Harlikwin

Yeah, I just tried this. I remember doing it once before and it worked, but it seems like there is something else I need to do to get it to forget the device or something like that because the Z-boot device or new drivers are not loading up.

Its impressive they have anything.

Yeah, currently with my 3090 and 12900k and a G2, on heavy MP missions in VR (ECW) I'm getting nearly identical CPU/GPU frametimes. Offline of course is much better, and there the GPU is usually the limit, so I might consider it, especially if ED actually delivers some performance boosts from multicore.

Jolly good, thanks for your contribution.

What does this have to do with the WMD7?

Interesting, is it just the 2 pages or is there more actual info in there? Also as I understand it China also had Lantirn "copy/equivalent" that preceded this pod right? Its always been weird to me that this pod has basically a 3rd gen imaging sensor(s), but can't do stuff like stabilize the image at a distance i.e. area track. Using the laser for coordinate generation makes perfect sense though. I have seen this language in several places related to the WMD7 though "Automatic tracking targets after identified" which seems to imply it needs some sort of point track lock maybe. Or it could mean it can be stabilized on a target after its detected, which in the case of a building could be quite a ways away. In terms of zoom the NFOV is comparable to western pods. Is the sensor resolution actually known? If not I can take a stab at back calculating it from these FOV and the detection range. WFOV: 4.3°×5.8° NFOV: 1.4°×1.9°

Depends on when you are talking about. And who has access to it. Early 2000's GPS coupled INS were 5-15meters accurate for ownship positions. With DGPS its better but that all depends if you have access to those signals. And well, it all depends on the WMD-7, which I'm not particularly familiar with since there isn't much info on it.

Modern point tracks are way more processed than "contrast" lock. But at any rate, I thought you were talking about being able to cue the pod at longer ranges and have it be stabilized. I guess its possible its not, but that implies a much lower capability than the date it was built would suggest as well as the rest of the components used. Well the issue with "simple geometry" is that it is in fact not simple. You have errors and error propagation for coordinate generation starting with the basic fact most aircraft only know their own absolute location within a few to 10s of meters with GPS. From there you have additional problems with pointing accuracy and so forth. Your primary sources of error are Ownship position error, Operator error (is it really on the right pixel), then you have heading error, bearing error, alignment error, pointing error, and range errors which then result in altitude errors which is generally why you need that laser range. So you will have pretty large CEP and LEP at longer range. So yeah, depending on the actual range to a target you have "rough" coordinates, maybe enough target a big building, but not a tank or truck.

I think your guys terminology is a bit imprecise. I generally agree that it seems odd that the WMD7 cannot do something like an "area" track, usually this is done just using image comparison, techniques like edge detection matching etc. That being said, while this type technology was originally developed for pods like LANTIRN, IDK if the WMD7 would have this tech or not. But its not exactly a "contrast lock", its more sophisticated. But to me it seems like 2000's era TGP would likely have this capability as the algorithms are generally known and processing (which was the original issue) at this point is no longer a limit. Where you have to break that out is can you generate actually usable coordinates from it for targeting weapons. For the best case of that with highly precise coordinates (like a tank) that you most likely do need an actual laser defined point on the ground. That being said there are angle rate tracking techniques to get rough ranges and "coordinates" that are employed by western pods, but IDK if the WMD-7 would necessarily do this. And the issue with these coordinates is that they will be wrong from longer ranges. No one use <profanity> like TOO for long range targeting for JSOW's or JDAMs etc. and this is something ED and other devs have gotten absolutely wrong. You can easily check the range of WMD 7 for detection of objects using johnson criteria if you know the sensor FOV, the sensor pixel count. But 22km doesn't seem to bad to me for those sized targets.

I mean maybe he has, still I doubt anything sensitive is gonna be modeled well if at all. I mean Pirate as an example is system with decades of development that went into it and only recently has it really reached somewhere near its actual initial promised potential from 30 years ago and it still has issues. And I very much doubt any of the more interesting things it can do and even more so its weak points will be modeled with any real fidelity. AFAIK, the whole deal with Deka is one of "just trust us", and I personally don't consider anything about the Jeff on a systems level to me modeled accurately, from the FM to the radar or other systems. They might be "close" but none of its accurate IMO. As for the 23 thing you have it backwards, the guy making the 23 had access to all the docs needed to do it. As for HB I'm far more interested how the F4 goes since that is pretty well documented publicly.

I mean I appreciate your optimism. But if ED can't decent documentation on literally 90's era MSI systems, I don't think the German MOD and certainly not the British MOD (Pirate), are absolutely not gonna be giving out details on current in-use systems. At best it will be entirely made up IMO.

HB is pretty close to ruining its rep by even trying to do the EF. They are relying on the ignorance of the customer base IMO. Because the customer base doesn't know what it doesn't know. Even the basic versions of the EF had systems like MSI that aren't in DCS, sensors like IRST currently aren't credibly modeled in DCS in any reasonable form. And lol AESA radars? Nope, not gonna happen IMO.

