MAV IR smoke image

[some1]

Latest openbeta. Instead of the smoke on the IR display there are white blobs.

Deleted the shaders, no change.

 

[Harlikwin]

Well in general thermal imagers should see through most smoke... That was one their "features" as used on tanks...

[NineLine]

Well in general thermal imagers should see through most smoke... That was one their "features" as used on tanks...

 

Through smoke generate by smoke dispensers, smoke from a fire or explosion with highly heated air would be a different story. As well, a new FLIR rendering system is being worked on and things like this would likely be tuned better.

[Fri13]

Well in general thermal imagers should see through most smoke... That was one their "features" as used on tanks...

 

Only the smoke grenades and generated smoke by injecting fuel directly to engine exhaust.

 

But for long time all the smoke covers the vehicle generates are IR and Laser blocked. You can't see through with TIS and you can't paint the target with laser as the smoke reflects the laser all over the places and missiles goes crazy. You can't range the target either as you get too many echoes and speed variations all over the places (same as chaff for radar) so you can't use any laser ranging to get your launch solution (guns, rockets etc).

 

And when you have a burning element that generates smoke, you can't see through it with TIS. That was already challenge in the first Iraq where military set the oil wells on fire, and all the smoke generates such thermal blockage that TIS and lasers couldn't be used. It was visual only flying there.

 

The same thing is with clouds, too much moisture and they block all the IR radiation and your laser range finder as well loses its power very quickly. So dog fights with gun metering range with laser, IR missiles etc are unusable when target is in the clouds or behind it.

[some1]

Looks like the FLIR / MAV IR in A-10C is completely broken. The smoke blobs are pitch black (in WHOT!) and whiteout the whole screen as the smoke gets in the field of view.

 

[Puma]

Its a new struggle for sure. I am also not convinced it wasn't an attempt to be more realistic. It feels like it may be a bit too far, but either way its hard to work with. Wish we got a better change description to know if this is a bug or a reality check.

[schurem]

a new FLIR rendering system is being worked on and things like this would likely be tuned better.

 

Any chance of progress updates on this? What are we looking at? Months? Years? Weeks?

[NineLine]

I don't have any info regarding timeline right now.

[Harker]

Looks like the FLIR / MAV IR in A-10C is completely broken. The smoke blobs are pitch black (in WHOT!) and whiteout the whole screen as the smoke gets in the field of view.

 

I had this until tonight. I checked again and my IR in all modules has returned to the "old" state. Anyone else has the same?

[Harlikwin]

Well, one of the "features" of FLIR and thermal imagining in general is the ability to see through smoke...

[some1]

I had this until tonight. I checked again and my IR in all modules has returned to the "old" state. Anyone else has the same?

 

Think it's been fixed in today's update.

[EasyEB]

 

I'd say it's gotten worse.

[Harlikwin]

Through smoke generate by smoke dispensers, smoke from a fire or explosion with highly heated air would be a different story. As well, a new FLIR rendering system is being worked on and things like this would likely be tuned better.

 

No offense 9L, but I've looked at fires through FLIR... There is a reason fire departments use FLIR, and a big part of it is that it sees through smoke, you know the burning kind.....

 

Maybe this can help your code team. Particularly the bit starting a 0.45 where it shows a side by side visual and FLIR camera, no smoke in the FLIR.

 

 

Edited May 9, 2019 by Harlikwin

[dcs76]

It depends on what the threshold of the IR sensors in WHOT (or BHOT) mode is. The IR sensor and TGP show everything white which has a temperature above a specific threshold that the pilot is able to locate and identify vehicles with running engines. An explosion of a bomb/rocket has definitely a higher temperature as this vehicle. So it is totally normal that also this cloud is shown in white (WHOT) and even blinds the rest of the screen.

 

Same Might also be said for smoke generated by WP. Phosphorus is burning at quite a high temperature and part of this heat energy is being carried by the smoke itself. I cannot say how hot this smoke will actually be though. Not sure if this would blind an IR sensor.

 

The example and reasoning given that IR cameras are used to look through smoke is just a matter of where the threshold for showing things white/bright is. For a burning building they would usually have this threshold very high to only see the fire through the clouds. As you watch the video, you can see that the firefighters are actually black (pretty much as black as the background). In the FLIR of the TGP those firefighters would be visible as white as the threshold there is different.

Edited May 11, 2019 by dcs76

[Harlikwin]

Most FLIR sensors tend to "auto calibrate" since the dynamic range of the commonly used vanadium oxide detectors is rather limited the algorithm tends to pick whatever gives the scene the best dynamic range. So when looking at something like a fire it would show firefighters as "cold". But in the presence of no fire they would look hot. Its been a very long time since you had to/could do that manually with most FLIR systems (like the 70's) I can do the experiment with my PAS-7 and report back, but at guess its gonna see right through the smoke. That being said, I'd just be happy if ED managed to not show "smoke" with FLIR, the reason you don't see it is due to the small size of the smoke particles versus the wavelength of the IR band, plus smoke tends to cool rapidly. Hold your hand 3" above a candle and then 1 foot higher to experiment with that.

