Radar not working after updated

[lionmagunas]

I'm practiced Radar and R3R ten hours ago,worked just fine

 

After today's update I cant get any radar contact in the same mission last night

 

Anyone got this problem or just me?

[2RGT Oberst]

same problem

[rossmum]

Did a quick test in the instant action free flight mission. Radar seemed even more temperamental than usual, wouldn't pick up hot targets especially at low altitude. Climbing a little and looking for cold targets, it seemed to be working normally.

 

I'll have to do some more testing later, it definitely feels like it's not seeing things it should.

[2RGT Oberst]

For this issue, I have opened a support ticket:

 

https://leatherneck-sim.mantishub.io/view.php?id=986

[Sirchuzzalot]

Has there been any updates to this issue? I bought the aircraft on sale and have not been able to use the radar yet.

I did find rolling back nvidia driver helped with the radar lag but it's still not receiving any contacts and is more or less useless at the moment.

[2RGT Oberst]

the problem is still there. The decline in FPS occurs when there are clouds on the radar or you fly at low altitude (visibility of the ground on the radar). At high altitude, planes are seen while at low/medium altitude, you don't see planes.
 

 

Edited December 2, 2020 by 2RGT Oberst

[ataribaby]

With latest patch radar seems work ok again and without FPS bug.

[Sideburns]

In a brief few tests on the cold war server it is nice to see that the MiG21 is now without the FPS hit that has plagued it for over half a year, but it does seem to work remarkably well now at low level. Reliable target detection of Sabre and F5s at max distance, while at low level and with hills in the back ground. Does anyone know what references are being used for how well the MiG21 radar performs? For comparison the Sea Harrier FRS.1 radar is stated as picking up fighter size targets at 8-12 miles and the MiG21 appears to outperform this atm. Granted I need to do some proper testing on this.

[rossmum]

I haven't noticed a substantial difference compared to before, though it does perhaps see things marginally better at low altitude without terrain backing (as long as you're below the contact). Earlier I had an F-5 cause me to lose lock by diving below the horizon and lost several as they dipped lower to the ground. I can definitely say it's affected by closure rate as I've found myself having to use the low speed filter, and being notched, a couple of times over the past week or two.

 

One thing I can say, is that a lot of the time people aren't flying quite as low as they think. Most of the contacts I've spotted on radar, I've then spotted visually, sitting clearly above terrain.

 

Considering the F-5 can detect and lock things up in ground clutter in search mode, let alone its dogfight modes, it's not exactly unique to the 21 - DCS doesn't seem to handle ground clutter that well in general, like the way people used to roll inverted to spot targets below them even at low alt with the 21 (which probably still works).

Edited January 27, 2021 by rossmum

[Sideburns]

3 hours ago, rossmum said:

I haven't noticed a substantial difference compared to before, though it does perhaps see things marginally better at low altitude without terrain backing (as long as you're below the contact). Earlier I had an F-5 cause me to lose lock by diving below the horizon and lost several as they dipped lower to the ground. I can definitely say it's affected by closure rate as I've found myself having to use the low speed filter, and being notched, a couple of times over the past week or two.

 

One thing I can say, is that a lot of the time people aren't flying quite as low as they think. Most of the contacts I've spotted on radar, I've then spotted visually, sitting clearly above terrain.

 

Considering the F-5 can detect and lock things up in ground clutter in search mode, let alone its dogfight modes, it's not exactly unique to the 21 - DCS doesn't seem to handle ground clutter that well in general, like the way people used to roll inverted to spot targets below them even at low alt with the 21 (which probably still works).

 

 

I'm fairly sure closure rate and notching means more to a doppler radar, target aspect and the variations in reflectivity as a result of target aspect would be more important to a pulse radar and possibly it is this variation you are seeing. Typically radar dishes and jet engine compressor blades make for good radar returns in a frontal aspect, planform aspects can also enhance return but unsure how far DCS or the MiG21 goes to model this. On clutter, the target/clutter distinction is a lot clearer in the MiG21, with a distinctive symbol for target, vs the F5 presenting clutter as potential targets so it is a bit trickier in this respect. Regardless if you have an issue with the F5 radar suggest you submit a bug in the F5 section of the forums, otherwise it could be inferred you are trying to balance the two aircraft off each other rather than have them accurately represented in DCS when you take the discussion this way.

