Widescreen Gaming Forum

[-noun] Web community dedicated to ensuring PC games run properly on your tablet, netbook, personal computer, HDTV and multi-monitor gaming rig.
It is currently 29 Apr 2024, 03:59

All times are UTC [ DST ]




Post new topic Reply to topic  [ 11 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: 10 Dec 2008, 22:35 
Offline

Joined: 10 Dec 2008, 19:43
Posts: 8
OFF-TOPIC
First of all, sorry for my bad english.
Second: Greetings from Italy and congratulations for this useful forum.
1st post here and dunno how to write for my presentation.

IN-TOPIC
I'm a online FPS game and, in another forum, I'm talking about FOV calculation, changing ratio and monitor dimensions.
One player send me here:
http://www.widescreengamingforum.com/node/
and here:
http://www.widescreengamingforum.com/fovcalc.php
because he found that link and found different results of fov than what I calculate.
(I found something about 93.3° instead 100.39° for a 90° fov calculation from a 4:3 to a 16:10 screen)

One of these calc is wrong.
If mine is, I hope someone tell me what's wrong here.
If not, I hope to be useful like this forum.

WHAT I FOUND WRONG IN THAT CALCULATION?
That calc is strickly correct, but forget all about dimensions.
It compare a 4:3 and a 16:10 (or other ratio) by assuming that their screen height are the same:

(from 1st link)
Resolutions: 1280x800 / 1680x1050 / 1920x1200
16:9.6 = 16.00:9.6
4:3 = 12.80:9.6

Of course they aren't.
A 19''@4:3 has less width than a 19''@16:10, but is more higher.
This things "corrupt" the rest of the calc.

MY FOV CALCULATION
(hard to write in english, hoping is easily to read)
First of all, 3 things to keep in mind, or what I write is uncorrect.
1) the goal to calculate the FOV from a screen to another, is to watch a same object with the same dimensions in both screens.
2) with SAME FOV, EVERY screen shows the SAME HORIZONTAL FIELD, regardless of the ratio. The different ratio affects the VERTICAL FIELD.
(The Unreal Tournament Game Engines 1,2 and 3 works like that, sperimented by me many and many times)
3) changing monitors, the distance between user eyes and the sceen panel does not change. If it does, this (and every) calculation is not correct.
It's hard to me to considerate this distance, so I assume it is a CONSTANT.

Start:
Let's consider a 19''@4:3 and a 19''@16:10
With the similar triangles Theoreme and The Pitagora's theoreme, I calculate:
19''@4:3 = 15.20'' x 11.40''
19''@16:10= 16.11'' x 10.07''
(and I have verified with a ruler just onto my panels)

Now Assuming that human eye are in the middle of the panel.
For simplicity consider half panel, from the center to the left border.

The FOV is the angle from left panel border, the eye and the center of the screen (X2, of course)
The (half)width of the panel is the tangent of the circle where distance eye-panel is the radius.
IF a 4:3 ratio has 90° as FOV, means that the radius is equal to the half-width panel.
(I know, is not real for most cases, but matematically is menaningless is we not assume that)

Tangent (relative tu radius) of 19''@4:3 = 16.11/15.20 = 1.0599
Arcotangent of 1.0599 = 46.666

19''@16:10 FOV (starting from 90° 19''@4:3) = 46.666*2 = 93.333 -> 93°

WHAT'S WAS WRONG??

Thank for those who read since here.

(edited for fixing type errors)


Top
 Profile  
 


PostPosted: 10 Dec 2008, 22:53 
Offline
User avatar

Joined: 24 Dec 2006, 18:56
Posts: 764
With a quick overview, i can't see the solution. Maybe i could dig up some old math program :)


Top
 Profile  
 
PostPosted: 10 Dec 2008, 23:10 
Offline
User avatar

Joined: 24 Dec 2006, 18:56
Posts: 764
Okay let's try it on the old school way.

First you have to find out how wide the 4:3 and 16:10 monitor are.
(we make the monitor with height 1)

4/3=1,3333
16/10=1,6

Then we have to know the half of those (working with triangles is bit handier)

1,3333/2=0,6666
1,6/2=0,8



Okay, now we have a triangle, with greenbase (see pic), let's calculate that out

Tan(45)=0,6666/greenbase => greenbase = 0,6666/tan(45) = 0,6666

Now for the new one

Tan(?)=0,8/0,6666 => ?=50,197

We had divided the fov with 2, so correct it with 2x
Fov 16:10 50,197*2= 100,394...