A renaissance man I see.

You shouldn't be. I very much doubt HB can actually even remotely model the actual capabilities of Pirate in DCS in any sort of dynamic sense. And that is a job thats probably 1000x easier to do than do anything resembling a good job on how MSI in the Typhoon works. Mind you specific tactics, profiles and flight paths have to be flown to make modern IRST systems work well. Again, is that information going to be public? How much "realism are you willing to give up" ED hasn't managed to do it with the hornet at all. and that in many ways is far simpler system/sensors than whats going in the typhoon since they are mostly going off 90's era info. Think about this. You are approaching how actual modern day fighter combat works from the standpoint of baron von richthofen. All he knew was you needed some speed, luck, and how to turn tight, and gun the enemy... All you know from DCS is turn burn, lock contact, shoot fox3 win... Modern fighter combat, especially in DCS for most end users at this point is very much a case of "what you don't know". You basically think the typhoon is really awesome 3rd generation fighter... It turns good, it goes fast etc... THAT is NOT what makes the typhoon a good "fighter" IRL for the most part. And while DCS can kind of model that 3rd gen stuff, (and it will be wrong in the actual details). It will be far more right in those 3rd gen details, which are largely irrelevant, because its the sensors that matter most, and thats what DCS does not model well, if at all. And thats before you get further into the knowledge swamp and start adding EW to that mix.

