to the side, and your mirrors, and you don't have huge/close/multiple screens, low FOV isn't for you.

Just to recap; there are 2 types of realistic here:

1. Being able to see about the same amount on screen as you would in real life.
2. Perspective being correct.

You can get both happening, but you need big screens. I suspect most people tend to favour option 1, because while proper perspective is nice it's even nicer to be able to see where you're going and where other cars are.

It's all relative of course, and personal preference.

 

[LesiU]

Yup, that's true. To really have good experience with natural FOV, you need really wide picture. Here are 2 examples of how it would like on a triple panoramic screen (I only have TV, so I fired up rF in a window mode, with SoftTH on).

Could you tell from them, that's "only" FOV 25*?

 

[Lazza]

Could you tell from them, that's "only" FOV 25*?

Of course not... because if I display those pictures fullscreen on my 23" screen it looks like a 61° FOV with the top and bottom cut off... unless I sit with my face just under 25cm (10") away from the screen

I hope this is helping some people understand the objective of this thread, and why they might not like what they're seeing with such low FOV values. I wouldn't recommend modifying your FOV for realistic perspective if you have a single screen in front of you - but, like extreme look-ahead, some people will do it and like it... each to their own.

I agree that having huge multiple screens and still using a high FOV is a bit silly, because the opportunity is there to make it look correct. But when people are used to that, it will take them a while to get used to driving with an adjusted FOV - even if it is what they'd see in real life. It's an interesting situation, where they'd sit there looking at correct perspective and say that it looks 'wrong' - purely because it's not what they're accustomed to

 

[LesiU]

Lazza, rF perspective has nothing to do with size of your display, and how far you sit in front of it. And here's why.
Let's assume 2 guys, both have same panoramic 16:9 screens, but in different size and sit at different distance in front of them:
1. 92,6" diagonal picture on a wall, sitting 1m in front of it. That should give you 60* of vertical FOV.
2. 21" diagonal monitor, sitting 0,5m. That is 20" vertical FOV.

Both have proper FOV, right? But do both have proper perspective? Launch 60* FOV and go drive round a track, and compare to what perspective you have with 20* FOV.
Your attempt just won't work like would like! If ISI would provide the same proper perspective and you, by changing the FOV, only "cut off" a slice from that picture for your calculations (so that 20 or 60" is a part of a bigger screen with proper perspective) then that will work. But what we have in rF 1, it just won't. Perspective on that 92.6" sized picture won't change if you step back from 1 to 5 meters It will still be the same, because of high FOV value.

That's why you have to set FOV with proper perspective (from my tests, it might be something between 28-31*) and if you want "get real", buy a screen big enough, to provide 1:1 scale sized elements from a place where they would be in real world. So if you sit 2m in front of your screen, having your steering wheel in front of you in normal driving position... then Motec display is usually 20-30cm behind the steering wheel. When you put in that place a ruller and measure from that point, Motec ADL width on your screen, then you should get.... 18cm, with picture with realistic in-game driver position. I hope you understand what I mean

Talking about FOV in general - yes, most people don't want proper FOV because they feel weird when see a picture like through a box...but that's how they would see the world in real life, if they will look through a box sized like their monitor/TV.

 

[Lazza]

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

Hmm... the distance from the screen, and the size of the screen, is what this is all about. I may be misunderstanding, but it sounds like you think rFactor itself is somehow doing something weird with the perspective so that a particular FOV is 'correct', while all others are wrong. This is not true.

To test the 2 layouts you describe, you need to be able to sit in those 2 layouts. If you record rFactor on your system with 25° FOV, and it looks good to you, but I then play that video in windowed mode, or use a smaller screen, or sit farther back, it will have pincushion distortion - because my FOV is lower than yours, same as yours is lower than rFactor's default. Perspective is all about making what's drawn on screen match what you would see through a plane at the same distance and size as your screen, if you were in that 'world'.

For a long time (maybe still) FPS games had 90° horizontal FOV - this gave a pincushion effect because for the perspective to work you needed the left edge of the screen 45° to your left, and the right edge 45° to your right - then the perspective drawn on screen matches what you'd expect - but of course with the small screens generally available that means being very close to it. No one sat that close, so the perspective looked 'wrong'. But, it was actually fine, and has been since the days of Doom.

