Previous Thread
Next Thread
Print Thread
Rate This Thread
Hop To
Page 4 of 7 1 2 3 4 5 6 7
#2401700 - 12/15/07 03:51 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Rama]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Rama
To put an end with the issues of LOD/relative sizes/supposed wrong scale.... at least for the Bf109F serie.

I just did a small experiment everybody can do to check the accuracy of the Bf109F size in game.
I did a short mission with the mission editor, choosing a straight road on a flat area, putting a P39 to look (player plane) facing BF109. All are flyable planes with 2 waypoints (to place them in the correct direction), first one with TAKEOFF waypoint type.
The corresponding mission is then edited to have for each plane: 0 fuel, Altitude 0, speed 0 (so they can stay in place without moving when mission launch).
The exact positionning is also corrected on the mission editor to have planes at exactly 100m, 200m, 300m, 500m, 750m, 1000m, 2000m, 3000m and 5000m

I then launched the missions, choose the FoV angle at exactly 30° (and not visor view) and take screenshots.
I could measure the wingspan on the first screenshot for 100m to 300m.... for the other distances, I had to eliminate the closest Bf109 in the mission to be able to see them and take screenshots.

First screenshot look like this:


I spare you all the other ones... everybody can redo what I did.
As an example, here how the mission looks like when only one Bf109F stays (the one at 5000m):

----------------------------------------------------------------------------------
[MAIN]
MAP Net2Summer/load.ini
TIME 12.0
CloudType 0
CloudHeight 1000.0
army 1
playerNum 0
[Wing]
r0100
g0103
[r0100]
Planes 1
Skill 1
Class air.P_39N
Fuel 0
weapons default
[r0100_Way]
TAKEOFF 48000.00 10900.00 0.00 0.00 &0
NORMFLY 47999.90 10900.00 0.00 0.00 &0
[g0103]
Planes 1
Skill 1
Class air.BF_109F4
Fuel 0
weapons default
[g0103_Way]
TAKEOFF 43000.00 10900.00 0.00 0.00 &0
NORMFLY 43000.10 10900.00 0.00 0.00 &0
[NStationary]
[Buildings]
[Bridge]
[House]
----------------------------------------------------------------------------------

Anybody can also build by hand the other missions by changing the red values.

Now the numbers:
What I find by measuring the wingspan for each distance (with The Gimp)
100m -> 313 pixels
200m -> 156 pixels
300m -> 104 pixels
500m -> 63 pixels
750m -> 43 pixels
1000m -> 32 pixels
2000m -> 16 pixels
3000m -> 11 pixels
5000m -> 6 pixels

All these of course with an accuracy of +/-1 pixel.

What can be immediatelly found: is that the ratios for the different distances are quite consistant (thales theorem)... if I take a ratio of 1 for 100m, I find.
200m -> 1.003
300m -> 0.997
500m -> 1.006
750m -> 1.030
1000m -> 1.022
2000m -> 1.022
3000m -> 1.054
5000m -> 0.95
with less than 5% "error" everywhere (and less than 1% at 500m and below), I think it's quite a good demonstration that there's no "increasing" and "shrinking" when changing LOD. The size change with distance is continuous and consistant.

Now about the size of the Bf109F wingspan.
.... no need to have any concrete bloc nead... we have everything we need.
A FoV of 30°, a monitor width of 1680 and a size of 313 pixels at 100m exact distance.

With an easy calculation (I'm sparing you also...), you find... 9.98m exactly....

Someone said there's an issue with Bf109 wingspan size?
Well... if anybody has a an issue with the wingspan size of any aircraft of the game..... no need to complain first wrongly.... you can verify with the method I just gave... and complain after if you find something REALLY wrong

I end with a screenshot of the mission editor for this small verification.



Hope this helps...

