Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TT]

This mod puts a new reticle for all 4.12 stock russian planes, including gunner positions. The reticle is from mangas I-16 repaint, so all credit goes to him.
This also fixes the problem of some reticles being "dim" and unrealistically looking. Editing gunsight glass mats was needed to fix the problem.





Here we see the same reticle texture with old gunsight glass mat and new gunsight glass mat.



Download

[greybeard]

IL-2 reflector gunsights would deserve a treatise.

While telescopic ones respect the obvious principle that to aim an object you, first of all, must see the object  , so they have very thin black reticles in a clear optical device, reflector gunsights try to "mask" target in every way, like by thick, dazzling reticles on a blurred, opaque glass  .

Mods often worsen this situation, inventing craft reticles like this:


and increasing exceedingly reticle brightness. Can't understand how to aim by such a gadget, especially at night.

Historical correctness, then, would open another chapter...

About Mangas arrangement, although glass is now more clear, I think reticle lines are still thick and luminosity excessive, Thanksgod, 4.12 allows to manage reticle brightness, but I think a color AMBER and not yellow should be the base, also to facilitate 4.10 and 4.09 users.

Looking at your first screenshot, frankly I can barely see the target in it, which is albeit very close and in full daylight, what if would be dusk and at longer distance from a smaller object like an aircraft?

[WxTech]

Indeed! I'm presently tweaking the MAT files for many modded gunsight reticles so as to *greatly* dim them down. And I've decided to leave out the extra layer which simulates the 'glow'. In reality, only if the reticle is *unnaturally* bright should such a glow be seen. Now, in day, dusk or night I can actually see my target. Yay!

Furthermore, without making up a new .tga image, I've discovered how to adjust both the scale and positioning of the reticle image so as to make the reticle size truly correct, if necessary.

By the way, this ol' dinosaur is still playing 4.08.

[Dreamk]

I'm not sure you're right - looking at some movies and pictures taken through authentic ww2 reticles, one can see that vision was indeed impeded by the reflected light lines - some being rather gross when the rheostat was pushed to maximum - these brought some optical companies to design reticles where the central part was "empty" - see for instance the reticles of the OPL and Baille-Lemaire. This was always a problem even before reflector sights - the well known ww1 and interwar tube sight Aldis and Chretien had also an "empty" central part in order to allow clear vision of the target.

[greybeard]

Don't forget we're speaking of reproduction.

In reality we've a stereoscopic sight with five times the screen resolution (about 500 dpi for human eye gainst the 96 dpi of a monitor). Also, WWII footage may be more or less degraded, not to mention it was taken thru' a limited optical too. You should watch personally an actual reflector gunsight to have a clear idea.

In the present day, we're so accustomed to look at reproduction devices (cinema, TV, monitor, smartphone, etc.) that we believe reality is as we see it in images. This took so root that people ask for optical defects to be modeled for realism!!! 

[WxTech]

I must point out that the eye's resolving power should be expressed as an angular measure, not linear. For most people this approaches 1 arcminute, or 1/60 (0.017) degree. Relating this to a monitor, we need to know the pixel array size and viewing distance.

For example, a 1920 pixel screen, to be seen just at the threshold of one's resolving the individual pixels, must subtend an angle of 1920 * 0.017 = 32 degrees. This angle is reached when the viewing distance equals the screen width times 1.74. For example, my 35" wide screen subtends 32 degrees if I sit away from it 35" * 1.74 = 61", or some 5 feet.

Interestingly, at the zoomed-in FOV of 30 degrees, when you are at the viewing distance where the screen subtends the same angle, you have a view which both matches the real life image scale and is seen with no apparent pixelation.

At closer distances, where the screen subtends more than some 30-ish degrees, then you can resolve the pixels.

Back to reticle brightness. A screen cannot replicate the huge range of scene brightness, and certainly not the astonishing dynamic range of our visual system. And so the on-screen representation tends to compress the brightness range and thus 'flatten' and compress the brightness range. Which is why it's fair to ease our lot by controlling the degree of interference imposed by a simulated reticle.

[greybeard]

Thanks! Interesting info! 

Can't understand following statement:
"Which is why it's fair to ease our lot by controlling the degree of interference imposed by a simulated reticle."

[Catahoulak9]

Wytech, I would love to see what another old dinosaur is up to. I believe your interpretation may be a lot more accurate than what's out there now.

Originally reticle "glow" was used to simulate a more lively, light based reticle instead of the stock IL2 image. I know because I did a lot of these in the early days. Greybeard is right, this new bright style reticle was based on images with scene averaged exposure and not the actual reticle brightness exposure, thus overexposing the reticle image making it appear much brighter than it actually was.