You can make rF default FOV look fine... just get a big enough screen or get very very close to the one you have. I'm not saying it's an ideal way to play, but the perspective works fine.

If ISI 'forced' a low FOV that made sense sitting at your 21" screen (say 20°), you would have correct perspective and a small window to look through, yes. But anyone with a huge screen and an actual FOV of 60° will have a barrel distortion effect using those values. (that is, objects at the centre of screen will appear too large in comparison to those at the edges - the opposite of pincushion distortion)

That's why this thread is about calculating your true FOV - so that you can set rF accordingly and the perspective is correct.

 

[ZeosPantera]

Hmm... the distance from the screen, and the size of the screen, is what this is all about.

For a long time (maybe still) FPS games had 90° horizontal FOV - this gave a pincushion effect because for the perspective to work you needed the left edge of the screen 45° to your left, and the right edge 45° to your right - then the perspective drawn on screen matches what you'd expect - but of course with the small screens generally available that means being very close to it. No one sat that close, so the perspective looked 'wrong'.

You can make rF default FOV look fine... just get a big enough screen or get very very close to the one you have. I'm not saying it's an ideal way to play, but the perspective works fine.

If ISI 'forced' a low FOV that made sense sitting at your 21" screen (say 20°), you would have correct perspective and a small window to look through, yes. But anyone with a huge screen and an actual FOV of 60° will have a barrel distortion effect using those values. (that is, objects at the centre of screen will appear too large in comparison to those at the edges - the opposite of pincushion distortion)

That's why this thread is about calculating your true FOV - so that you can set rF accordingly and the perspective is correct.

Absolutely correct. I would say this thread is actually about Playing rFactor as a game Vs Using it as a simulator. A game is allowed to skew your perspective, break the laws of physics and exaggerate explosions and such to make your experience fun. A true simulator should pretty much force everything to be as close to realistic as possible even if it no longer becomes fun. What annoys me is when they put a real race driver in a "sim" and I can see the FOV is completely incorrect and the rig itself is setup wrong. http://www.virtualr.net/mark-blundell-tries-f1-1991-he-for-rfactor-video/

I first noticed it in the Jeremy Clarkson @ Laguna Seca segment of Top Gear when he raced it on the PlayStation and then tried to match it in real life.. Very first thing he says is "I'm going down a strait that's not strait!! That is not a corner on the playstation" and that can be blamed solely on Gran Tourismo's FOV.

Furthering my testing I have applied my 20.5x32 Degree RL-FOV to other games and had mixed results, it is great in Garry's mod for showing exactly how huge things that you build actually are and the maps are astonishingly scaled. Less successfully HL2-EP1 because the levels aren't designed for such a limited perspective and you have no idea whats going on around you. I have also played arma2 in a "close to" accurate FOV which makes it easier to see enemies over a distance but more dangerous around urban settings, Thus the squeezed 21x50 Fov. Another example is a friend set the game "Amnesia" to his correct 34 Degrees Horizontal fov and said it took the experience from scary to pant browning.

If you go in understanding the compromises and the science behind it I find most people will get annoyed at how they were forced to play previously. I for one can no longer drive in anything but my correct fov. It switched in the video 4-5 posts up and the terrified scream I let out was due to the horror of changing from 20 to 65 fov, it took me right out of what I was doing and I nearly crashed it. Most new people I tell to just throw the FOV slider down to 35 and use that for a while. I'll ask the same of you Lazza. Don't go all the way. But ease yourself down until your comfortable... then go a little bit more.

 

[Lazza]

Most new people I tell to just throw the FOV slider down to 35 and use that for a while. I'll ask the same of you Lazza.