EDIT in red



except of course that you have just proven my point !

you have incorrectly used a 30 FoV for your measurements of those 109 on-screen wingspans, and this has artificially magnified what you see on your screen. you stated previously you were using a 22' monitor, so really your FoV setting for "normal" should have been something like 50 or 55 FoV, not 30 ! (presuming you are sitting about 75 cm from your screen, which would be about the correct distance for normal viewing with a monitor at that resolution and size)

the 30 FoV is a zoom feature which artificially serves as a magnifier (as clearly stated by oleg), it is not a 1:1 Fov like the 35 to 90 FoV's that you can select depending on your monitor size, so you never should have used the 30 FoV, no matter what size screen you use.

the list of measurements you produced would suggest the 109 is in fact about 25% to small on screen, since you should have used about a 55 FoV setting which would have significantly shrunk the LoD models you just measured.

be aware of the bored laptop armed executive on a long trans continental flight, aircraft have net access these days \:D

i think i'll just order another scotch now and a few more of those nice packets of roast cashews \:\)



Be advised, we got zips in the wire
Inline advert (2nd and 3rd post)

#2401974 - 12/15/07 11:38 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Nov 2005
Posts: 679
Rama Offline
Member
Rama  Offline
Member

Joined: Nov 2005
Posts: 679
Toulouse France
No, you don't understand anything
I displayed at 30° FOV and the display is correct for a 30° FOV, and if I display at 55° FOV, it would also be correct for a 55° FOV (I did the same experience). This is totally independant from the distance of the viewer to the screen. When the graphical calculate the size of planes according to FOV (choosen by the player), it only takes in account this FOW and the size (width) of the monitor in pixels... which is of course totally normal.
Mathematically, the display is perfectly correct. All the rest is fantasm talk.

But of course, as you prooved many time, you're totally unable to understand simple mathematical facts.



Last edited by Rama; 12/16/07 01:15 AM.
#2402098 - 12/16/07 02:43 AM Re: scale and proportional model size issues: p/a SaQSoN [Re: Rama]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Rama

I displayed at 30° FOV and the display is correct for a 30° FOV, and if I display at 55° FOV, it would also be correct for a 55° FOV (I did the same experience). This is totally independant from the distance of the viewer to the screen. When the graphical calculate the size of planes according to FOV (choosen by the player), it only takes in account this FOW and the size (width) of the monitor in pixels... which is of course totally normal.
Mathematically, the display is perfectly correct. All the rest is fantasm talk.


you didnt read correctly what i posted, or you have a problem with understanding the english language.

try and focus on this part, you had previously posted this:

 Originally Posted By: Rama
Now the numbers:
What I find by measuring the wingspan for each distance (with The Gimp)
100m -> 313 pixels
200m -> 156 pixels
300m -> 104 pixels
500m -> 63 pixels
750m -> 43 pixels
1000m -> 32 pixels
2000m -> 16 pixels
3000m -> 11 pixels
5000m -> 6 pixels

Now about the size of the Bf109F wingspan.
.... no need to have any concrete bloc nead... we have everything we need.
A FoV of 30°, a monitor width of 1680 and a size of 313 pixels at 100m exact distance.


and that is the part i have a problem with !

you previously stated you were using a 20' apple lcd (when you wernt inventing new apple monitor models, but lets skip that part for now), or in another thread you gave a different story and said you were using a 22' wide lcd, so the 1680 wide pixels for your monitor you mention fits in fine with either that 20' or 22' sized lcd, but...

to do any of your pixel counts from screen shots you should have set your FoV to about 55, not the 30 you just used ! this will then shrink the Lod models significantly down in size (for your 20' mac monitor the pixel size is 0.258 mm, for a 22' wide lcd it is 0.285 mm ), and you should then be able to confirm what a number of people have said all along, when you use the correct FoV for your monitor size the wingspan for that il2 109 displayed on screen is smaller in size than it should be seen as in real life.

comprende ?

and unless my high skool memories are letting me down:
at 100m a 10 m wide 109 should be seen as 10 cm, not the 313 pixels x 0.258 mm = 8.07 cm you give
at 1000 m a 10 m wide 109 should be seen as 1 cm, not the 32 pixels x 0.258 mm = 0.825 cm you give

because you seem to come up with different stories about what monitor you use for il2 in different threads, the above measurement was in case you were using a 20' wide imac monitor. if you were using a 22' wide lcd like you had said in another thread here before, then you still have a similar problem.
at 100m a 10 m wide 109 should be seen as 10 cm, not the 313 pixels x 0.285 mm = 8.92 cm you give
at 1000 m a 10 m wide 109 should be seen as 1 cm, not the 32 pixels x 0.285 mm = 0.912 cm you give


and those results you gave will be even smaller if you would use the correct 55 FoV for your 22' lcd or a 50 FoV for a 20'lcd

a completely different point you were trying to illustrate is that there should be a linearly scale in which the LoD models reduce in size from 100 to 5000 meters. that is a different issue which we can discuss later.

