Will the Eurofighter have an ir seaker like the su-27 have for example?

[Dragon1-1]

We do not need to simulate real radar physics, except maybe with very early analog systems. More advanced radars to significant processing, and enough information can often be gathered to figure out just what the display should look like in given conditions. Physics can be used to figure that out. Things like RCS based on a lookup table are fine, as long as they can account for things like different RCS at different aspect angles (this is something that's missing from DCS at the moment). In fact, physically simulating radar waves would significantly tax our already strained CPU resources.

[Harlikwin]

10 hours ago, F-2 said:

Neither of those are physics based though. There are only a handful of radars with very detailed documentation in the public domain, I’m really only aware of AWG-10 and the IEEE papers on APG-66. Almost all of it is educated guesswork based off what info is available and cleaver implementation. this is very true with old radars and their quirks and pre solid state hardware. If Raz can do a convincing multi mode pulse Doppler radar like RDI and it seems APG-70 then I don’t see why the benefit of the doubt shouldn’t be extended to a proven team with a number of subject matter experts and industry support. This should be considered in light of True Grits alter ego The Adams group which is aiming to develop a professional version of the sim. 

I mean radar physics/function are pretty well understood at least for the older stuff. Once you know things like antenna gain, power out, etc you can get some vaguely useful models since what a radar signal does when it bounces off the ground or an airplane is known and can simulated. IRST, its sort of the same in that regard, but the problem is what makes modern IRST "good" isn't the physics generally speaking. Its literally <profanity>loads of processing/algorithms, and how all that works is most definitely not public domain. And thats saying nothing of the current state of IR modeling in DCS which is terrible, I mean stuff like clouds blocking IR (and also reflecting it) should be a basic feature of the sim, and its not. Not to mention ground reflections etc. I had high hopes for the "flir" model from ED, but well, lets just say its not good despite the work they put into it, primarily because they focused on units not "background".

6 hours ago, Dragon1-1 said:

We do not need to simulate real radar physics, except maybe with very early analog systems. More advanced radars to significant processing, and enough information can often be gathered to figure out just what the display should look like in given conditions. Physics can be used to figure that out. Things like RCS based on a lookup table are fine, as long as they can account for things like different RCS at different aspect angles (this is something that's missing from DCS at the moment). In fact, physically simulating radar waves would significantly tax our already strained CPU resources.

I'm not talking about doing any of this in real time necessarily. But you do need to simulate/display things like false contacts. And aside from the M2k RDI radar I can't think of another radar in DCS that does it. Or simple stuff like merging far contacts into a single return. 

Alot of people have talked about, we just need to model what the pilot sees, well yes thats right, but aside from the M2k radar no one is doing it. I guarantee you that both the APG73/68/awg9 etc will have a false alarm rate, and they don't have magical resolution at long range to break out contacts. And they all suffer in terms of detection range when flying at low alt due to sidelobe issues, and the resulting processing therof. And then there is the meme that is "notching" in DCS, and as a hint, for modern radars that mostly doesn't happen. And yet all of these "pilot side" things aren't being modeled currently on the majority of modules.

Edited November 1, 2022 by Harlikwin

[Dragon1-1]

Notching is overmodeled for anything that's more advanced than the F-14, but that's a known issue, and besides, F-15s were notched in Gulf War, modern radars (besides AESA) aren't magic. It's just a matter of how big the notch actually is. I agree, we need improvements, particularly to contact separation, which is implemented in the F-14, but not in others. Things you mentioned are not left out because of lack of data, they're left out because the devs didn't bother implementing them.

All this has little to do with physics, though. False contacts on the Mirage are probably just a clever dice roll, not radar waves reflecting off moving vehicles or tertiary echoes or whatever.  It doesn't need to be anything else. Also, note, even in the Mirage they're filtered out in some modes, so a radar significantly smarter than RDI (say, in the F-16) would probably not display those to the pilot. The F-14 would probably only display those to the RIO.

