Rlaxoxo Posted December 1, 2016 Posted December 1, 2016 What happens with two consecutive launches at less than 10 seconds apart? Are two lines drawn, or does the second launch override the first one visually? Also, why was a line chosen instead of a symbol like the A-10C does? Good question [sIGPIC][/sIGPIC] Youtube Reddit
Zeus67 Posted December 1, 2016 Author Posted December 1, 2016 What happens with two consecutive launches at less than 10 seconds apart? Are two lines drawn, or does the second launch override the first one visually? A line is drawn for each missiles. In your case two lines are drawn. Also, why was a line chosen instead of a symbol like the A-10C does? Because French? ;) Actually because the IR sensor cannot determine range just azimuth and perhaps signal strength. Another reason is that I don't want to copy the A-10C instruments either. The functionality is the same but interface should be different. Remember that these are highly classified systems. Any implementation, is just pure guess work or as I call it WAGUESS. "Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning." "The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch and a user with an idea."
myHelljumper Posted December 1, 2016 Posted December 1, 2016 I like the line :thumbup: Helljumper - M2000C Guru Helljumper's Youtube https://www.youtube.com/channel/UCK3rTjezLUxPbWHvJJ3W2fA
Eddie Posted December 1, 2016 Posted December 1, 2016 A line on the relevant azimuth is not at all unreasonable. ;)
tombeckett2285 Posted December 1, 2016 Posted December 1, 2016 A line on the relevant azimuth is not at all unreasonable. ;) /OFFTOPIC AVIONICS GEEK ALERT/ The ARI 18216 Wideband Homer on the Buccaneer functioned much the same, except it was based on radar transmissions. When transmissions were detected a line illuminated on the CRT along the azimuth that the transmissions were being detected from, along with an audio alert. 12 Sqn retained these even after the introduction of conventional RWR equipment because of the superior range detection of the WBH equipment for anti-ship strikes. /OFFTOPIC ENDS/ ;) Looks great guys, a fantastic example in both customer service interaction and programming ingenuity. Look forward to testing the update! Sent from my iPhone using Tapatalk "The only replacement for a Buccaneer is a Buccaneer".
Maulkin Posted December 1, 2016 Posted December 1, 2016 I think the line is a wonderful idea. I cannot wait to test this out. Imagine my embarrassment when I thought it was already implemented and I was trying to test this in Nevada with no results (and a very frustrated Igla launcher). :) --Maulkin Windows 10 64-bit - AMD Ryzen 9 5900X @ 3.7 GHz - 32 GB DDR4 3600MHz RAM - EVGA FTW3 RTX 3080 - Asus Crosshair VIII Hero motherboard - Samsung EVO Pro 1 TB SSD - TrackIR 4 Pro - Thrustmaster Warthog - Saitek rudder pedals - Lilliput UM-80/C with TM Cougars
Schmidtfire Posted December 2, 2016 Posted December 2, 2016 Spot on! Great idea with the line. (old-school RWR's used lines for radar contacts instead of symbols)
Flagrum Posted December 2, 2016 Posted December 2, 2016 I like the line symbology! But I still wonder, if such representation isn't maybe a bit too acurate? I mean, the exact bearing - are the sensors really capable of that, in terms of precision?
Zeus67 Posted December 2, 2016 Author Posted December 2, 2016 I like the line symbology! But I still wonder, if such representation isn't maybe a bit too acurate? I mean, the exact bearing - are the sensors really capable of that, in terms of precision? Yes they are. After all, the sensors are not different from the ones that are on the seeker heads of IR missiles. "Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning." "The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch and a user with an idea."