 Originally Posted By: Rama
What can be immediatelly found: is that the ratios for the different distances are quite consistant (thales theorem)... if I take a ratio of 1 for 100m, I find.
200m -> 1.003
300m -> 0.997
500m -> 1.006
750m -> 1.030
1000m -> 1.022
2000m -> 1.022
3000m -> 1.054
5000m -> 0.95
with less than 5% "error" everywhere (and less than 1% at 500m and below), I think it's quite a good demonstration that there's no "increasing" and "shrinking" when changing LOD. The size change with distance is continuous and consistant.




Be advised, we got zips in the wire
#2402285 - 12/16/07 02:28 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Sep 2004
Posts: 10,618
Ming_EAF19 Offline
Babelfish Immune
Ming_EAF19  Offline
Babelfish Immune
Veteran

Joined: Sep 2004
Posts: 10,618
London
Counting pixels as units of measurement may be misleading. All sorts of functions transform pixels, LOD and so on. If the concrete block object is released things may be simpler but would you want someone measuring the real size of anything you might claim to be a certain size and telling all the flight-sim world about it? Or an uncertain size in some cases. Happy Christmas, mine's a large one check them pixels

Ming


'You are either a hater or you are not' Roman Halter
#2402321 - 12/16/07 04:10 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Ming_EAF19]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Ming_EAF19
Counting pixels as units of measurement may be misleading. All sorts of functions transform pixels, LOD and so on. If the concrete block object is released things may be simpler but would you want someone measuring the real size of anything you might claim to be a certain size and telling all the flight-sim world about it? Or an uncertain size in some cases. Happy Christmas, mine's a large one check them pixels
Ming


i dont really care about one pixel or the other

what does bug me however is that when i look out of my cockpit in il2 most aircraft i look at seem smaller then they should be, and that does matter when you are trying to shoot at them or hit a specific part of the target. same problem when you try and hit trucks or tanks as ground targets, i have a feeling i am hunting for dinky toys and can see diddly squat compared to what real ww2 pilots reported in the many books i have read of that period. it also doesnt equal what i can see from a real aircraft, great or small.

in theory in il2 it should just be a matter of setting your FoV correctly for your monitor size, and all should be well, but even then many objects still look to small to me. the only way to settle this question is to measure these distant aircraft, and for that i am afraid we are down to counting pixels

i think my observation in that last post is correct. rama made his measurements with a FoV setting of 30, which is an artificial zoom feature which magnifies everything you look at. he should have used 50 or 55 for that monitor size instead, and with his pixel counting he has exactly confirmed what i have been saying all along, that the 109 at those various distances is indeed significantly to small.


Be advised, we got zips in the wire
#2402327 - 12/16/07 04:17 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Ming_EAF19]  
Joined: Nov 2005
Posts: 679
Rama Offline
Member
Rama  Offline
Member

Joined: Nov 2005
Posts: 679
Toulouse France
Ming: The Sim compute the size in pixels from the FOV, the distance of the plane to the virtual camera, and the size of the object. If you know the 2 later, you can calculate the first (size in pixels) and compare to qwhat you see on the screen.
This is the only wat to show if screen size are correct or incorrect.

Rootango: Again (and for the last time) you don't understand geometry and mathematics. You can make sentences as long as ou can, include as many good words in it as you want ... it wont replace a formula.
I'm sure you have a lot of other competences... it would be better if you focus on what you know and understand.
For the last time (but you probably wont understand better... I just to it in case a miracle gives you some simple geometry comprehension...)...
---- SIZE OF THE MONITOR IN INCHES HAS NOTHING TO DO WITH THE CORRECT DISPLAY OF THE OBJETCS OF THE SIM, ONLY THE SIZE IN PIXELS HAS TO BE TAKEN IN ACCOUNT (SO 1680 IN MY CASE). THIS IS THE ONLY THING THE GRAPHICAL ENGINE KNOWS, ALONG WITH FOV AND OBJETCS REAL SIZE. THE GRAPHICAL ENGINE DON'T KNOW ANYTHING ABOUT YOUR MONITOR SIDE IN INCHES OR DOT PITCH SIZE AND DON'T NEED IT ----


With FOV, real object size and monitor width in pixels, you get all what you need for correct geometrical calculations (with very simple formula using Sinus, Cosinus, etc...)
Got it?