Oh, I'm not adjusting mine. Mine's set to 60, which already gives me a lot less peripheral vision than the guys I race against with similar FOV and triple screens

My priority is being able to race, so getting more realistic perspective will have to wait until my setup changes a lot. But I can see the merits, so just thought I'd try and clear up some of the confusion in here

 

[LesiU]

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 Picture on the monitor won't change when you change your sitting place. You can move to the left, right, up and down... the picture on your screen will remain the same!
Screen size also doesn't matter at all. rF have no idea if you have 7" or 150" display. So it doesn't matter if you have FOV 60* on big screen or small. Perspective with that FOV will always be the same, no matter what monitor you have, of what proportions (4:3, 5:4, 16:9... doesn't matter). Of course you keep in mind that in rF, you are working with vertical FOV?

 

[ZeosPantera]

Calm down LesiU. He does get it I am sure. I don't need a misunderstanding argument going back and forth for 8 pages.

 

[LesiU]

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, because with different FOV, you will get different perspective.

 

[MaXyM]

Maybe problem comes from misunderstanding between FOV and size of area you are seeing virtual world

- FOV - is angle between perspective edges. If this angle is wider you can see more at once, but perspective is going to be fish-aye. If perspective is low (real in that case) you can see less in the same window (camera view) than with larger FOV
- view window (camera view, or view plane in 3d) - the window you see virtual world through. this window may be moved over scene (ie simulating moving driver eyes closer or farer to/from windscreen).

We are using FOV therm describing 'computer science value' rather than physical one which is view window.

 

[ZeosPantera]

I plan to eventually make a quick "understanding FOV" tutorial video with my camera and some cardboard to represent my monitor. I explain things better by speaking about it them trying to type it.

 

[LesiU]

I think, problem is not with us (me and Maxym) but you, with setting different FOV with relatation to how far you sit in front of a display and it's diameter.

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

 

[ZeosPantera]

um.. actually you would have to.

Example:: If you had your steering wheel turned on in-game (and you shouldn't) and you properly adjusted for a 12 inch high screen @ 34 inches away (~20 Vfov). Then adjusted your seat to perfection your in-game wheel in that car would measure say 12 inches across if you actually measured the screen.

Now go upgrade to a 50 inch plasma with a screen height of 25 inches and stay 34 inches away. If you calculate correctly (36 Vfov) and the seat is in the same position the wheel should measure exactly the same size on this new screen. All you have done by upgrading the display is add more area to see.

But, If you put that new 50" screen 65 inches in-front of you instead of 34 the Vfov calculates to about 20 again and you have gained nothing. And the in-game steering wheel would now be about 24 inches across.

The goal of sim design should be to get as close to as big a screen as you can handle staring at.

 

[LesiU]

Now go upgrade to a 50 inch plasma with a screen height of 25 inches and stay 34 inches away. If you calculate correctly (36 Vfov) and the seat is in the same position the wheel should measure exactly the same size on this new screen. All you have done by upgrading the display is add more area to see.

OK, Now I upgraded to a display that is 50 inches high and stay 34 inches away. FOV should be 72deg. Now launch rF and see for your self, how perspective gets unnatural (fish-eye like). If you still don't see this, then go with FOV 90 (68" high display and 34 inches away). Still don't see any difference in perspective, compared to what is with low FOV?

FOV 28: http://www.youtube.com/watch?v=HoARQ1P_nFY
FOV 60: http://www.youtube.com/watch?v=0c-S0F0Sj78

 

[MaXyM]

Now go upgrade to a 50 inch plasma with a screen height of 25 inches and stay 34 inches away. If you calculate correctly (36 Vfov) and the seat is in the same position the wheel should measure exactly the same size on this new screen. All you have done by upgrading the display is add more area to see.

it would be true in case of mounting wheel at 0 Z-position. Because it isn't - you have move the sit instead of fov.
remember that fov in rf changes perspective. Making object closer or bigger is only (natural) side-effect. But we want to get in sim correct perspective to get correct feel of speed/distance. We should at least

The goal of sim design should be to get as close to as big a screen as you can handle staring at.

you can handle any size of screen with more or less correct view.
Problem is that rf is able to change only fov - not panning a screen. Or in other worlds, you can set size of camera clip-plane but not size of far-plane (settings exist but doesnt work). It makes perspective unreal (doesn't apply perspective correction). As I remember iR does in a right way.