[Harlikwin]

7 hours ago, Dragon1-1 said:

Notching is overmodeled for anything that's more advanced than the F-14, but that's a known issue, and besides, F-15s were notched in Gulf War, modern radars (besides AESA) aren't magic. It's just a matter of how big the notch actually is. I agree, we need improvements, particularly to contact separation, which is implemented in the F-14, but not in others. Things you mentioned are not left out because of lack of data, they're left out because the devs didn't bother implementing them.

All this has little to do with physics, though. False contacts on the Mirage are probably just a clever dice roll, not radar waves reflecting off moving vehicles or tertiary echoes or whatever.  It doesn't need to be anything else. Also, note, even in the Mirage they're filtered out in some modes, so a radar significantly smarter than RDI (say, in the F-16) would probably not display those to the pilot. The F-14 would probably only display those to the RIO.

Precisely. And honestly I think devs need to held to a higher standard on this stuff. Especially ED when it comes to core stuff. And double plus for actual IR modeling. But to do that right they need to actual model things like object IR reflectivity/emissivity correctly for background/map objects which is no small task even if implemented simply.  (you could do it by there being some time of day based probability over different terrain types of there being some reflective/emissive down there, with a very high probability for areas with man made objects i.e. towns/roads/rail etc)

False: They actually do use some ray tracing for it so it interacts with the ground, they basically put all that processing on a second core to do an actual "simulation" albeit a simple one. Do I think that its 100% correct? No, of course not but it does give you some basic behavior that should be there. Also I think the mirage does model merged contacts as well. 

[Dragon1-1]

Are you sure you're not confusing ground clutter with ghost contacts? The latter don't depend on ground, and in fact, appear to pop up in more or less random locations, even if the radar is looking up. Raytracing is for simulating ground clutter, and Mirage's isn't the only radar that shows that (in fact, pretty much every early radar in DCS will be flooded with false contacts when it sees too much of the ground).

I do agree about holding the devs, particularly ED, to a higher standard when it comes to radar modeling. That a system is modern doesn't mean it's magic.

[Harlikwin]

3 hours ago, Dragon1-1 said:

Are you sure you're not confusing ground clutter with ghost contacts? The latter don't depend on ground, and in fact, appear to pop up in more or less random locations, even if the radar is looking up. Raytracing is for simulating ground clutter, and Mirage's isn't the only radar that shows that (in fact, pretty much every early radar in DCS will be flooded with false contacts when it sees too much of the ground).

I do agree about holding the devs, particularly ED, to a higher standard when it comes to radar modeling. That a system is modern doesn't mean it's magic.

Nope not at all, I mean all radars will have a false alarm rate. Then you get into stuff like side lobe clutter at low alt, which again is barely modeled (i.e. autogain circuits tend to deal with this on modern radars but it basically means you get less range out of a radar at low alt)

[Dragon1-1]

Modern radars get less range than they theoretically could due to filtering out stuff like this. The Viper has, or should have, a special uplook search mode which disables most of the filtering, letting you get more range out of the otherwise rather small radar. Not sure if the Hornet has something like this, it might be done in a completely different way.

3 hours ago, Harlikwin said:

Nope not at all, I mean all radars will have a false alarm rate.

I'm not saying they don't, I'm saying this would not be modeled with raytracing, because then it wouldn't be a random "false alarm", but a filterable ground return. 

[Harlikwin]

1 hour ago, Dragon1-1 said:

Modern radars get less range than they theoretically could due to filtering out stuff like this. The Viper has, or should have, a special uplook search mode which disables most of the filtering, letting you get more range out of the otherwise rather small radar. Not sure if the Hornet has something like this, it might be done in a completely different way.

I'm not saying they don't, I'm saying this would not be modeled with raytracing, because then it wouldn't be a random "false alarm", but a filterable ground return. 

Yeah correct on all counts, modern radars know which way they are pointed. And yes false alarm rates are their own thing, though if you point it at the ground you are gonna get more clutter/false targets so its not exactly unrelated.