Sure no... but I won't insist. What you need is basic mathematical and geometrical courses - In any case you wont accept them since you're so sure to be right.

Take your last word. Discussion is over for me... as it was 4 month ago.

#2402482 - 12/16/07 10:30 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Rama]  
Joined: Sep 2004
Posts: 10,618
Ming_EAF19 Offline
Babelfish Immune
Ming_EAF19  Offline
Babelfish Immune
Veteran

Joined: Sep 2004
Posts: 10,618
London
The concrete block in the shot has a number of pixels making up one metre

Four times as far away one would expect four times less pixel count but-

Pixels get messed around with by the renderer

I use pixels all the time to measure things Rama. But they can only be used reliably when they are orthogonal shots. Perspective makes things less reliable

I only mention this because I see you using words like 'FOV'. There is no field of view parameter for a flat bitmap. Only flat bitmaps can reliably be used for making measurements: orthogonal 2D images. This can be done for example from the FMB in zoomed-in overhead view. Once you move into the 3D viewport and bring in perspective errors the rules change

 Quote:
For the last time (but you probably wont understand better... I just to it in case a miracle gives you some simple geometry comprehension...)...
---- SIZE OF THE MONITOR IN INCHES HAS NOTHING TO DO WITH THE CORRECT DISPLAY OF THE OBJETCS OF THE SIM, ONLY THE SIZE IN PIXELS HAS TO BE TAKEN IN ACCOUNT (SO 1680 IN MY CASE). THIS IS THE ONLY THING THE GRAPHICAL ENGINE KNOWS, ALONG WITH FOV AND OBJETCS REAL SIZE. THE GRAPHICAL ENGINE DON'T KNOW ANYTHING ABOUT YOUR MONITOR SIDE IN INCHES OR DOT PITCH SIZE AND DON'T NEED IT ----


I think you mean-

 Quote:
A BLOCK MEASURING FIFTY PIXELS ON MY MONITOR MAY BE ONLY TWENTY-FIVE PIXELS ON YOUR MONITOR. BUT ALL OTHER PIXEL MEASUREMENTS ON YOUR MONITOR CAN BE SCALED BY A FACTOR OF TWO TO MATCH MY PIXEL READINGS. MY TWO HUNDRED PIXELS ARE YOUR ONE HUNDRED PIXELS IS ANY OF THIS GETTING THROUGH TO YOU THEY ARE RELATIVE, PROPORTIONALLY-SCALED READINGS. YOU SAY POTATO I SAY POTATO OVER TWO. STOP FAKING YOU'VE DOZED OFF I CAN SEE YOUR EYES MOVING UNDER THE LIDS


Ming


'You are either a hater or you are not' Roman Halter
#2402535 - 12/17/07 12:16 AM Re: scale and proportional model size issues: p/a SaQSoN [Re: Ming_EAF19]  
Joined: Nov 2005
Posts: 679
Rama Offline
Member
Rama  Offline
Member

Joined: Nov 2005
Posts: 679
Toulouse France
Ming:
I'm not talking at all about the concrete block, I talk about the Bf109 wingspan measurement at different distance done by me (without help of teh concrete block) and quoted by rootango..... this has absolutly NOTHING to do with the concrete block.
When making the measurement, I made sure while placing the BF109 on the map road that the wings are totally ortogonal to the direction plane->camera.
The measurement is correctly done. I also use pixels for measurement everyday, I'm a map engineer and photogrametry is my background.
Read the part quoted by Rootango on 15/12, and you'll understand.