 

[Lazza]

OK, Now I upgraded to a display that is 50 inches high and stay 34 inches away. FOV should be 72deg. Now launch rF and see for your self, how perspective gets unnatural (fish-eye like). If you still don't see this, then go with FOV 90 (68" high display and 34 inches away). Still don't see any difference in perspective, compared to what is with low FOV?

FOV 28: http://www.youtube.com/watch?v=HoARQ1P_nFY
FOV 60: http://www.youtube.com/watch?v=0c-S0F0Sj78

You can't watch videos with different FOV on the same screen - if, by luck, either of them actually produces correct perspective on your setup, the other will not.

For a 28° FOV video to look right, your screen has to occupy 28° of your vertical [field of] view. And for the 60° to look right, the screen has to occupy, yes, 60° of your view. This is why the size of your screen and your distance from it is critical in calculating your ideal FOV if you want to see correct perspective with your particular setup.

FOV determines perspective. If you match it to your situation, so that the in-game window (plane) covers a vertical angle from the in-game cameras perspective equal to the vertical angle your screen covers from your perspective, it will be correct.

This, as I said, has been true since the first 3D games came out... I'll admit I'm not sure what you're getting at MaXyM, but unless rF really is doing something very unusual I just don't see how setting the FOV is in any way independent of the 'perspective' produced.

This isn't some computer game phenomenon... any captured image, be it camera footage or a photo, covers a certain FOV. Unless you view it from exactly the right distance (which will vary depending on the image's size) you will get perspective distortion - barrel, or pincushion, if you are too close or too distant respectively.

Edit: Here's a simple practical example to show the importance of viewing position. Have someone stand a distance away from you, and place an object between the two of you. Tell them to look at that object. They will appear cross-eyed to you. Now, if they keep looking at that object, and you move closer to them until you are right at the object's position, they will no longer look cross-eyed. Finally, if you can raise the object slightly so that they can keep looking at it over your head while you continue to move closer to them, it will appear that their eyes are looking in opposite directions (slightly!).

The fact that they appear cross-eyed, 'normal', or their eyes are 'apart', is purely dependent on your position - and where they're looking.

In the case of rFactor, and any game/video/picture, the other person's focus represents the set FOV; your position, is, your position. Sit too far away from the screen and you get pincushion (cross-eyed). Sit too close and it's barrel distortion (the other one!). Bearing in mind, obviously, that you can also adjust the FOV to make it correct whatever your screen and position - at the likely expense of losing a lot of visibility in the game.

 

[LesiU]

But it doesn't matter if I view that 60* FOV video on tiny 3" screen or 10000" wall of a castle. Doesn't matter if I have my nose 2cm from the picture or 200m. At 60* FOV perspective will be already distorted sort of "fish-eye" on any screen, because it's the high FOV that makes it look unnatural, not screen size or my distance to it, and it will remain unchanged, until you change rFactor FOV
The higher the FOV, the more distorted perspective! On the other hand, setting too low FOV will also change it in the other way, which is also not good.... and it won't magically change to normal perspective with my different position in front of the picture, size of my TV or whatever

You can't compare what you see in real life, because in real life you move towards that object, in relation to the guy, while rF generates picture with it's own perspective, defined by its FOV. rFactor is not asking you about your display size or how much away of the display you sit, right? I mean... you know that, right? So whatever position you take in your room, picture in the screen will be exactly the same, all the time, no matter what is your position in relation to the display
rFactor has its own world, with it's own perspective defined by FOV and it has nothing to do with where you sit with your steering wheel.

If you want to have realistic position, then you have to set FOV with proper perspective. Which is?
Well, in photography, for portrait photos, photographs use 45-55mm lenses for 35mm film. I had somewhere a picture that shows FOV for few different type of lenses (makro, wide, potrait, few zooms - all that is related with focal length of course). Those FOVs were diagonal. Picture is 3:2 so it's easy to recalculate vertical FOV from it. That was something between 28-31*.