So, the whole MSI thing in this community is badly misunderstood as to what it actually is and can do, and more specifically as to when it could do what. And mind you what I'm going to write, is VASTLY oversimplified. On a very simple level, in case 1 you have say 3 sensors, your radar, your IRST, and your RWR and lets say a basic INS system. And lets say the year is the early 90's for these capabilities. Which in large part determines how much actual processing power you have to deal with and sort all this data. The accuracy of all 3 of these sensors is limited by physics, and I'll simplify that to just dealing with lets say wavelengths and how they are detected. So lets start our hypothetical scenario, you are flying along all passive like at 500kts heading 000. Your RWR is gonna be your worst sensor with the most ambiguity but it can get you a rough bearing of lets make up a number say 10 degrees. So, this sensor tells the MSI "computer" hey boss ping bearing 340 (the submarine analogy here will become a bit more clear later). So the MSI computer gets an target at an angle of 340 "ish" at time and position X and time Y (time is well known, ownship position less so, but lets say internally referenced its perfect at time Y to 0,0,0), the RWR periodically updates this contact every time Mr. Enemy radar paints over your ship. So you being the ace you are think to point your amazing IRST sensor in that general direction and tell it to search whatever volume of space along that bearing line. IRST of course works not on really bad radar wavelengths of whack units like "centimeters" holy cow, how crude is that. IRST works on in the micrometer range (way way way smaller wavelength wise), and therefore the angular accuracy of IRST's is "phenomenal" relative to things like some poor mans 10m long antenna (your RWR) or your radar. Well off your IRST chugs along and lets say a minute later Mr IRST gets a ping at bearing EXACTLY 342.3 degrees. The sensor of course keeps track of this contact. So now at Time Y+1min, at location 0,0,0 + 500kts in the north direction for 1 minute, PLUS whatever INS drift you have in your system in all 3 coordinates. Mr MSI computer now has 3 radar ping points over that 1 minute however, and they are still roughly coming 340, 345, and 335 according to your RWR but you have moved in space as well. Whats a poor MSI computer to do? Well ok, so the target is either at 340, 335, 345 or 342.3 degrees. The RWR has a track file that is "on average" 340 degrees. Luckily Mr MSI computer was programmed by a smart man that knew something about physics that realizes well, the IRST sensors is the better sensor but hedging bets the target is likely at 340 degrees or 342.3 degrees, now that is a volume of space, or "solid angle" if you want to think about it this way. Of course Mr. MSI computer can also choose to ignore the RWR track, or in a "computation" give it less weight because we know its a low tier sensor. Well you keep this up for like 10 minutes flying as due north as you can, the track gradually changes from 342.3 degrees to like 330.7 for the IRST, and you get a bunch of data points on that. Your RWR is still getting lit up, however the tracks are now moving "away" from the IRST track... So now for any "submarine sim" fans will know what is happening. We are now building a ranging baseline track on our enemy contact, but instead of relying on absolutely huge sonar waves (many meters long) we are now using IRST tracks, and radar tracks to do the same thing. Now we can start to get range data for our contact from either our IRST or our RWR. However, due to our INS drifting our own ship position is now not a point, its a "sphere" (technically a spheroid). So we now have an estimate of enemy range from 2 sources, our RWR track file, and our IRST track file... They aren't the same, the RWR is giving us one range and the IRST another. Which do we believe. They are roughly coming from the same direction... The other fun bit we have is that our RWR has also classified the target as a mig29 radar, we know the output power of that radar, and we also understand the radar equation, so we now also have in addition to that angle data, some very rough range data from the power of that radar for each "ping" and its getting "closer". So at this point the MSI algorithm has merged with various weights, all of these angle rate tracks, along with estimates of range from these sensors, be they angle rate measurements, or radar "strength" measurements. And its now munged all that "data" into an "estimate", which includes within a circle of X miles, heading direction approximately Y, at speed Z. Just like Mr submarines much larger and more capable computer (cuz hey its a submarine), which of course has the same problem, but only one (ish) sensor to solve it with.... Now disaster strikes... the IRST has been totally and utterly defeated by its greatest nemesis. Mr Fluffy McCloud... "sad bzooop" and R2D noises from Mr. IRST ensue... But you just lost that track. Luckily Mr RWR still works through clouds, but now your track file is only being updated by your absolute worst sensor. And Mr. Mig29 in the MSI computers estimate is coming close enough to detect you. Being a smart and well trained Ace fighter bro, you quickly cue up the track, and point your might Mr. Radar at it. BZZT... Scan, scan, scan, send cancer out. Ping... Ah well there he is, and now he's at bearing 328, but the range is waay off from your RWR estimate and your IRST estimate of range. But hey this is your fancy fire control radar and your #1 most trusted sensor. So now Mr. Radar starts sending Mr. MSI range/bearing/azimuth data. With the standard uncertainty of Mr. Radar (which is really fairly good). And Mr. radar gets weighted MUCH more heavily by Mr. MSI than Mr. RWR who no one trusts, and Mr. IRST that we kinda trust, but he's always a bit dodgy. So now the track files are getting updated quite precisely and you now are starting to build up a very good firing solution on poor Mr. Mig29. Your rather large spheroid of uncertainty of the enemy has now shrunk considerably to a weapons grade firing solution. But wait it gets better, Mr. Cloudy McCloud suddenly isn't in the line of site anymore. And the IRST picks up Mr. mig29 again. And starts sending much more accurate bearing data to Mr. MSI which now knows that Mr. IRST is providing superior bearing and azimuth data to further refine the firing solution to a VERY good firing solution for Mr. AAMRAM... FOX3.... Splash, glory awaits.... So that was the early 90's version of what these systems were capable of doing in general. The F18C has this system, The F15E had this sort of system, the F15C MSIP had this sort of integration, even Blk 40 vipers had some level of this integration. What happens next.... in say 2005... You know "DCS MODERN" Radios is what... (yes turns out "datalinks" are just fancy radios, also turns out "GPS" is a bunch of fancy radios) Now instead of INS with dodgy errors and bad absolute position error data (I don't know where I am, please god help me) I don't know if I'm here or half a km over there plz halps. In 2005 (or earlier) you have INS plus GPS. You know where your ABSOLUTE ownship is at all times within some error margin of well a few meters, or worst case 10-20meters.... So now all your ownship tracks are much more accurate. But your own tracks are just the start. Now instead of just 1 ship. you have 2 ships (or more). Now with a 3 body system, I'm sure the smart guys here will figure out you can do really advanced stuff like... GEOMETRY... Holy cow we can now make "TRIANGLES". Pythagoras would be so proud of us and how we weaponized math. And for those that are in possession of advanced, weapons grade geometry will know, if you have two known points in space, and good angle data, you can fix that 3rd point in space pretty well. But wait it gets better for Mr. MSI... Not only does that work with a 2 ship, it also works even better with different sorts of triangles that you can draw from "multi ship" pictures... And even better Mr. Awacs also gets a vote in the MSI equation (though no one trusts his ass cuz he's using low grade UHF radars with absolute garbage accuracy....) So whats my point with absolute "WALL OF TEXT"... In the context of MSI in DCS. Not ED, not Heablur, not Deka, not anyone has access to the level of information to make even a meaningful guess of how these algorithms and sensors work. And no one in the mil industrial complex will give up ANY details on these types of systems or how they work, even if they are 30 years old or more at this point. "best guesses" from ED or heatblur can be made on "basic physics". But since DCS will not, or cannot actually model "basic physics" of these sorts of systems its gonna pretty pointless beyond very general facts like RWR bad at bearings, IRST good, Radar in-between. And the companies that have developed the multi sensor type inegration algorithms and techniques pointed out here at the kindergarten level of understanding that I have laid out will absolutely laugh in ED/subsidiaries faces when asked about details. So, ask yourself this. Do you really want a "simulator" that can't even begin to simulate how your "uber fighter" actually works in the modern world? And then ask yourself, is that world even simulated all that well, are SAMs as simple as they are in DCS? How "good" is DCS IRST modeling when they currently see through clouds (same question for IR missiles)? OR is actually realistic combat systems much more possible to simulate (and more fun) in the pre 1990 world? This post brought to you by Mr. Coffee null