#2402602 - 12/17/07 03:40 AM Re: scale and proportional model size issues: p/a SaQSoN [Re: Rama]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Rama

Rootango: Again (and for the last time) you don't understand geometry and mathematics. You can make sentences as long as ou can, include as many good words in it as you want ... it wont replace a formula.
I'm sure you have a lot of other competences... it would be better if you focus on what you know and understand.
For the last time (but you probably wont understand better... I just to it in case a miracle gives you some simple geometry comprehension...)...
---- SIZE OF THE MONITOR IN INCHES HAS NOTHING TO DO WITH THE CORRECT DISPLAY OF THE OBJETCS OF THE SIM, ONLY THE SIZE IN PIXELS HAS TO BE TAKEN IN ACCOUNT (SO 1680 IN MY CASE). THIS IS THE ONLY THING THE GRAPHICAL ENGINE KNOWS, ALONG WITH FOV AND OBJETCS REAL SIZE. THE GRAPHICAL ENGINE DON'T KNOW ANYTHING ABOUT YOUR MONITOR SIDE IN INCHES OR DOT PITCH SIZE AND DON'T NEED IT ----

With FOV, real object size and monitor width in pixels, you get all what you need for correct geometrical calculations (with very simple formula using Sinus, Cosinus, etc...)
Got it?

Sure no... but I won't insist. What you need is basic mathematical and geometrical courses - In any case you wont accept them since you're so sure to be right.



bollocks and poppycock !

your doing lots of blablabla yourself there but not giving anything that refutes me observing that the previous measurements you posted were indeed confirming that the 109 is displayed to small in il2.

if it looks like a duck, quacks like a duck, and waddles like a duck, it probably is a duck.

there is only 3 main issues that matter here afaics to determine if the 109 being displayed on screen is indeed the correct size, and it needs to be seen onscreen exactly the same size as it would be seen as in real life !

1) whatever monitor size you use in il2, and how wide that monitor is and the distance it is from you will determine how much of your field of view it occupies, and that determines what FoV you set in il2 as "normal" . using a smaller FoV in il2 will then magnify objects, and using a wider FoV will zoom out of objects and shrink them.
- the correct "normal" FoV for your monitor can be calculated by anybody, its simple and you know that method to.
2) we know how large a 10 meter wide object looks in real life when it is seen from 100, 200, 500, .... 1000 meters etc. and this is determined by a very simple formula
3) in il2 you can take a screenshot to represent what you see from the cockpit, so you can then measure the size of the object correctly and others can confirm those measurements. the pixel measurements you posted were taken with a 109 directly and squarely in front of you, so there is no distortion argument about them being seen at an angle. the only reason pixel size matters is because when you look at the il2 objects and scenery with the correct "normal view", then knowing the pixel size of your monitor (and in lcd's we presume you are using the native resolution), allows an exact measurement of the object.

you also failed to recognize that even if a 19', 20' or 22' wide lcd all use the same resolution, and they might use a slightly different pixel size, but that doesnt really matter because for our purpose each of those monitors would be set at a slightly different "normal FoV" in il2 ( because they occupy a slightly different % of your 180 degree forward view). the whole issue about different pixel sizes only matters because it gives us the accuracy of the measurement and error rate if we are measuring an object that is made up of about 30 or 300 pixels. i am quite able to discuss this in a normal manner, all i am interested in is resolving the issue, you often seem in a bad mood and make out of the blue rude posts to people (on most forums i have seen you on).

lets focus on the simple facts here. the pixel measurements you posted for the 109 at those different distances is not only producing an incorrect size for the 109 compared to what it should be, it also artificially magnified the 109 you are looking at because you used a more zoomed view on your monitor when you took the screenshot, and this was even the biggest zoom view possible in the sim and this then added a further artificial magnification (rather than simply changing the FoV itself). even so, it still shows the 109 as being slightly to small.

you didnt answer that argument, did you ?


the way to solve it is to do the same measurements with your monitor (since you are claiming to be the expert in this), but now set to the correct FoV (i suspect that is about 50 or 55 for you), and like even you can predict that will show the 109 to be even smaller, which is why i suspect you dont want to do it.

within a very small margin of error we can figure most of this out. and we dont have the same problems as with the measurement of the concrete block,truck and 109 from the saqson pictures, which used screenshots taken in external view (which distort thing we see further) and were taken at a slight angle.

measuring these 109's is a much more accurate method because we know the exact size the aircraft should be, and the distance we are from it. right now it looks by your own post and pixel count that you exactly confirmed what i said right at the start of this thread, some of the objects we see onscreen in il2 are just displayed to small (in partic aircraft in the sky). the only exception is for people with very large monitors, like a 32 or 37 inch TV screen where everything is stretched out even when they use the the max 90 FoV.