Having that set, you have to decide, how far from the screen you would like to sit and based on that, prepare large enough screen. Also, at what point you want to have transition into rF world and objects.. I mean, If I have my steering wheel base/case 40cm from my TV screen (and that point is 20cm after the steering wheel, so the steering wheel is 60cm from my screen) and I know that in real car the dials/ADL or whatever, is also 60cm after the steering wheel (something like that is in DBR9 GT1), then you want to have a picture on the screen that is exactly scaled 1:1 with real dash. You can do that with combination of virtual seat position and your screen size.

If in real world car, the dash (let's say, there is MoTeC ADL) is 20cm behind the steering wheel and my screen is 60cm then I have to have the pictured scaled more than 1:1. All that must be measured from your POV (the real world POV, where you sit in your seat) so that after you enlarge the picture, that ADL will have proper size at that spot where it "should" be placed (but on the display it will be bigger because of that).

 

[Lazza]

But it doesn't matter if I view that 60* FOV video on tiny 3" screen or 10000" wall of a castle. Doesn't matter if I have my nose 2cm from the picture or 200m. At 60* FOV perspective will be already distorted sort of "fish-eye" on any screen, because it's the high FOV that makes it look unnatural, not screen size or my distance to it, and it will remain unchanged, until you change rFactor FOV

Ok, this is where our disagreement is, so let's please focus on this.

Let's use 60° as the example. If you run rFactor with this FOV, on your single screen at a 'normal' viewing distance, you'll have 'pincushion' distortion. (you're referring to it as fish-eye) Objects near the edges of the image will be stretched and appear larger than those in the centre. I'm sure we can agree on that.

I'm also sure we both understand the effect of perspective on objects - the closer you get to something, the larger it appears, and the more extreme its apparent perspective becomes. (so distant objects appear 'flatter', while close ones have more depth)

So, to interpret the 'pincushion' image a different way, it's like we're too close to the objects around the edges, or too far from the centre ones. Hence your 'fish-eye' effect. (I believe fish-eye more correctly refers to barrel distortion by the way, but I know what you meant)

Now, take that same image, and project it so that it fills an entire wall of your room.

If you stand back so that it fills your view to the same extent it did on your screen, you will see exactly the same distortion you saw before. You might be 5m away instead of 0.5, but everything's just scaled up so it's all equivalent.

Now, if you walk forward until you are 1m away from that huge image, the apparent perspective will change.

Why? Because the objects in the centre of the image, which as we said before appear too small and distant, are now only 1m away from you. The edges of the image, too 'close' before, are a couple of metres to your left, right, above and below. From your current position they are more distant in relation to the centre of the image - thus you are introducing your own barrel distortion to an image that, from a distance, had pincusion distortion.

If you only get close enough so that the image FOV matches your true FOV (the angle the image covers from your current viewpoint) the two cancel each other out and the perspective looks correct.

I hope that makes sense.

 

[ZeosPantera]

At 60* FOV perspective will be already distorted sort of "fish-eye" on any screen, because it's the high FOV that makes it look unnatural, not screen size or my distance to it, and it will remain unchanged, until you change rFactor FOV

The higher the FOV, the more distorted perspective!

I think I know what your on about. The reason the edges distort at higher FOV is because screens are flat. Your eye may be 34 inches from the center of a 50 inch screen but at the far upper left corner it would be more like 65 inches away so knowing that games make things in the corners bigger to compensate. If your screen was a partial sphere and you could be the same distance from all parts of it and distortion would not occur. This is the reason triple-head is preferable over a single massive display and why the Jdome comes to mind when discussing this.

You can't compare what you see in real life, because in real life you move towards that object, in relation to the guy, while rF generates picture with it's own perspective, defined by its FOV. rFactor is not asking you about your display size or how much away of the display you sit, right? I mean... you know that, right? So whatever position you take in your room, picture in the screen will be exactly the same, all the time, no matter what is your position in relation to the display

To a point. TrackIR 6 axis head tracking fixes that.

rFactor has its own world, with it's own perspective defined by FOV and it has nothing to do with where you sit with your steering wheel.

Until you set it up. And force it to be correct. If you want to have realistic position, then you have to set FOV and move the camera to the proper perspective.