*Some rounding errors could occur, i could make it a bit more accurate if people want.
**Don't laugh my paint skills :lol:

[Edit]

MY FOV CALCULATION

2) with SAME FOV, EVERY screen shows the SAME HORIZONTAL FIELD, regardless of the ratio. The different ratio affects the VERTICAL FIELD.
(The Unreal Tournament Game Engines 1,2 and 3 works like that, sperimented by me many and many times)


This method you describe is a vert- method. you keep the horizontal fov the same, and vertical fov scales.

What you need is Hor+ what gives you more space.

In the MGL (Master Game List) you can see that unreal is hor+


Top
 Profile  
 
PostPosted: 10 Dec 2008, 23:44 
Offline
User avatar

Joined: 14 Nov 2006, 15:48
Posts: 2356
What he said^

And technically theres no correct way to determine FOV, Its purely a personal preference.


Top
 Profile  
 
PostPosted: 11 Dec 2008, 01:07 
Offline

Joined: 10 Dec 2008, 19:43
Posts: 8
mh, let's start from the bottom.

@ DAFOX:
I agree: It's a personal preference. Here are only mathematical time-wasting.
Maybe technically theres no correct way to determine FOV.
But i compare my calculations with what I see on screen and SEEMS correct.

@ P-STORM:
U calculate the fov in the same way was calculated in that link, assuming the same things (same identical vertical dimensions)
No surprise if the results are identical.

I Need to know why MY CALC is different.
In my opinion ratio AND dimension are important.
a 19'' and a 42'' both @16:10 cannot be set at the same fov.
If so, U must calculate and change the distance between eyes and panel, or you play like with binocular in your eyes with the 42''.

I describe the "vert-" method because Unreal Engine use a vert- method.
There's an error in the Master Game List.


The proofs:




Every image was rescaled at 50% both axes.
U could find that the horizontal field is the same for all the pics (take a look at the "angles" shown in the 1st pic)

These pics "talk" only about resolution and ratio, not about dimensions.
Thats why I measured with a ruler all my calcs on all the monitors I owned:
17'' @ 4:3 CRT
19'' @ 4:3 CRT
(remember: a CRT loose al least one inch of diagonal. LCD not)
19'' 16:10 LCD (1440x900 native)
22'' 16:10 LCD (1650x1060 native)
26'' 16:9.6 LCD (1280x768 native)
SAME results in every cases.

The couterpart:
Try to take a 19''@4:3 and a 19''@16:10 side to side.
the 1st with FOV90°, the 2nd with 103, like P-STORM said.
U'll see "innatural" differences and U realize immediately that something is wrong.

That's why I'm here.
Thx.


Top
 Profile  
 
PostPosted: 13 Dec 2008, 03:54 
Offline
Editors
Editors
User avatar

Joined: 10 Jun 2005, 21:24
Posts: 1371
:welcome

Its an interesting subject you are bringing up. :)

The reason why your calculations are different, is probably because your constants differs from WSGF. Our goals are different in the calculations.

Read this:
http://www.widescreengamingforum.com/wiki/index.php/FAQ#Screenchange

Horizontal+ while vert stays constant is the optimal screenchange. Visually, HL2 describes it best. Check the mouseover function here:
http://www.widescreengamingforum.com/screenshots/hl2.php

Our FOV calculations are not based upon a constant distance to the screen, but rather a constant vertical keeping the 4:3 FOV intact and just adding to the side upon change of aspect ratio.

This means we have a different starting point then your calculations. Since our FOV solutions are based upon the game developers choice of FOV in 4:3, distance to the screen might be different from game to game. Not everyone is comfortable with the default FOV in 4:3, so not everyone chooses the same in widescreen.

In general, most people have a distance to screen chosen from when they can read text comfortably. This means its impossible to operate with a constant distance that works for all. Especially considering the difference in size of screen as well. The FOV calculator is more as a starting point then a "one fit all" solution.