 

A quick test in the MiG21 shows it is picking up an F5e at 30km / ~16 miles hot aspect and co-altitude 6200ft as per the attached images. I haven't gone into too deep a dive on the Internet for Sapfir documents but Wikipedia states:

 

"In theory it was able to detect fighter-sized targets from a range of 20 km, and lock on to them at a range of 10 km though in practice this got reduced to 13 km and 7 km respectively."

 

https://en.wikipedia.org/wiki/RP-21_Sapfir

 

Granted these ranges appear to be for the RP-21 rather than RP-21M or RP-22. Best I can find on the RP22 is "Search range 30km, tracking range 15km versus a 16 sq m target." but I'm guessing the F5e is not a 16 sq m target.

 

As before is the MIG21 radar performance in game correct? Feels like it is overperforming at the moment and I don't recall picking up fighter targets at such a range a year ago when I seriously tried to use the MiG21 radar before the FPS issues. Others also seem to have also spotted this.

 

 

[rossmum]

The RP-22 is a pulse-doppler radar, that's what it uses to filter clouds. It just isn't able to fully filter ground clutter. This is why there is a filter specifically for use against low closure rate/low speed targets, and why the radar is able to at least partially filter out weather. It is not a straight pulse radar.

 

Detection ranges (accurate or not, it does seem to pick fighters up quite far out) haven't changed - it has always spotted things at 30km. This isn't a new behaviour, it was doing this before the radar bug set in and has been doing it for years. It's also worth noting that RCS is uniform regardless of aspect in DCS - aircraft have a fixed RCS value. That property isn't unique to the DCS RP-22.

 

A Wikipedia article (about the RP-21, no less - which is not the same radar set) with no citations for the entire relevant paragraph, along with a thread on Hoggit, is not exactly the kind of source I'd be expecting before demanding any changes. I'll have a look through the POH later but I don't remember if it had much in the way of radar performance details, that might be mostly confined to separate documentation.

 

E - as I thought, POH describes controls and combat employment but gives no performance data at all on the set. I have seen some pages out of the German language RP-22SMA documentation floating around somewhere but it was literally a few pages, not the whole manual. I'm pretty sure I saved what I could find, but don't remember where I put it

Edited January 27, 2021 by rossmum

[smilodon]

manuel , page 137

Edited January 28, 2021 by smilodon

[Sideburns]

18 hours ago, rossmum said:

The RP-22 is a pulse-doppler radar, that's what it uses to filter clouds. It just isn't able to fully filter ground clutter. This is why there is a filter specifically for use against low closure rate/low speed targets, and why the radar is able to at least partially filter out weather. It is not a straight pulse radar.

 

Detection ranges (accurate or not, it does seem to pick fighters up quite far out) haven't changed - it has always spotted things at 30km. This isn't a new behaviour, it was doing this before the radar bug set in and has been doing it for years. It's also worth noting that RCS is uniform regardless of aspect in DCS - aircraft have a fixed RCS value. That property isn't unique to the DCS RP-22.

 

A Wikipedia article (about the RP-21, no less - which is not the same radar set) with no citations for the entire relevant paragraph, along with a thread on Hoggit, is not exactly the kind of source I'd be expecting before demanding any changes. I'll have a look through the POH later but I don't remember if it had much in the way of radar performance details, that might be mostly confined to separate documentation.

 

E - as I thought, POH describes controls and combat employment but gives no performance data at all on the set. I have seen some pages out of the German language RP-22SMA documentation floating around somewhere but it was literally a few pages, not the whole manual. I'm pretty sure I saved what I could find, but don't remember where I put it

 

 

Not demanding changes, asking for a confirmation that the radar is performing as intended. Trying to promote a discussion on this following the recent big changes and based on publicly available information. Also as we have seen many times with DCS just because it has been that way before doesn't mean it is correct.

 

I stated the RP21 ranges as some accounts indicate the radars are related. I also provided some RP22 ranges from http://www.flyshark.ayz.pl/Stacja_angielska/sapfir.htm, but you seem to have either overlooked or ignored this. The MiG21 / F5e / typical small fighter plane RCS size is generally considered to be 2-6 sqm. As before this source indicates a 30km search range for a 14sqm bomber size target, which seems reasonable for a 1960s radar and it follows that it should significantly less for a small fighter size target.