The exposure for this gunsite image is based on overall scene brightness (including background) and not the reticle brightness. This is what the camera sees and not the human eye. Unfortunately, this is what a lot of the original reticles were modeled after. Well mine anyway.

I'd love to sit in the cockpit of a P-40 and peer through an accurately set up gunsite to see what a pilot actually saw back then. Right now I'm afraid it's only a guess.

[WxTech]

Greybeard,
What I mean is that because of the limited dynamic range of a video display, it's fair to choose to 'under represent' reticle brightness. The on-screen representation more egregiously masks the target than t would in real life.

-  -  -  -

To be sure, the original reticle .tga files were uniformly flat and boring. Their chiseled edges and absolutely even, dark hue just did not look at all like they're generated by a light source. The current treatments, with a glow layer added, are *far* superior. It's just that I find the balance of brilliance from the game play perspective can have them toned down; but that's my preference. The great thing is that the user can get into the .mat files and exit things to taste. And even the .tga images can be re-worked with freeware Irfanview software.

Very briefly... For some reticles I've adjusted the scale so that on screen they do subtend the historical angular size. In the .mat file I've edited these lines, which most oftenresemble:

TextureCoordScale 0.0 0.0 1.0 1.0
ColorScale 0.8 1.0 0.4 1.0

The first line is the .tga image's X and Y offset, as a fraction of the graphic's size, and the horizontal and vertical scale factors. If the reticle dimension within the image and the allocated region centered on the gun sight's optical axis are in harmony, the values shown above will apply.

The second line are the R G B A values for colouring and controlling transparency. (The Shine parameter also affects transparency, in a fashion at least.)

Following is an example for a modified P-40 reticle from Non Wonder Dog a few years back. It was too large, and I wished to make it the historical 50 milliradians in diameter. I measure the reticle's on screen at the 30 degree FOV setting. I calculate the image scale at view center, based on the game's tangential projection. For a 1920 pixel wide screen, at the 30 degree FOV setting the central image scale is 28.8 pixels per degree (from which follows the scale in milliradians, there being 100 mils in 5.729 degrees.)

The altered scaling:

TextureCoordScale -0.095 -0.095 1.19 1.19

The image is dynamically reduced by the inverse of 1.19 (to make smaller, the value increases, as paradoxical as it might seem!). And the image center must be relocated by offsetting in both X and Y, here by -0.095. The latter number exactly equals 1/2 the difference between 1 and 1.19 (if the reticle was originally already nicely centered.)

The second altered line:

ColorScale 1.0 0.8 0.6 0.25

The accentuates the red color (R > G > B), and makes the transparency 25%, or much more see-through.

Incidentally, for the B25 tail gun I tweaked the reticle vertically (the Y axis) by suitably altering the second value in the TextureCoordScale line. Previously, the bullet stream immediately fell below the sight line, so I lowered the reticle (about 5 mils). I just took a guess in the fraction by which to adjust, and checked, repeating until the desired result was achieved.


By the way, I've only just got into this kind of stuff about a week ago, figuring it out by poking around.

[santobr]

Forgive me for my ignorance, but don't you think changing the scale of the reticle will make you calculate wrong the distance from the target.
I'm asking this because maybe the best option could be changing the FOV that would change the scale of the entire environment.
I put a customized FOV in my "end" key, so every time I hit the "end" key, it changes back to my customized FOV.



santobr.

[BravoFxTrt]

So, sorry for off topic, but WxTech, your a Weather Tech? In Military or where?

[WxTech]

Santobr,
Irrespective of the FOV chosen, the reticle has a fixed scale in the environment. If you pause while a plane is centered in the sight, note how both plane and reticle and gunsight always change scale exactly in step as you zoom in/out. Now imagine if the reticle graphic is not at the correct scale; instead of the Revi circle being the actual 100 mils in diameter, suppose it's 90 mils. Now you're underestimating the target distance by 10%. No matter what your FOV setting, that same error accrues.

One could create a new graphic, of the same pixel array (often 512X512 or 1024X1024), but with the reticle pattern made the required 10% larger. Or you could alter the.mat file, which the game reads and applies to that .tga file, resizing and positioning to suit.

Interestingly, if one reduces a reticle graphic via the .mat file beyond a certain point, one will see repeated edges appearing in the outer part of the gun sight, like a kind of 'cloverleaf'. I've gotten around that by adding more black space around all 4 sides of the reticle pattern in the image, so that the reticle remains centered in the graphic. But this is inelegant, for it results in a larger file size.

I'll probably get around to creating my own graphics from scratch, sized exactly based on my now divined requirements demanded by the specific plane. Note that it's not necessarily the case that all planes using a certain reticle accept the one and same image file without some rescaling.

Bigfoot,
For some years I was a met tech/weather station manager For Environment Canada, working most of that time in the Arctic. I figured my chosen nick would be 'decoded' here!