Why485

Something odd has happened to the Tu-142's damage model. I shot off both of its wings, but it remained in control flight and was able to eventually fly to an airfield and land.

I get a black screen whenever I try to change any graphics options. I don't have an Oculus Rift. Here are the files requested, although in my case I had to grab the options.lua from my regular DCS folder in Saved Games because the one in openbeta isn't the one the game is using. Edit: Adding the VR section fixed this crash. Data.lua options.lua

ED has never officially spoken on this issue.

The switch to impostors in the mod will happen at closer ranges compared to the current settings that DCS uses. At medium ranges they should be about the same, and at long ranges small should be easier to see. If you want, you can try changing the maxsize in Config\Effects\impostors.lua of all the settings to 16 from 12 if you want to try the setting that vanilla uses. I was actually surprised to see that ED raised it so high, especially for small, compared to the values it used to have. That's actually a good thing, as I think the system works best with settings like that. That's about all I can say, as the rest of the features are so different. If small works best for you, then you should use that. This mod was never really meant to be for people to practically play with, but just a visual aid in a point I was trying to make about how flawed the current system is and ideas on how to make it work better.

After a quick look at the test scenario I always use, it doesn't seem like it's any different from what it used to be. It's always been an awkward band-aid onto a flawed system. The snapping between different angles in particular will create sudden changes in visibility when they should be more gradual over time. There's also the fact that no matter how visible I make an impostor, once the model gets close enough that it switches to the 3D model, that 3D model will be as difficult to see as it always has been. Are you sure it used to be better? I also want to point out that in most situations, this mod will make things harder to see versus the vanilla medium/large visibility settings. The vanilla impostors are much bigger at most ranges because they don't scale with distance.

I'd have to check to see if something has changed with recent patches, but assuming nothing is different, it sounds to me like you're just getting close enough that it switches to the regular models.

It's local to your computer. There's no way currently for the server to force the mod on clients. If you join a vanilla server with the mod activated, you would actually be putting yourself at a disadvantage in most cases because this makes the impostors overall more difficult to spot than the medium and large settings. Versus small, impostors should be easier to see at close range while vanilla would be easier to see at a distance.

I just wanted to drop a message saying that I really appreciate the work that goes into this mission and server. It's the best of the "fly around and shoot stuff" missions out there because there's a wide variety of targets and it's just overall more involved than your typical 104th mission. Keep up the good work.

I installed this on the latest version (1.5.3.52875) and while the radar seems to work a lot better, the IFF no longer works at all. As Deathbane noticed, the IFF switch can not be toggled on/off anymore and stays in the off position.

Agree. I would really like to know ED's opinion on where it stands now, where they're planning to go with it, and in particular, what the intent of the system is.

That's the real question, isn't it? Did ED implement this with the intent of it being a realistic solution to a problem plaguing DCS for years? Or was the intent just to have an alternative to labels? Frankly, I think it's the latter. ED showed no interest in the slightest until they started experimenting with VR, a situation which greatly exacerbates all the spotting problems. The way the system works plays right into the low pixel densities and limitations of current VR headsets, and I don't think that, or the timing, is a coincidence. By this point, I've given up all hope of model visibility ever getting the treatment it deserves. I get the impression that ED genuinely doesn't believe the inability to see aircraft is a problem. I would love nothing more than to be proven wrong, but ED doesn't ever talk about these kinds of things so I can only go by their actions, and their actions have not been encouraging. Inspired by the paper you posted earlier in the thread, I experimented with the scaling method they described. The combination of reading the paper and seeing the result of it first hand really changed my mind on a lot of things I had previously thought about aircraft visibility and how it should work. I'd post a direct quote and the video itself, but I can tell you from my own experience that doing such things is grounds for a ban on these forums. If you're interested in those experiments and my thoughts on how they apply to DCS check out this link. TL;DR: This is a solved problem, and it's a solution that DCS needs.

It's mostly the ancient LOMAC era models that suffer from this, with the Tu-95 being an especially bad example. Those models tended to use bigger/darker models for their LOD models to help a tiny bit with spotting. The newer models don't do this.

At the moment the best you can do is put model visibility to high. Model visibility is a big problem in DCS and has been for over 10 years. The model visibility system is a flawed band-aid in the right direction but that's unfortunately all it is at the moment.

Wow, I wish I had seen this sooner. I've always heard that BMS based their scaling off of research papers used in actual military simulators but I could never find a source for that. This whole paper is fascinating. Thanks for pointing this out. For the record, there is no fixed distance at which impostors start drawing. The distance they draw is determined by the pixel size of the fighter. For example with medium settings, the model switches to an impostor when it takes up 8 pixels. As this is basically a measurement of the objects apparent size, the exact distance that it turns into an impostor varies based on the targets aspect (when head on, a target takes up less pixels, thus would switch closer), true size (a B-52 switches much further than an F-5), and screen resolution (the higher your resolution, the further the switch occurs).