 

Here is some very light technical reading for anyone interested.

 

https://www.flirmedia.com/MMC/CVS/Tech_Notes/TN_0001_EN.pdf

Edited May 12, 2019 by Harlikwin

[KTFBGB]

Most FLIR sensors tend to "auto calibrate" since the dynamic range of the commonly used vanadium oxide detectors is rather limited the algorithm tends to pick whatever gives the scene the best dynamic range. So when looking at something like a fire it would show firefighters as "cold". But in the presence of no fire they would look hot. Its been a very long time since you had to/could do that manually with most FLIR systems (like the 70's) I can do the experiment with my PAS-7 and report back, but at guess its gonna see right through the smoke. That being said, I'd just be happy if ED managed to not show "smoke" with FLIR, the reason you don't see it is due to the small size of the smoke particles versus the wavelength of the IR band, plus smoke tends to cool rapidly. Hold your hand 3" above a candle and then 1 foot higher to experiment with that.

 

Here is some very light technical reading for anyone interested.

 

https://www.flirmedia.com/MMC/CVS/Tech_Notes/TN_0001_EN.pdf

 

^^^^

This. I am retired from the fire dept. Every engine carried FLIR, or a TIC as we called it. I was a Recruit Training Officer and trained a lot of guys how to use these. Even inside a burning structure where you can’t see 6inches in front of your face, with a temp differential of 1200 at the ceiling, and 350 at the floor, the cameras auto calibrate and you can see everything. Even when going from say a hallway where there is good temp variation for the camera to pick up, to a room fully involved where there is very little temp differential, all you get is a momentary flash on the camera then the screen updates with the new calibration and you are back in business.

[QuiGon]

Did some LANTIRNing with the F-14 yesterday and my pilot did have the same problem on his LANTIRN display. On my RIO LANTIRN display everything was fine though :dunno:

[randomTOTEN]

I'd say it's gotten worse.

I don't think i've seen intensity calibration in DCS before. This looks like new tech. Is this the early stages of the new IR render?:thumbup:

[Nealius]

Can someone list which specific version(s) of DCS has this issue? I'm currently on 30386 and do not experience this on the A-10s Maverick or TGP screens. I did not have this problem on my previous version either, but I skipped one or two updates prior to 30386.

[Gierasimov]

I see this too in the latest beta. Solid smoke through MAV F on DDI in Hornet. Same for the smoke marker. Interestingly Static Object - Big smoke (with preset of a small fire and smoke) renders differently - kind of like it was rendering so far.

[Nealius]

This is a very strange bug. I encountered it for the first time in the A-10C using the TGP, after flying many hours and never encountering this bug. After clearing my fxo/metashaders2 folders, and changing my graphics settings, I replayed the track file of the exact same flight for some video footage. During the replay track, the smoke bug was not present.

[9echo]

I have the same issue still with 2.5.4.30386 updated two days ago. AT the sime time of updating this I've experimented with Nvidia Profile Inspector settings following recommendations on the forum here so as to enhance performance, so not sure if it is the update or the experimentations that has messed it up. Is there a fix? Can't fly with this, it's an absolute nightmare to do missions. SPI one MBT, whack it, want to slew to the next for immediate engagement with AGM and can't see sh*t! :doh: :helpsmilie:

 

@Nealius: Did you simply delete the entire Bazar/shaders/fxo and Bazar/shaders/metashaders folders? Have you tested it flying a live flight, not just replaying the track?

Edited May 31, 2019 by 9echo

[Ironwulf]

in the A-10C you can play with the gain to get a slightly better image but gain is not modelled in the Hornet, for example.

 

Probably need to have an automatically adjusted gain....

 

In any case smoke from a smoke marker should not be 'hot' - perhaps its source canister might be, but the smoke it self isn't so hot it would cause such a huge issue with FLIR

[Ironwulf]

I should note this is also a problem in the T-Pod of the harrier...

 

Anyway, here's a video I am sure most of us have seen from an AC-130 showing real explosions in a FLIR, and the resultant smoke... neither of which 'blind' the sensor.

 

[The RDN]

Hi everyone!

 

I have this issue too… I have hope that it will be corrected with OB 2.5.5 but no… It seems that this problem appeared after version 2.5.4.29167. I had to leave my stable version in 2.5.4.29167 in order to be able to use descently my A10c / AV8B / F18C / F14 modules.

What I find annoying is that there is no official reaction to this topic to know if this bug is taken into account or if we will have to be dealt with. Can somebody say if is that bug reported or not? Thanks to all...

 

Sorry for my basic English :-P