 

The general consensus is that the RP22 aka Jaybird was a pure monopulse radar, not pulse doppler (I appreciate monopulse techniques are also used by pulse doppler radars). Monopulse radars are a form of amplitude or phase change comparison radar to accurately point the radar against a target and discern range, popular in the 1960s, the f5e also uses a monopulse radar. The presence of ground returns on the radar screen also suggests a lack of doppler ground return processing, a fairly common feature of airborne pulse doppler radars to enhance their range by filtering out this noise.

 

 

It has been "stated" on that thread, presumably sourced from somewhere, that the RP22 was a simplified version of the early MiG25 Smerch A radar, also a pulse radar, whereas Wikipedia states the RP-22 was development of the RP-21MA. Either way it appears to be pulse radar derivative, with the MiG23 and later MiG25 variants being the introduction of primitive pulse-doppler radars (and obviously some MiG23 export versions being seriously downgraded with the RP21/22).

 

https://forums.eagle.ru/topic/136731-rp-22-sma-sapphir-radar-bug-with-153-update-1/

 

https://www.secretprojects.co.uk/threads/uragan-5b-smerch-smerch-a-radars.23306/

 

Before you continue do you have any evidence or sources that the RP22 was a pulse doppler radar as you state? It would be polite of you to present some sources for your own information before trying to dismiss other's sources so readily. If you have any evidence or sources supporting a 30km range for the RP-22 against fighter targets these would also be useful. The Polish manuals I have found emphasise the MiG21bis radar is still intended for GCI use, rather than independent operations, implying it has poor search and track qualities. Also, as before the Harrier FRS1 Blue Fox monopulse radar, developed in the 70s and being a digital radar system, was picking up fighter targets 8-14 miles. A lesser range than the DCS MiG21bis currently picks up fighter targets.

 

Finally, interested to know the source on "It's also worth noting that RCS is uniform regardless of aspect in DCS"? I wouldn't be surprised if this was the case, mine was originally intended as a suggestion that this might be modelled, but you say it absolutely as if it has been stated somewhere by ED or you have analysed the code?

[Skysurfer]

2 hours ago, Sideburns said:

 

 

 

Finally, interested to know the source on "It's also worth noting that RCS is uniform regardless of aspect in DCS"? I wouldn't be surprised if this was the case, mine was originally intended as a suggestion that this might be modelled, but you say it absolutely as if it has been stated somewhere by ED or you have analysed the code?

No, it literally is in the code. If you actually knew what you are talking about and where to look you'd know. There is no way a simplified desktop sim is going to have a dynamic RCS based on aspect, insidence angle, frequency and closure or whatnot. Every plane in DCS has a fixed RCS, regardless of aspect, ladout or anything else.

Edited January 28, 2021 by Skysurfer

[Sideburns]

55 minutes ago, Skysurfer said:

No, it literally is in the code. If you actually knew what you are talking about and where to look you'd know. There is no way a simplified desktop sim is going to have a dynamic RCS based on aspect, insidence angle, frequency and closure or whatnot. Every plane in DCS has a fixed RCS, regardless of aspect, ladout or anything else.

 

 

No need to try to put people down in your response, I'd also be careful not to assume people's backgrounds and experience.

 

It seems to be repeated a lot that the RCS value is static in DCS but with no reputable source I have found so far. Just wanted to check the source of this information and by my own admission I wouldn't be surprised if DCS did it this way to keep things simple. I would take your direction on where this information is if you want to share? Though having done some analysis of DCS files and submitted a few bug reports I suspect I have probably seen a few of the data files you are thinking of and that the actual radar simulation is buried in a dll / dll's.

 

On RCS aspect simulation you could / I would use a precomputed lookup table to keep it quick and limit it to a few of the factors you mention, such as aspect in the first instance. I know from experience this is realistic to implement on desktop grade computers. It need not be a full simulation as you imply.

[rossmum]

The MiG-21's radar includes a weather filter along with anti-chaff measures and a 'low speed' mode. I had been under the impression (having heard it) that this was achieved through very, very rudimentary PD filtering, but still not sufficient to fully eliminate sidelobe returns at low altitudes. Perhaps I heard wrong and it is indeed straight pulse, at which point who knows what it's using to filter weather or passive (chaff) countermeasures, and why it needs a "low speed" filter. In any case, in DCS, the way the radar is modelled makes it possible to notch, although relatively uncommon.

