Universal Model Visibility Setting Mod

[Esac_mirmidon]

I´ve tested this mod and for me it´s the best setting out there for Air to Air. Ingame looks a lot better but here are some pictures from a Mig-21 intercepting an A-10C from 9 NM.

 

 

 

 

 

 

 

 

 

 

It´s a lot better than the stock one. The transition from FAR to NEAR it´s very good, the imposters are darker and easier to pick but progressive with distance, and the size its very good in relation to the distance you are.

 

BUT:

 

For ground targets it´s not good at all. The contrast is to much, a lot darker than the stock ones and even the ground targets doesnt show so far, thats good, they are very dark and visible when they start to appear. The size of ground targets are ok for me but they are very dark and visible.

 

Some pics.

 

 

 

 

 

 

I hope with time ED could consider to split Air contacts from Ground contacts with imposters for each indovidually.

[YoYo]

Ground targets are important also. Is it possible to add this mod but only for air objects?

[Esac_mirmidon]

I don´t think so at this moment.

 

It´s a pity because air contacts with this mod are great, but ground targets not are over darker.

[Why485]

Ground targets are important also. Is it possible to add this mod but only for air objects?

 

It's all or nothing. At least as of right now I haven't been able to figure out a way to tell the difference between ground and air targets given the very little information I have in the shader. The mod is optimized for air to air, as that's where I feel the biggest problem lies, and while it's certainly better at air to ground spotting (as in, it's harder) compared to the stock ED solution, it's still too strong.

 

That is something that ED needs to fix. There needs to be separate values for air targets versus ground targets.

[Rikus]

In a 27" inch monitor 1080p a "contact" looks bigger than in a 27" 1440p

 

Shouldn´t has the contact the same area in both monitors, but in the 1440p same area but with more definition?

[ron533]

Important mod,thank you!

Does it address target size as well?

I expect to spot an F15 or SU27/31 at twice the range I spot the F16/MIG21.

 

Cheers:)

[98abaile]

I have no idea how shaders work, but would it be possible to change the opacity or size depending on the contrast between what the imposter is silhouetted against?

[SinusoidDelta]

I imagine you could modify some of the fog parameters to mask objects below a certain altitude.

[EntropySG]

I almost spilled my coffee when I looked at the first picture in post#1.

Never seen such ugliness with scaling before, guess one needs to run low res to get this?

Im running at 4k, that means I will see ground targets a lot later than anyone with low res screen setups? wow, did not know that.. thats screwed up for sure :O

[Markindel]

Thanks, great work.

[Oznerol256]

I almost spilled my coffee when I looked at the first picture in post#1.

Never seen such ugliness with scaling before, guess one needs to run low res to get this?

Im running at 4k, that means I will see ground targets a lot later than anyone with low res screen setups? wow, did not know that.. thats screwed up for sure :O

 

The scaling in DCS works by placing an image over the small target. So the actual 3d model is not rendered bigger, but instead a big image is placed over the target. The problem: That image always has the same amount of pixels (resolution). On a low res monitor the same amount of pixels make up a much bigger portion of the image than on a high res monitor.

 

To answer your question: No, you dont see targets later. You only see them smaller.

[EntropySG]

The scaling in DCS works by placing an image over the small target. So the actual 3d model is not rendered bigger, but instead a big image is placed over the target. The problem: That image always has the same amount of pixels (resolution). On a low res monitor the same amount of pixels make up a much bigger portion of the image than on a high res monitor.

 

To answer your question: No, you dont see targets later. You only see them smaller.

 

ok, interesting to know thanks! :)

[Ramsay]

A very elegant solution.

 

I've been using your mod for the last couple of days without issue and find it works really well.

 

Flying the L-39C, I still need to use the F10 map or AWAC's to get a BRAA, then FOV zoom/scan to spot air targets - which is as it should be. For some aspects (head-on/trailing) and ranges the imposter sprite still disappears but overall your mod feels far more realistic than ED's 1.5.2 implementation.

 

Ground targets are little too easy to spot, as I don't have to FOV zoom/scan to find them but it's a fair compromise. As noted, it's something ED might want to address with separate air/ground settings.