Be advised, we got zips in the wire
#2402751 - 12/17/07 01:29 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Nov 2005
Posts: 679
Rama Offline
Member
Rama  Offline
Member

Joined: Nov 2005
Posts: 679
Toulouse France
Only one answer:

Dpix: Display size in pixel on the screen
OBjw: Object Width (in meter)
Dist: Distance between object and camera (in meter)
Mon: Monitor Width in pixels
FOV... you're supposed to now what it is (in degree)

Thales: D/Dist = Dpix/OBJw => D = Dpix*Dist/Objw
tgt(FOV/2) = (Mon/2)/D => D = Mon/2*tgt(FOV/2)

=> Dpix*Dist/Objw = Mon/2*tgt(FOV/2)

=> Dpix = Mon*Objw/2*Dist*tgt(FOV/2)

You get everything you need for the measurement. This is a mathematical fact, and only accurate method to verify is size of the objects in graphical engine is correct or not... this is not blabla.




Everything else is talking about the sex of the angels.

#2402810 - 12/17/07 02:46 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Rama]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere

so in your Dpix = Mon*Objw/2*Dist*tgt(FOV/2), is * a multiplication symbol ?



you just used that same method earlier to produce your own measurements, so why are they then producing a measurement for the 109 that is to small ?

i dont see a way around that, and i have done similar measurements on my own monitor that indicate the same thing.

my conclusion is that even if you calculate the correct FoV to use for your monitor (or at least for the ones i used being 19, 20 and 27') the enemy planes like 109's are still displayed to small on screen.


Be advised, we got zips in the wire
#2402815 - 12/17/07 03:09 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Mar 2001
Posts: 1,089
Tree Offline
Member
Tree  Offline
Member

Joined: Mar 2001
Posts: 1,089
Derby, England
The scale of all objects in Il2 is 100% correct, its just our lying eyes!! :-)

#2402823 - 12/17/07 03:25 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Nov 2005
Posts: 679
Rama Offline
Member
Rama  Offline
Member

Joined: Nov 2005
Posts: 679
Toulouse France
 Originally Posted By: rootango

so in your Dpix = Mon*Objw/2*Dist*tgt(FOV/2), is * a multiplication symbol ?

Yes

 Originally Posted By: rootango
you just used that same method earlier to produce your own measurements, so why are they then producing a measurement for the 109 that is to small ?

They are not - check the calculation again

 Originally Posted By: rootango
my conclusion is that even if you calculate the correct FoV to use for your monitor (or at least for the ones i used being 19, 20 and 27') the enemy planes like 109's are still displayed to small on screen.

Where do you see the monitor size in the formula?
Again, the formula is independant from monitor size, as it should be.

 Originally Posted By: rootango
the enemy planes like 109's are still displayed to small on screen.

You can repeat that ad nauseam without any proof... it wont be true
The formula and the measurements are correct... period.

But I do feel stupid to repeat again and again the same obvious ultra-simple mathematical fact.

@Tree: The measurement are done are for Bf109 (no time to do it for all IL2 planes). For other objects, measurement are to be done the same way. I don't pretend ALL objects in IL2 are 100% correct (and they aren't... as an example, most buildings are probably too small)

#2402832 - 12/17/07 03:39 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Tree]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Tree
The scale of all objects in Il2 is 100% correct, its just our lying eyes!! :-)


its already been officially confirmed what other users including myself had been reporting for a long time, many objects are the wrong size ;\)

 Quote:
Originally posted by SaQSoN:
The smallest axis armor in the game is Pz.IIF.

In the game it is about 2m wide and 4m long. While it should be 2.220m wide and 4.800m long. Which means, it is about 17% smaller, then should be.
Other axis armor has size variation relatively to the correct size within 5% margin.


 Quote:
Originally posted by SaQSoN:
One IJA and one USN fighters are about 20% larger then should be (subtypes not counted).
Other planes have correct scale.


 Quote:
Originally posted by SaQSoN:Many IL-2 buildings were modelled too small, something like 50-70% of their real(-istic) size.