zaelu Posted December 2, 2016 Posted December 2, 2016 (edited) Yes they are. After all, the sensors are not different from the ones that are on the seeker heads of IR missiles. The only difference is the seeker head is actively pointed towards the target and kept like this for few moments before launch. A maneuvering plane (with an unaware pilot) would have little chance of detecting with precision the direction of launch if not the launch at all. It's like trying to lock an IR missile on a target that moves all over the sky in front of the plane. IMHO... taking in consideration all the story about this, those sensors would in best case detect a launch or not and not point to it instantly as it will need quite of work for this (neither the seeker head do that it just uses a feedback mechanics to turn the missile so it will point towards the target after launch). I think something like the SPO-10 from Mig21Bis would be best case (meaning just 4 cardinal points detectors and overlapping in their alerts). I have my reservations about this system implemented with so many "shortcuts" towards "arcade" (not implemented in C probably for good reasons not "money", 3 sensors on each side instead of one, laser precision and probably instant reaction, probably impossible to disable in MP games) not being a "can of worms" for multiplayer. But... sometimes we need to see before believing. Edited December 2, 2016 by zaelu [sIGPIC][/sIGPIC] I5 4670k, 32GB, GTX 1070, Thrustmaster TFRP, G940 Throttle extremely modded with Bodnar 0836X and Bu0836A, Warthog Joystick with F-18 grip, Oculus Rift S - Almost all is made from gifts from friends, the most expensive parts at least
T_A Posted December 2, 2016 Posted December 2, 2016 I like the line symbology! But I still wonder, if such representation isn't maybe a bit too accurate? I mean, the exact bearing - are the sensors really capable of that, in terms of precision? Why too accurate? the plane has all the variables it needs to show a pretty real direction. 1. Plane`s own attitude 2. IR seeker is basically a camera , it knows that the Missile is coming from an azimuth X with elevation Y and those are constants relative to the plane. the computer just needs to add that to the plane`s attitude and it will show where the missile is on the RWR Accuracy depends on distance , size of IR signature , resolution and IR sensitivity of the seeker. IAF.Tomer My Rig: Core i7 6700K + Corsair Hydro H100i GTX Gigabyte Z170X Gaming 7,G.Skill 32GB DDR4 3000Mhz Gigabyte GTX 980 OC Samsung 840EVO 250GB + 3xCrucial 275GB in RAID 0 (1500 MB/s) Asus MG279Q | TM Warthog + Saitek Combat Pedals + TrackIR 5 [sIGPIC][/sIGPIC]
zaelu Posted December 2, 2016 Posted December 2, 2016 How do you think it looked the image from those sensors from '90s and what computer-software was available to make such accurate detection? Do you think it had more than 320x240@30fps resolution and a 386 behind it? What lenses do you thing it used (if any)? [sIGPIC][/sIGPIC] I5 4670k, 32GB, GTX 1070, Thrustmaster TFRP, G940 Throttle extremely modded with Bodnar 0836X and Bu0836A, Warthog Joystick with F-18 grip, Oculus Rift S - Almost all is made from gifts from friends, the most expensive parts at least
T_A Posted December 2, 2016 Posted December 2, 2016 I don't know , but who`s to say it wasn't analogue, or that there aren't several seekers chained together, these systems are classified , we may only sometimes get a clue as to their general workings but not the technique and components that were used. and generally speaking - military technology especially back then , was a years ahead of anything you could find on the civilian market. it`s true that some systems that are simulated in DCS work a little 'too good' the M2K will not be the first. IAF.Tomer My Rig: Core i7 6700K + Corsair Hydro H100i GTX Gigabyte Z170X Gaming 7,G.Skill 32GB DDR4 3000Mhz Gigabyte GTX 980 OC Samsung 840EVO 250GB + 3xCrucial 275GB in RAID 0 (1500 MB/s) Asus MG279Q | TM Warthog + Saitek Combat Pedals + TrackIR 5 [sIGPIC][/sIGPIC]
lemoen Posted December 2, 2016 Posted December 2, 2016 How do you think it looked the image from those sensors from '90s and what computer-software was available to make such accurate detection? Do you think it had more than 320x240@30fps resolution and a 386 behind it? What lenses do you thing it used (if any)? Missile seekers don't all work that way. The missile usually stays upright (using spinerons) and the seeker head spins. So if the heat source is off centre, it'll wobble about on the detector, so you can determine (without using a computer) in which direction the heat source is, because you know the seeker's angle at which the source is a maximum. Look up "Spin Scan". Kinda like the reverse of a beam rider. DISCLAIMER: I don't know if the DDM works like that, just pointing out you don't need a camera and a computer to do these things.
microvax Posted December 2, 2016 Posted December 2, 2016 How do you think it looked the image from those sensors from '90s and what computer-software was available to make such accurate detection? Do you think it had more than 320x240@30fps resolution and a 386 behind it? What lenses do you thing it used (if any)? Nobody would use a CPU, simply nobody. You would use some kind of silicone which has exactly that one job to identify something worth tracking. That would then report to a CPU. Major criteria would be the LOSR but hey. We are all just forum warriors. What I can tell you, is that the computing power should not be the problem to realize something like 5-10hz update rate. Today so much computing power gets thrown out of the window, its really unreal. [sIGPIC][/sIGPIC] *unexpected flight behaviour* Oh shiii*** ! What ? Why ? What is happening ?
red_coreSix Posted December 2, 2016 Posted December 2, 2016 The D2M should be pretty accurate for angle tracking as are all modern IR sensors. It's much more likely that the RWRs we have in DCS are way too accurate. IR is bad for ranging and radar bad for angle tracking (comparatively obviously).