Edited February 16, 2016 by Ramsay

[Chrinik]

Thanks for this. I also hope ED will make one single size of "impostors" lockable and verifiable by server (for multiplayer sake) and maybe adopt your approach.

 

I don´t know if you are one of those guys that runs a beast of a machine, turns everything on and expects it to look amazing and sharp...because if you are one of those guys, there are a couple options in the grafics menue that result in the exact OPPOSITE happening.

 

So you eat fps for a worse looking game.

 

Since the grafics options menue seems to have a few options missing or I´m just not remembering them, try turning depth of field off.

I know there was a setting that caused distant objects to blur like crazy.

[Kuky]

The scaling in DCS works by placing an image over the small target. So the actual 3d model is not rendered bigger, but instead a big image is placed over the target.

 

Not correct, its quite the opposite. There is no image over the object, the object itself (actual 3D model) is resized. For ground units currently all units have basic rectangle shape object show at 10km and this "box" size is maintained when it reaches certain distance (whatever value is used in Impostors.lua)

[zaelu]

I don´t know if you are one of those guys that runs a beast of a machine, turns everything on and expects it to look amazing and sharp...because if you are one of those guys, there are a couple options in the grafics menue that result in the exact OPPOSITE happening.

 

So you eat fps for a worse looking game.

 

Since the grafics options menue seems to have a few options missing or I´m just not remembering them, try turning depth of field off.

I know there was a setting that caused distant objects to blur like crazy.

 

I am sorry but I don't understand how your answer is related with what you quoted from my post. :huh:

[Why485]

I have no idea how shaders work, but would it be possible to change the opacity or size depending on the contrast between what the imposter is silhouetted against?

 

I imagine you could modify some of the fog parameters to mask objects below a certain altitude.

 

Due to the way shaders work, I have no idea what's behind the plane, so I can't blend based on background. I also can't use fog to adjust based on altitude because altitude is an absolute number but AGL isn't. I spent a lot of time trying to see if I could somehow calculate AGL and use that to hide ground vehicles, but unfortunately I couldn't find anything.

 

Not correct, its quite the opposite. There is no image over the object, the object itself (actual 3D model) is resized. For ground units currently all units have basic rectangle shape object show at 10km and this "box" size is maintained when it reaches certain distance (whatever value is used in Impostors.lua)

 

No, he was right. You're either mistaking impostors for LOD models or mixing up the concepts completely. They are totally separate and unrelated systems.

 

When an impostor is drawn, the model behind it stops being drawn and a sprite of the model is instead drawn on top of it. Hence the name impostor. It's a "fake" stand in for the real model.

 

The boxes you're thinking about are the LOD models, which are models that swap out based on distance. The further you are, the less complex the model is, eventually resulting in the box you are talking about for ground objects. Changes in LOD models have absolutely nothing to do with impostors or how they're rendered. It's a completely separate system.

 

Ozneral is spot on with his explanation for why resolution has such a great effect on the apparent size of impostors. It's something that's easily correctable, but ED doesn't for whatever reason.

Edited February 17, 2016 by Why485

[Rivvern]

Great mod! Works very well. I think we will force use of this in our community until DCS gets a better model enlargement.

[Justin Case]

And push ED to fix the presentation of ground vehicle imposters.

[Oznerol256]

Not correct, its quite the opposite. There is no image over the object, the object itself (actual 3D model) is resized. For ground units currently all units have basic rectangle shape object show at 10km and this "box" size is maintained when it reaches certain distance (whatever value is used in Impostors.lua)

 

Very interesting and certainly a more elegant and better solution. This means all little details of the aircrafts, like angle and used skin, are retained, right?

[Why485]

Very interesting and certainly a more elegant and better solution. This means all little details of the aircrafts, like angle and used skin, are retained, right?

 

No, what he said was completely wrong and not at all how impostors work.

 

I think he's confusing and mixing up the LOD system with the impostors, but they are totally unrelated and independent of each other. All LODs do is reduce the amount of polygons on a model based on distance. They're not dynamically scaled models based on distance. Wags has actually explicitly stated in the past that for some reason they cannot scale the models as a form of visibility aid because it would mess up radar returns.