So, MSI has a long history at least with US planes and its not some singular thing. Rather it is a set of systems and techniques and integrations that evolved over time in parallel to sensor capabilities. So for example MSI on an early 90's hornet is going to be far more primitive and less capable than on say our 2005 era hornet, which due to things like DL integration as well as EGI/GPS integration makes MSI work far better due to the techniques that can be used. Same story for the F15E, and post MSIP 15C that had similar integrations, and even Vipers had it, though I'm not sure exactly how well they were integrated. The EF absolutely has MSI as well and its employment is built around the use of those systems to provide massively superior SA. But again, I doubt HB will even include it, which is an absurd situation. My guess is that the J10 and possibly even the JF17 have very similar capabilities. As for what information has or doesn't have on it IDK. But one of the core difficulties in implementing it is that each sensor or type of signal received is going to have "errors" and how those errors are actually correlated into various types of tracks is likely information they don't have. As well as issues such as tolerance stacking when you might have some sort of info from sensor 1, and sort of correct but conflicting data from sensor two, and then other possible sources of right, but not quite right information from say another flight member. And the fact is DCS sensor modeling as implemented by ED has 0 uncertainty at all. In DCS you are targeting a "point", whereas in the real world your BVR target is basically a "probability spheroid" of where the target is "localized". One example of this is how RAID modes are implemented in the current modules. IRL at long range if you have 2 or more aircraft in a radar range cell, they will show up as a single contact (yay peak top-gun moment right). Well RAID modes/techniques can be employed to figure out if that "single" contact is actually multi-ship formation. But in DCS currently there is no point to raid modes since all targets get their own unique "blip" so you never have to wonder about it.

I mean PESA is certainly better and more complex than MSA radar. For example the Mig31 Zasoln PESA can STT multiple targets for its R-33 missiles. And that is a very primitive radar compared to modern AESAS, and I'm sure there still aren't any real details about it out there despite being from the 80's. I think there is certainly "theory" out there on AESA and the very basics of what it can do, but then that gets to the actual details of can actually do and how well it can do it and so forth. And no one is gonna tell Deka that. So it will be a "made up" radar if they end up doing it. That also doesn't touch on a modern MSI system, or modern IRST system and how those actually function. Again, it would be pure guesswork because once again, no one is giving them modern fighter capabilities. And before someone says I'm being mean to Deka, I will make these same exact arguments against the EF. Personally I'd also rather see a J10A as technologically it fits better with the current crop of "modern" DCS fighters.

The thought was the 10B had some sort of PESA. And the question still stands, Is there enough known about modern PESA/AESA tech to even try to bother model it, or will it be "just made up" with a double dose of handwavium, cuz folks want to have some modern day DCS fantasy? I'll remind you the F18C should have MSI... We don't have MSI despite knowing the basics of how it works. AESA radar systems are a decade newer than that, and for China, nearly 2 decades... So are we gonna get a "realstic" representation of less than 10 year old radar tech in DCS? I'm gonna give that a hard pass, just like anything on the EF being "realistic".

IDK if it had some sort of speed gate, or if some bagguete grade tech still let it work, but yeah in general it shouldn't work well in those environments. Also keep in mind, signal strength in that case is still gonna be relative, i.e. a radar reflective target is still gonna be a stronger return than random ground clutter by some db limit.

Meh, we will see. Maybe the 9.12 can still get made, but I doubt anything newer.

Yeah I get it. Though personally I've pretty much given up on "modern" DCS. Too much modern stuff is unknown, or not modeled or broken with way too many modules. The overall quality of the CW stuff is generally better and getting better from most devs. Plus CW gameplay is way more fun than just slinging aamrams back n forth, depending on how they work on any given patch cycle. But I understand people want 4th gen stuff, even if the gameplay is pretty sus.

J10B was a PESA radar, so similar issues. I mean if Deka can build a convincing AESA radar model maybe. I.e. Razbam is about the only one with a good well explained radar model for the M2k. I see deka has some "features" for the JF17, but really they need a white paper to explain it all IMO.

I'm pretty sure at least the 7F will be an "option" on the 74 F4E we get. Since they were in service after 74...

Yeah IDK on the M sparrow thing. Like I said I've heard it both ways. What you say does make sense though, but I IIRC what I heard was that seeker doesn't require monopulse encoding, but can use it or older schemes or monopulse. IDK if that really tracks though. As for the rest of it yeah, I do think HB will do a decent job with the radar model. And yeah it would be nice if ED modeled some differences between the different seekers at some basic level, the fact that an R3R or 530 is tracking as well as Aim-7M or 7P is well, not particularly realistic.

Yeah fair enough, I haven't been following the EF cuz my interest in it is near zero.