 

I have no evidence of the radar showing fighters at 30km because I have made no assertion it's accurate - I simply stated that it's always been that way in DCS. This is not a recent development. It probably is too sensitive, I wouldn't be surprised, but unfortunately legitimate RP-22 documentation is near impossible to find without a very large sum of money or knowing the right people or we wouldn't be having this conversation and the radar wouldn't still include several pieces of guesswork or gameplay abstractions (locking in fixed beam mode, for a start).

 

I did go hunting through the Russian forums and found a former 21 pilot saying that the only time he was able to successfully see targets were from astern at 20km when talked onto them, but that was about the sum of his discussion of the matter. No mention of what the target was (one could assume a practice reflector or perhaps a bomber, but that's too big of an assumption to make). Without the actual manual and/or seriously in depth engineering documents, or a working example, there's not a whole lot to go on and making changes based on vague pilot testimony is not exactly a good idea.

 

The radar being intended for GCI use is neither here nor there, it means absolutely nothing. Most radars of the time had fairly poor independent search and if you think only the Soviets relied on aircraft being directed onto targets, I don't know what to tell you. Comparisons to the Blue Fox are vaguely useful at best because it is a British radar, not a Soviet one, and other than the most basic operating principles it is unlikely to have much in common with the RP-22. This is dangerously close to "well this aircraft has X flight characteristic and looks kinda like that aircraft, so it must also therefore have X flight characteristic", which these forums have far too much of already.

 

Source on RCS being constant is hearing it from people who have a lot more invested in figuring out the game's mechanics than I do (including modders). I'm not going to pretend I've seen the code, I wouldn't know what I'm looking at anyway, but the people who have stated it would have no reason to throw that one out for funsies and I have come across threads on these very forums which show how the IR signatures are calculated - as a multiplier against an arbitrary measurement value, that being an AL-31F at military thrust. If the sim is using the Flanker's engine as a yardstick for IR signatures, I wouldn't be very surprised by RCS being a per-aircraft value.

 

Anyway, with all that said - I do actually think the radar is overperforming slightly, and should probably not pick things up quite as far away as it does. This is not a new factor and it has always detected targets at this range, so I'm not sure why it wasn't brought up earlier. The problem is finding out the margin and using either single anecdotes or the figures for another set, or estimates off of the one semi-detailed English language page, isn't really a great idea because it becomes very easy to end up with yet another value that swings one way or the other wildly and which nobody is ever happy with, in a sim where most equipment is assumed to always function to the best of its design specs. Radar performance in DCS is universally messy and even radar sets we have better information for either overshoot or undershoot their known performance. Realism is an important goal, but making guesses on the basis of "well this predecessor radar is similar enough" or "this foreign counterpart is probably about the same" is not realism any more than taking the max assured detection range of a target as gospel. God forbid we start applying this to FMs or the mechanics of the aircraft's controls as well.

 

Performance over ground clutter seems unchanged from how it's always been, the possible exception being in situations where there are so many contacts on scope that the ground clutter thins out or stops rendering (a side effect of the FPS drop fix, as far as I know). I don't know if this is actually happening or not, or how consistently, so it needs further testing. In any case, the radar is still unable to effectively look down and still reliably loses lock when a target dives for the floor. Every target I have successfully locked was either above me by a decent margin, or clear above the highest local terrain feature.

 

On the plus side, I've learnt a few interesting things while combing the Russian side of the forums and solved something else that was a bit of a sticking point relating to the FM.

[Conker4]

TLDR : Radar is unchanged and is working as it always has. As for the pilot in the video, they were using raw radar instead of comp. Simple pilot error.

[KAKTYC_11]

Radar is broken again , this time -at low altitudes . It causes no effect on radar screen after turn it at low altitude mode

[Frederf]

On 1/27/2021 at 5:57 PM, rossmum said:

The RP-22 is a pulse-doppler radar,  This is why there is a filter specifically for use against low closure rate/low speed targets, and why the radar is able to at least partially filter out weather. It is not a straight pulse radar.