I think I mentioned it somewhere in this thread, if not in the OP, but if you fly with this against people who are using the vanilla settings you are putting yourself at a disadvantage. This mod doesn't make spotting easier, it makes it more realistic compared to the default settings, as well as compensates for resolution differences and a few other things. I do not see this mod as the final solution to this problem. Rather I'd like to see ED use it as inspiration for a better implementation of their visibility system. The problem though is that I don't know what ED thinks of the issue, or if they think an issue exists at all. For all I know they are perfectly content with the current model visibility setting and don't see it as anything more than an assist, like labels, to make the game easier. If that's the case, then I doubt they will ever be improved from where they are now. The current settings are reasonable if that's what you're going for, but I don't think that's what ED should be going for. I'd love to hear ED's stance on this.

Unfortunately there's only so much I can do with the limited information and access I have, so it's not perfect by any means. There's some flickering and resizing that sometimes happens because of the way sprite sizes and aspect change that this mod exacerbates a bit. Presumably an ED natively implemented solution would have none of these issues. This is more of a proof of concept to show to visualize what I think a smartly implemented model visibility setting would look like.

What do you mean? That's too vague a question to really answer.

They're all the same. From the OP: One of the points I wanted to make was that this shouldn't be a configurable setting and there should just be an on/off with on being realistic and off being for the diehards who hate being able to see anything out the window. Sort of like Falcon BMS's smart scaling.

Nothing's changed about how impostors are drawn. ED just set the values back to values that actually work. The 1.5.2 values were completely broken and made the whole system effectively worthless. You can see impostors again (too easily IMO) now with the new settings. They're a middle ground between 1.5.1's insanely visible sittings and 1.5.0's very conservative and hard to see settings. However, the ED impostors still suffer from all the same problems as they originally did. As I said before, nothing has changed about how they render,they just tweaked the values again. It's appreciated that the system is back to working, but it's still too simplistic and flawed. The mod still works as intended with the original impostors.lua. If you use the much larger impostors.lua from 1.5.3 you'll get very different and unintended results from what this mod intended. Further, all the variables in the shader files are tweaked to get the right view distances for the specific pixel sizes in the mod's impostors.lua. TL;DR: Use the mod's impostor's.lua if you want to keep using the mod. In 1.5.3, ED tweaked the broken values of 1.5.2 so that the impostor system is back to what it was like in 1.5.1, but they didn't fix any of the fundamental problems of the system.

I really wish I had a 4k screen to test this on, because I'm not exactly what it looks like. From testing with a baseline res of 1080 versus 1440, I noticed that the scaling gets a little jumpy when it has to scale up. The impostors also get noticeably blurrier if scaled up. They just don't scale up very well in general, but they can scale down perfectly fine. It's one of the things that if I had access to the layer between the impostors.lua and the shaders would be trivially easy to address, but unfortunately that would be in the DCS source code. I only made this mod as a proof of concept of what I'd like to see ED do. The whole thing is kinda hacky because I'm working with what little I have available to tweak. If ED were to fix this themselves, they can do it "right" and prevent the weird little anomolies that sometimes happen with this mod.

Please keep in mind that OP is talking about the MiG-29A or MiG-29G, which nobody flies ever and is basically a gimped MiG-29S that's worse all-around with worse weapons. It is very much limited in terms of capability, and that's on top of the MiG-29's already limited capability compared to the Flanker and F-15C. It is by no means a "bread and butter" plane. The OP already has a very compelling list of arguments for why specifically the MiG-29A (or G) and only the MiG-29A (or G). Nobody is asking for a free MiG-21, which is a module that I believe is absolutely worth the money and has no reason to be free. I'm sorry but saying that a free downgraded FC3 DCS fast jet that nobody with FC3 ever uses anyway would attract only a "shallow hot headed lone ranger Call of Duty player" is a disgustingly short sighted and arrogant viewpoint.

I guess I shouldn't be surprised by the quality of the replies considering where we are, but I think it's a real shame that hardly anybody replies with meaningful criticism beyond just laughing and dismissal of what I believe is a very reasonable proposal. OP outlined all the reasons for why it's a good idea, and they're ideas that I completely agree with. Even though I bought A-10C, I didn't really get into modern DCS until I started flying around with the FC3 jets, and it took me a lot of convincing to eventually get to the point of essentially rebuying LOMAC for a third(!) time. However, once I flew around a bit in the F-15C (which back then had just gotten its swanky PFM) I was completely sold, I've bought many new modules since then, and DCS is one again one of the main ways I spend my video game time. I don't care for slow ground attack planes or prop planes. Su-25T does nothing for me, and I can count on one hand the number of times I've flown the TF-51. If there had been a modern fighter, and it seriously doesn't have to even be a good one, then I can guarantee you that I would have been sold right away and jumped right in. This isn't about you. This isn't about current DCS players, although it does benefit them. It's about drawing in new players to further support the game and to grow the community. Having a teaser plane with a very attractive role (who doesn't want to fly fast jets and get into dogfights?), you'll give people a much more compelling argument. On top of that, it's also a great way for new players to unwittingly ease themselves into the more complex modules by learning the basics in a simplified systems plane. Don't give me that crap of "why do you want everything for free?" This isn't about me. This is a strategic long term decision that I believe would be in DCS's best interests in attracting new players.

It just crashes if you try to load the Nevada map when you select it under maps in the mission editor. The fact that all the GUI Is there for it, and the fact that Caucasus now runs in 2.0.3 makes me think that the 2.0 merge is coming very soon.

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.