IL-2 pilot models are about 15-20% larger, then the real average human (i.e., they are about 2-2.1m tall) ;\)


etc... there are other examples.

my main complaint is that when i chase an enemy 109, and it is 300 meters or so ahead of me, it looks smaller than what it should, even when i use the correct normal FoV. so far all indications are that the onscreen measurements indeed confirm this. if i am wrong i'll be happy to accept that, i just dont think that is the case.

i suspect this whole issue is a residue from the original il2 design, when we only had 3 FoV's, normal, wide and zoom. those were the only 3 options you could use, no matter what monitor you had, but that is about 7 years ago when most of us might have had a 17'crt monitor. since FB we have the graded FoV's from 30 to 90, and with 5% accuracy you can select your "normal" FoV for your monitor size.

my argument is that you need to set the normal FoV about 10 degrees smaller than the "normal" view as calculated by the % of view method, and only then do distant aircraft look roughly right in size (ie for my 27' lcd i should use 65 FoV as normal, but only by setting it to 55 FoV do i see those distant aircraft correctly in size).





Be advised, we got zips in the wire
#2404138 - 12/19/07 11:39 AM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Dec 2002
Posts: 19,381
Ajay Offline
newbie
Ajay  Offline
newbie
Veteran

Joined: Dec 2002
Posts: 19,381
Brisbane OZ
i want the funny tasting pomegranites please , in correct size with no orthogonal sizing.


My il2 page
Seelowe Campaign
Cliffs of Dover page
CloD
My Models
Tanks/Planes/Ships


#2404167 - 12/19/07 01:06 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Mar 2003
Posts: 1,209
Dozer Offline
Member
Dozer  Offline
Member

Joined: Mar 2003
Posts: 1,209
 Originally Posted By: rootango

its already been officially confirmed what other users including myself had been reporting for a long time, many objects are the wrong size ;\)

Yes some are wrong. But not the 109, as has been comprehensively proven using MATHS and SCIENCE and the FAMILY VON TRAPP.

 Originally Posted By: rootango
my main complaint is that when i chase an enemy 109, and it is 300 meters or so ahead of me, it looks smaller than what it should, even when i use the correct normal FoV. so far all indications are that the onscreen measurements indeed confirm this. if i am wrong i'll be happy to accept that, i just dont think that is the case.


Peace! Peace in our time! I have a solution:

SIT CLOSER TO THE SCREEN!!

Then all the objects will appear larger to your eyes!

Or everyone can accept that we're not sitting in an aircraft when playing a sim, we're sitting at a desk watching a screen fed by a camera with a zoom lense mounted on a gimbal in the cockpit of the aircraft, roughly where the pilot's head should be.


Dozer
#2404216 - 12/19/07 02:20 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: Dozer]  
Joined: Dec 2000
Posts: 831
rootango Offline
Member
rootango  Offline
Member

Joined: Dec 2000
Posts: 831
everywhere
 Originally Posted By: Dozer_EAF(T)
... everyone can accept that we're not sitting in an aircraft when playing a sim, we're sitting at a desk watching a screen fed by a camera with a zoom lense mounted on a gimbal in the cockpit of the aircraft, roughly where the pilot's head should be.


hey your happy, good for you. does that mean we should now shut the rest of the universe down maybe, would that be more convenient for you ?

to me it is import to see the aircraft around me in their correct sizes, so you can judge distance to them correctly when you take a shot at them. i also want to see the aircraft around me correctly at various distances during combat, the same way they would be visible with the naked eye from a normal cockpit, so i can maintain normal SA.

those are the 2 critical parts about simulating flight on a pc, and if what i see onscreen in il2 doesnt correspond to what it should be, then i'll try and figure out why. it would also take very little effort in my opinion to improve those factors.


Be advised, we got zips in the wire
#2404227 - 12/19/07 02:37 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Sep 2001
Posts: 24,712
Dart Offline
Measured in Llamathrusts
Dart  Offline
Measured in Llamathrusts
Lifer

Joined: Sep 2001
Posts: 24,712
Alabaster, AL USA
I doubt we'll ever get a 1:1 representation of size in a sim, rootango.

One can accomplish the feat with dedicated simulators with code written specifically for the monitors and distance from the virtual pilot to them, but not with entertainment software.

Too many variables, from monitor size to how far one has it placed from the seat.

Judging distance is in some degree a matter of learning the simulation. When an aircraft is in the front of the aircraft, all bets are off, as we have the sight to work as a scale.