 

This is false. RP-22 is not Doppler. It's a monopulse radar. The filter is there to reject chaff.

[Sideburns]

On 1/28/2021 at 11:50 PM, rossmum said:

The MiG-21's radar includes a weather filter along with anti-chaff measures and a 'low speed' mode. I had been under the impression (having heard it) that this was achieved through very, very rudimentary PD filtering, but still not sufficient to fully eliminate sidelobe returns at low altitudes. Perhaps I heard wrong and it is indeed straight pulse, at which point who knows what it's using to filter weather or passive (chaff) countermeasures, and why it needs a "low speed" filter. In any case, in DCS, the way the radar is modelled makes it possible to notch, although relatively uncommon.

 

I have no evidence of the radar showing fighters at 30km because I have made no assertion it's accurate - I simply stated that it's always been that way in DCS. This is not a recent development. It probably is too sensitive, I wouldn't be surprised, but unfortunately legitimate RP-22 documentation is near impossible to find without a very large sum of money or knowing the right people or we wouldn't be having this conversation and the radar wouldn't still include several pieces of guesswork or gameplay abstractions (locking in fixed beam mode, for a start).

 

I did go hunting through the Russian forums and found a former 21 pilot saying that the only time he was able to successfully see targets were from astern at 20km when talked onto them, but that was about the sum of his discussion of the matter. No mention of what the target was (one could assume a practice reflector or perhaps a bomber, but that's too big of an assumption to make). Without the actual manual and/or seriously in depth engineering documents, or a working example, there's not a whole lot to go on and making changes based on vague pilot testimony is not exactly a good idea.

 

The radar being intended for GCI use is neither here nor there, it means absolutely nothing. Most radars of the time had fairly poor independent search and if you think only the Soviets relied on aircraft being directed onto targets, I don't know what to tell you. Comparisons to the Blue Fox are vaguely useful at best because it is a British radar, not a Soviet one, and other than the most basic operating principles it is unlikely to have much in common with the RP-22. This is dangerously close to "well this aircraft has X flight characteristic and looks kinda like that aircraft, so it must also therefore have X flight characteristic", which these forums have far too much of already.

 

Source on RCS being constant is hearing it from people who have a lot more invested in figuring out the game's mechanics than I do (including modders). I'm not going to pretend I've seen the code, I wouldn't know what I'm looking at anyway, but the people who have stated it would have no reason to throw that one out for funsies and I have come across threads on these very forums which show how the IR signatures are calculated - as a multiplier against an arbitrary measurement value, that being an AL-31F at military thrust. If the sim is using the Flanker's engine as a yardstick for IR signatures, I wouldn't be very surprised by RCS being a per-aircraft value.

 

Anyway, with all that said - I do actually think the radar is overperforming slightly, and should probably not pick things up quite as far away as it does. This is not a new factor and it has always detected targets at this range, so I'm not sure why it wasn't brought up earlier. The problem is finding out the margin and using either single anecdotes or the figures for another set, or estimates off of the one semi-detailed English language page, isn't really a great idea because it becomes very easy to end up with yet another value that swings one way or the other wildly and which nobody is ever happy with, in a sim where most equipment is assumed to always function to the best of its design specs. Radar performance in DCS is universally messy and even radar sets we have better information for either overshoot or undershoot their known performance. Realism is an important goal, but making guesses on the basis of "well this predecessor radar is similar enough" or "this foreign counterpart is probably about the same" is not realism any more than taking the max assured detection range of a target as gospel. God forbid we start applying this to FMs or the mechanics of the aircraft's controls as well.

 

Performance over ground clutter seems unchanged from how it's always been, the possible exception being in situations where there are so many contacts on scope that the ground clutter thins out or stops rendering (a side effect of the FPS drop fix, as far as I know). I don't know if this is actually happening or not, or how consistently, so it needs further testing. In any case, the radar is still unable to effectively look down and still reliably loses lock when a target dives for the floor. Every target I have successfully locked was either above me by a decent margin, or clear above the highest local terrain feature.

 

On the plus side, I've learnt a few interesting things while combing the Russian side of the forums and solved something else that was a bit of a sticking point relating to the FM.

 

It feels like you are attempting to introduce strawmen to the argument? For what it is worth I think the F5e radar might be also overperforming as well based on some basic research.