zaelu Posted December 2, 2016 Posted December 2, 2016 Nobody would use a CPU, simply nobody. You would use some kind of silicone which has exactly that one job to identify something worth tracking. That would then report to a CPU. Major criteria would be the LOSR but hey. We are all just forum warriors. What I can tell you, is that the computing power should not be the problem to realize something like 5-10hz update rate. Today so much computing power gets thrown out of the window, its really unreal. I was just trying to make a comparison, obviously it wouldn't use something like "Fenêtres '95", etc. :D About computing power thrown out... that's what we all do with our brains isn't it? :) [sIGPIC][/sIGPIC] I5 4670k, 32GB, GTX 1070, Thrustmaster TFRP, G940 Throttle extremely modded with Bodnar 0836X and Bu0836A, Warthog Joystick with F-18 grip, Oculus Rift S - Almost all is made from gifts from friends, the most expensive parts at least
jojo Posted December 2, 2016 Posted December 2, 2016 All these talks are useless. Zeus did what he could with a code designed to simulate A-10C MLWS. Mirage fanatic ! I7-7700K/ MSI RTX3080/ RAM 64 Go/ SSD / TM Hornet stick-Virpil WarBRD + Virpil CM3 Throttle + MFG Crosswind + Reverb G2. Flickr gallery: https://www.flickr.com/gp/71068385@N02/728Hbi
Flagrum Posted December 2, 2016 Posted December 2, 2016 I would imagine, such sensors are rather conservative in their design (rather "electrical" than "electronical"?). For reliability reasons, for cost reasons and for lack-of-necessity reasons. I mean, there is no point for great accuracy for such sensors, is there? It doesn't help the pilot if he knows that the missile is coming at him at 99 degrees. All he needs to know is "MISSILE FROM RIGHT!"
jojo Posted December 2, 2016 Posted December 2, 2016 I would imagine, such sensors are rather conservative in their design (rather "electrical" than "electronical"?). For reliability reasons, for cost reasons and for lack-of-necessity reasons. I mean, there is no point for great accuracy for such sensors, is there? It doesn't help the pilot if he knows that the missile is coming at him at 99 degrees. All he needs to know is "MISSILE FROM RIGHT!" It's never too accurate. For instance, DDM NG on Rafale is said to be accurate enough to point DIRCM. Mirage fanatic ! I7-7700K/ MSI RTX3080/ RAM 64 Go/ SSD / TM Hornet stick-Virpil WarBRD + Virpil CM3 Throttle + MFG Crosswind + Reverb G2. Flickr gallery: https://www.flickr.com/gp/71068385@N02/728Hbi
Zeus67 Posted December 3, 2016 Author Posted December 3, 2016 I would imagine, such sensors are rather conservative in their design (rather "electrical" than "electronical"?). For reliability reasons, for cost reasons and for lack-of-necessity reasons. I mean, there is no point for great accuracy for such sensors, is there? It doesn't help the pilot if he knows that the missile is coming at him at 99 degrees. All he needs to know is "MISSILE FROM RIGHT!" The A-C's Bitching Betty does that. It just tells you where the missile is coming from. "Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning." "The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch and a user with an idea."
zaelu Posted December 3, 2016 Posted December 3, 2016 I think only in "game" (arcade) mode not "simulator". [sIGPIC][/sIGPIC] I5 4670k, 32GB, GTX 1070, Thrustmaster TFRP, G940 Throttle extremely modded with Bodnar 0836X and Bu0836A, Warthog Joystick with F-18 grip, Oculus Rift S - Almost all is made from gifts from friends, the most expensive parts at least
Maulkin Posted December 5, 2016 Posted December 5, 2016 Saw in CF's latest video that this is now working in v1.5.5: Does anyone know if the pylon model for the Magic shows the sensors? --Maulkin Windows 10 64-bit - AMD Ryzen 9 5900X @ 3.7 GHz - 32 GB DDR4 3600MHz RAM - EVGA FTW3 RTX 3080 - Asus Crosshair VIII Hero motherboard - Samsung EVO Pro 1 TB SSD - TrackIR 4 Pro - Thrustmaster Warthog - Saitek rudder pedals - Lilliput UM-80/C with TM Cougars
grunf Posted December 5, 2016 Posted December 5, 2016 Saw in CF's latest video that this is now working in v1.5.5: Does anyone know if the pylon model for the Magic shows the sensors? No, they don't show the sensors yet.
Zeus67 Posted December 13, 2016 Author Posted December 13, 2016 DDM test flight: I made this video using one of Esac_mirmidon's track files. It shows that the missile azimuth tracking is working and it also answers other observations made in the bug thread. "Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning." "The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch and a user with an idea."
Recommended Posts