Considering that the human eye also have a flexible focal length the optimal distance is not a single number, but a range. :)


Top
 Profile  
 
PostPosted: 13 Dec 2008, 19:00 
Offline
User avatar

Joined: 24 Dec 2006, 18:56
Posts: 764


@ P-STORM:
U calculate the fov in the same way was calculated in that link, assuming the same things (same identical vertical dimensions)
No surprise if the results are identical.

I Need to know why MY CALC is different.
In my opinion ratio AND dimension are important.
a 19'' and a 42'' both @16:10 cannot be set at the same fov.
If so, U must calculate and change the distance between eyes and panel, or you play like with binocular in your eyes with the 42''.


That's why I'm here.
Thx.


First thing what i want to say, i don't hate you for calculating different or have other idea's. So if i sound like that, sorry for that.

the thing you say is that if you have a 42'' then the fov is higher. The thing in my eyes is that you can't lock a fov then. See it as a paiting. If the size is bigger, the the horizontal FoV en vertical FoV will both increase. And this is a thing why the calculations are so shaky, because you have to take account for 2 FoVs.

And yes to fix that binocular effect you have to sit futher back, but that the calculations don't like. Maybe i could make a explained fov part 2 with this in it :)


Top
 Profile  
 
PostPosted: 13 Dec 2008, 19:26 
Offline

Joined: 10 Dec 2008, 19:43
Posts: 8
First thing what i want to say, i don't hate you for calculating different or have other idea's. So if i sound like that, sorry for that.

LOL! of course not! Forgive me if MY posts isn't so clear. (bad english)
I agree with U about the 42'' problems... but this is a extreme case.

Tamlin says all I need to know: the start point (the goal) is different - the results may be different.
The choice is:
- to preserve the original V-view of a 4:3 game;
- to preserve the original dimensions of a rendered object.

I DO prefer the 2nd. I'ts an opinion.

However I know that, like many people do, the goal is to find a good choice to enjoy the game in widescreen (regardless of calculations).
As I said, my post is only "mathematical time-wasting".

Thanks for your answers.

EDIT:
Since I notice and sperimented well that's an error in MGL about Unreal Tournament... (It's a Vert-, not a Hor+ game) can U fix It, Tamlin?


Top
 Profile  
 
PostPosted: 13 Dec 2008, 20:18 
Offline
User avatar

Joined: 24 Dec 2006, 18:56
Posts: 764
Ah I remember the [url=http://www.widescreengamingforum.com/node/6404 report[/url]
The thing is, that you can set the horizontal fov, and then you can set it to hor+. That is why it is hor+, if you can set it hor+, then it is hor+.

(allmost all the insiders can edit the wiki if it's needed)


Top
 Profile  
 
PostPosted: 13 Dec 2008, 20:46 
Offline
Editors
Editors
User avatar

Joined: 10 Jun 2005, 21:24
Posts: 1371

Tamlin says all I need to know: the start point (the goal) is different - the results may be different.
The choice is:
- to preserve the original V-view of a 4:3 game;
- to preserve the original dimensions of a rendered object.

I DO prefer the 2nd. I'ts an opinion.

However I know that, like many people do, the goal is to find a good choice to enjoy the game in widescreen (regardless of calculations).
As I said, my post is only "mathematical time-wasting".

Thanks for your answers.

EDIT:
Since I notice and sperimented well that's an error in MGL about Unreal Tournament... (It's a Vert-, not a Hor+ game) can U fix It, Tamlin?


Its not a mathematical time-wasting. :) I'm all for giving people options.

Your method has the advantage of adjusting from viewing distance. Ours are based that viewing distance is the same and that they have done the calculations when creating the game. Going 4:3 to 16:10, our solutions are about having the same distance also after change. Objects should have the orginal dimensions, but with added content only then.

As for UT, please check our Detailed report and respond there if you discover some errors. According to Cranky, the solution is based upon an .ini hack, so settings depends if its Hor+ or not.
http://www.widescreengamingforum.com/node/6283

As P-Storm correctly pointed out, almost all of the insiders can edit the wiki if needed. I even think he would be better at this then I. But, first it should be run through Cranky via the above link who created the detailed report. :)


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 11 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  




Powered by phpBB® Forum Software © phpBB Group