 

The comparison with the Sea Harrier radar is because it is also a small dish, light fighter monopulse radar but with a more modern design and digital signal processing, so you would expect the Sea Harrier radar to perform better whereas this doesn't appear to be the case, hence the comparison. I also note some modules do model degradation of radar based on the environment and usage, so this can be modelled developer dependent and not all radars in game are perfectly operating all the time. On the GCI comment atm the MiG21 radar performs well enough that you can use it as a search radar fairly reliably. If the real life MiG21 radar performed this well I suspect its use would not be as dependent on GCI tactics for air to air.

 

@Hiromachior @-Rudel- any chance of a comment on the MiG21's radar performance and if it is considered to be in the right place at the moment. I know there have been some fairly big changes to the radar recently, are further developments and adjustment ongoing?

Edited February 5, 2021 by Sideburns
Typos

[-Rudel-]

@Dolphin887 or @rolanderowill be better suited to answer.

[2RGT Oberst]

the radar in the map of Syria still gives problems of FPS drops

[Katte]

hello to all, I also confirm that the radar of the mig 21 has a big drop in fps in the map of Syria

brgds

[Koty]

On 2/2/2021 at 10:24 PM, Frederf said:

This is false. RP-22 is not Doppler. It's a monopulse radar. The filter is there to reject chaff.

Those two are not exclusive.

The "filter" that is there to reject chaff is a doppler filter  monopulse technique cannot do that. What it can do however is help the radar reduce side lobe clutter. Which then works alongside the doppler clutter rejection filter.

 

Mind you, RP-21 already had a low altitude clutter reducing filter (it would still get overwhelmed if low enough). RP-22 added weather rejection and chaff rejection. All working with doppler information. 

 

There's also claims that you cannot have a PD radar without hight PRF, which is just false - all it is is impractical, as you get velocity ambiguity. Dealing with ambiguity requires a lot of computing power - which is why you only see a fully efficient low PRF PD filter on RP-23 (and even there you reduce radar sensitivity). (Which is where Soviets decided that it is a dead end and used more standard Medium/High PRF doppler radars from there on, except on the MiG-25 which received an upscaled RP-23...). And with analogue computers like on the RP-21 and 22, not only have you reduced sensitivitty, it also lets through a lot of clutter simply because the computer is not precise enough.

 

Going back to range, yes, 30km is vs a 16 m^2 target - since we know this, the beam width and I believe I had emitting power somewhere, it won't be difficult to calculate at what range roughly should you see an F-5. But does anyone know its RCS in DCS?

 

EDIT: Got the figures from Hiro. After doing the math, you should see the F-5, with 5m^2 RCS, at 22.5-23 km.

Edited September 24, 2021 by Koty

[Sideburns]

57 minutes ago, Koty said:

Those two are not exclusive.

The "filter" that is there to reject chaff is a doppler filter  monopulse technique cannot do that. What it can do however is help the radar reduce side lobe clutter. Which then works alongside the doppler clutter rejection filter.

 

Mind you, RP-21 already had a low altitude clutter reducing filter (it would still get overwhelmed if low enough). RP-22 added weather rejection and chaff rejection. All working with doppler information. 

 

There's also claims that you cannot have a PD radar without hight PRF, which is just false - all it is is impractical, as you get velocity ambiguity. Dealing with ambiguity requires a lot of computing power - which is why you only see a fully efficient low PRF PD filter on RP-23 (and even there you reduce radar sensitivity). (Which is where Soviets decided that it is a dead end and used more standard Medium/High PRF doppler radars from there on, except on the MiG-25 which received an upscaled RP-23...). And with analogue computers like on the RP-21 and 22, not only have you reduced sensitivitty, it also lets through a lot of clutter simply because the computer is not precise enough.

 

Going back to range, yes, 30km is vs a 16 m^2 target - since we know this, the beam width and I believe I had emitting power somewhere, it won't be difficult to calculate at what range roughly should you see an F-5. But does anyone know its RCS in DCS?

 

 

Do you have any sources for your RP22 doppler filter information? Interested as I find the "techniques" of radar fascinating as a physics grad, I found a few sources on the Rp22, mostly East German and Polish operations manuals with some low level details.

 

Also, on RCS (figures from version 2.5)