Having distance set, you have to decide, how far from the screen you would like to sit and based on that, prepare large enough screen, then you want to have a picture on the screen that is exactly scaled 1:1 with real dash. You can do that with combination of virtual seat position and your screen size.


Here are a few of the demo models I have built.

This one shows how different size screens at different distances can equate to the same Vertical Field of View so getting a new larger monitor and sitting further from it can be a detriment. Sizes are 22" -- 32" -- 58" -- 120"



here I am building a few mock desks with different monitor sized keeping the distance the same. Showing how the FOV's are measured.



I will have to work up a script so that I can discuss it without stuttering when I record it.

 

[Lazza]

I should add I'm making the assumption rFactor renders based on projection through a flat plane in the virtual world - and varies FOV by altering the distance between this plane and the virtual eye (so to speak). Generating an image suitable for a curved surface would, I believe, be a different matter entirely.

Hey, could be wrong

 

[MaXyM]

here I am building a few mock desks with different monitor sized keeping the distance the same. Showing how the FOV's are measured.

Good try but it shows where you're missing the point.
FOV in rF is not representation of FOV of player's eye which watching monitor. It is perspective setting factor for 3d scene (it's rather connected to eye of camera on 3d scene)
in short: 3d scene is rendered between 2 rectangles (usually with the same aspect ratio): viewing plane and far/clipping plane. Even if scene may be bigger, all objects rendered between those planes are generated. Note, that related corners of planes may be connected by a lines making (reversed) perspective

Perspective gives to us feeling how far objects are.
In rF, changing FOV means that only one plane is changed (probably viewing plane goes bigger). Clipping plane stays untouched (the setting is present in camera files but doesn't work as we referred before). So... if you change size of one plane but not another one, you simply change perspective.

Changing perspective is unnatural things for human eye. It cannot change this factor. That's why we (me and Lesiu) saying that in case of rFactor there is only one correct FOV (means real), doesn't matter on other things. So if changing perspective is unreal, your approach cannot be true.

Hope it make the situation more clear

 

[ZeosPantera]

Are you referring to the plane the monitor represents (viewing plane) and how objects in game that are closer to you then the monitor's distance should not be rendered. IE steering wheel and part of the dash would most likely be on the wrong end of the screen's plane?

 

[MaXyM]

No. every 3d object rendered on scene is between both planes.

Point is that manipulating FOV value in rfactor, you are changing ratio between both planes dimensions. In results you changes perspective which is unreal. Call it 'bug of rf'

 

[Lazza]

Here's 2 images showing, as much as I could manage with a quick Windows Photo Viewer setup, the same perspective:

Note: Viewing this image in a separate window at full-size would be good idea

You might be interested to know the left image is a screenshot taken at 35° vertical FOV. Why is the one on the right pixelated?

Because, it's just zoomed in from a wider angle shot.

Here's the original right-hand side image in its entirety:

That is 90° vertical FOV.

The last image appears (more) distorted because your distance from the screen makes your true FOV much smaller than that of the image. If you move your face much closer to the screen, as I've simulated with the right image up top, the FOV starts to match much more closely and the perspective makes more sense.

I hope the pictures help... and sorry for the image size...

You really do have to open these images up separately... they get squashed down on the forum and appear to have very similar FOV, making the illustration very difficult

 

[MaXyM]

so another example to crash myths
Screen is from GTL but this is the same issue/the same engine

fish-eye effect caused by big fov makes definitely feeling that 2 cars close have different size (which is incorrect). It doesn't matter on what big screen you display this picture - it will be always distorted.
Note you are watching always whole screen. Not clipped area.

 

[Lazza]

It doesn't matter on what big screen you display this picture - it will be always distorted.

You say that with a lot of confidence. Have you actually done it?

 

[MaXyM]

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.

 

[MaXyM]

strange question.
If you see amount of objects in unnatural proportions, it will always be looking strange. on 14", 32" or 100". Because brain compares the same objects to how those would look in real world. For example a car would always have wide about 2m even if you display it on 1000" big screen. Any other objects will be compared to its 'known by brain' size during judging the distance.

You may crop the picture (as you do) minimize the bad feeling. no less no more