 

For some models (the game is very inconsistent about this, and typically it's the older LOMAC legacy models) the lower LOD models have a slightly bigger fixed size, but not in any meaningful way.

[Kuky]

No, what he said was completely wrong and not at all how impostors work.

 

I think he's confusing and mixing up the LOD system with the impostors, but they are totally unrelated and independent of each other. All LODs do is reduce the amount of polygons on a model based on distance. They're not dynamically scaled models based on distance. Wags has actually explicitly stated in the past that for some reason they cannot scale the models as a form of visibility aid because it would mess up radar returns.

 

For some models (the game is very inconsistent about this, and typically it's the older LOMAC legacy models) the lower LOD models have a slightly bigger fixed size, but not in any meaningful way.

 

Please check things before making such claims, I am not confusing anything.

 

Impostors work in such a way that once 3D model of an object reduces to certain pixel size, and it reaches max value in impostors.lua settings, at this point the 3D model switches to impostor effect and then reduces in size (up to min value in impostors.lua settings) and then stops reducing in size and just starts fading (increase opacity) until its no longer visible.

 

You can clearly see how this works if you change the min and max value to something like 100 & 300 pixels.

[Why485]

Please check things before making such claims, I am not confusing anything.

 

Impostors work in such a way that once 3D model of an object reduces to certain pixel size, and it reaches max value in impostors.lua settings, at this point the 3D model switches to impostor effect and then reduces in size (up to min value in impostors.lua settings) and then stops reducing in size and just starts fading (increase opacity) until its no longer visible.

 

You can clearly see how this works if you change the min and max value to something like 100 & 300 pixels.

 

What you say in this post is true, yes, and is what myself and Ozneral were saying before.

 

Although, the impostors themselves never fade with distance. (Which is a decrease in opacity, not increase.) Technically there is a gray area with how alphaExp works at extremely short and impractical distances, but in practice alphaExp functions as a flat alpha modifier. I wish the default ones faded with distance, but unless that's changed in 1.5.3, which I'm fairly sure they haven't based on some gameplay earlier today, they don't. They only get covered in more fog which makes them slightly harder to see with the distance, as the haze causes a lack of contrast. Eventually, when the model stops drawing completely (when this happens is based on when the LOD tells the model to stop drawing), it will sharply disappear.

 

What you were saying before where there is no image rendered over the model, (There is. This is the very definition of an impostor sprite.) and that the actual object itself is resized (Model geometry is never dynamically resized based on distance.) were patently false. I know how the impostors work. Dissecting the system was literally the first thing I did when EDGE dropped. There's no way I could have "fixed" them without having an in depth understanding of how they work.

 

You seem to have the concept of how impostors work correct, but perhaps your terminology is what's wrong and throwing me off.

Edited February 18, 2016 by Why485

[Kuky]

Well there is no image rendered over the model, the model itself changes. You can see this if you try increase max value to something large to make the impostor pop up very early.

[Why485]

Well there is no image rendered over the model, the model itself changes. You can see this if you try increase max value to something large to make the impostor pop up very early.

 

An impostor sprite is a picture of the model that gets drawn in place of the model. I don't know how I can explain it more simply than that. That is the very definition of an impostor sprite.

 

Typically it's used as a performance saving strategy, as drawing a sprite is significantly less expensive than drawing a 10k polygon model. At a distance, you can't tell difference anyway. Here's a few examples of what impostor rendering looks like

 

Here's a scene with half of it rendered using regular models, and the other half rendered using impostor sprites.

 

Here's what it looks like in wireframe. The impostors are literally just a billboard sprite of the model they're representing.

 

The model does not change. The model simply stops being rendered and instead you get the impostor sprite drawn over it. It's really that simple. Are you just trying to argue semantics? Is it that I keep saying "over" instead of "in place of?" Admittedly the latter is more correct, but the model itself is simply not drawn. It's not like the model itself is changing in any way. Even then, you keep saying things like "there is no image rendered over the model" when that's precisely what is happening.

 

I want to believe we're saying the same thing, but you need to get the terminology right or else the things you're saying are just flat out incorrect.

 

I'm starting to sound like Mojo Jojo. All I'm doing is repeating the same thing over and over again just with slightly different words each time.

Edited February 18, 2016 by Why485