And in relationship to each other, the scale of objects in the IL-2 series is incredibly good. Not perfect with a few, but for common aircraft bang on the money.


The opinions of this poster are largely based on facts and portray a possible version of the actual events.

More dumb stuff at http://www.darts-page.com

From Laser:
"The forum is the place where combat (real time) flight simulator fans come to play turn based strategy combat."
#2404257 - 12/19/07 03:37 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Mar 2003
Posts: 1,209
Dozer Offline
Member
Dozer  Offline
Member

Joined: Mar 2003
Posts: 1,209
 Originally Posted By: rootango
 Originally Posted By: Dozer_EAF(T)
... everyone can accept that we're not sitting in an aircraft when playing a sim, we're sitting at a desk watching a screen fed by a camera with a zoom lense mounted on a gimbal in the cockpit of the aircraft, roughly where the pilot's head should be.


hey your happy, good for you. does that mean we should now shut the rest of the universe down maybe, would that be more convenient for you ?
Thinking of universes, are you in a different one, where it's possible to use a different system to the one I've just described for the visual output of Il-2 (or any other desktop simulator)? \:\)
 Originally Posted By: rootango
to me it is import to see the aircraft around me in their correct sizes, so you can judge distance to them correctly when you take a shot at them. i also want to see the aircraft around me correctly at various distances during combat, the same way they would be visible with the naked eye from a normal cockpit, so i can maintain normal SA.

those are the 2 critical parts about simulating flight on a pc, and if what i see onscreen in il2 doesnt correspond to what it should be, then i'll try and figure out why. it would also take very little effort in my opinion to improve those factors.
I don't know what you're talking about I'm afraid. I can roughly judge the distance between myself and the target. Thanks to Dart's excellent video I know how to judge distances much more accurately when the target's in the gunsight ring. I could maintain SA with the hat-switch, although it is much easier and more natural to use TrackIR instead.


Dozer
#2404280 - 12/19/07 04:22 PM Re: scale and proportional model size issues: p/a SaQSoN [Re: rootango]  
Joined: Jun 2001
Posts: 9,189
Osram Offline
Hotshot
Osram  Offline
Hotshot

Joined: Jun 2001
Posts: 9,189
Bielefeld, Germany
 Originally Posted By: rootango

i suspect this whole issue is a residue from the original il2 design, when we only had 3 FoV's, normal, wide and zoom. those were the only 3 options you could use, no matter what monitor you had, but that is about 7 years ago when most of us might have had a 17'crt monitor. since FB we have the graded FoV's from 30 to 90, and with 5% accuracy you can select your "normal" FoV for your monitor size.


You seem to think that a model in a modern game could be correct in one FoV and incorrect in another.
That is impossible unless you don't use fairly basic 3D features of the gfx cards (which is almost certainly untrue for Oleg) or the error is in the gfx hardware/software, in which case it would affect every 3D app/game and I am sure it would be known by now. The reason is simply that the effects of FoV are calculated by the T&L feature in the driver or hardware of the 3D card you use.

IOW; If the 109 has correct size in one FoV - and everything in this thread points to it - then it is correct at all FoVs.

Page 4 of 7 1 2 3 4 5 6 7

Moderated by  RacerGT 

Quick Search
Recent Articles
Support SimHQ

If you shop on Amazon use this Amazon link to support SimHQ
.
Social


Recent Topics
Carnival Cruise Ship Fire....... Again
by F4UDash4. 03/26/24 05:58 PM
Baltimore Bridge Collapse
by F4UDash4. 03/26/24 05:51 PM
The Oldest WWII Veterans
by F4UDash4. 03/24/24 09:21 PM
They got fired after this.
by Wigean. 03/20/24 08:19 PM
Grown ups joke time
by NoFlyBoy. 03/18/24 10:34 PM
Anyone Heard from Nimits?
by F4UDash4. 03/18/24 10:01 PM
RIP Gemini/Apollo astronaut Tom Stafford
by semmern. 03/18/24 02:14 PM
10 years after 3/8/2014
by NoFlyBoy. 03/17/24 10:25 AM
Copyright 1997-2016, SimHQ Inc. All Rights Reserved.

Powered by UBB.threads™ PHP Forum Software 7.6.0