correct as is T-72B3 (All Russian MBTs, BMP-2/3, Russian Thermal) cannot spot at night.

[Shadow KT]

T-72B3 is supposed to have the latest generation thermal imaging, yet it cannot spot units at night, unless they are firing at it or around 1KM.

Although, all these vehicles I am about to list have different thermal capabilities in real life, they have pretty similar spotting ranges at night in DCS, from 0.9 to 1.4KMs. These include: T-90, T-80, T-72B, T-72B3, T-55, BMP(D)-1/2/3

Meanwhile M1A2, Leopard-2 and Challenger 2 all have a 3.9KM + spotting distance at night. Althoughthese vehicles do have better thermals, than the older russian tech, it is not that much better, especially compared to the T-72B3/90/80. Even the M2A2 Bradley outspots all these vehicles with a spotting distance of around 1.5KMs. Either the NATO MBTs have to be nerfed, or the Russian ones need a buff.... or best, both... as the truth is somewhere in the middle.

The British MCV-80, which doesn't even have thermals in game... has a spotting range of 3+ KMs.

This creates a different problem. As I mentioned in the beginning, the Russian MBTs can momentarily spot the enemy, when the M1 (for example) are shooting at the... I guess the flash from gun, illuminates them enough. One might say ah... so it isn't that bad ? Well, it is.... as the spots keep dropping an re-appearing. If above a certain range, russian MBTs would want to shoot ATGMs at the hostiles, but for some reason they need 11-12 seconds to "AIM" their shot (why do they need 10+ seconds to aim an ATGM, I don't know). In that time, the spot will drop and the AI will forget where the enemy was, resetting the aiming process. Spot will re-appear as the M1s fire again and the russian MBTs will get stuck in a limbo, not being able to engage. They can return fire, when firing AP, as there is enough spot time to put a shot out, but they are always shooting second and at a slowed down rate of fire, as they reset every time. First two tracks show how a T-55 and T-72B3 (which is supposed to have one of the best thermals, from the variant of vehicles we have in DCS) have roughly the same spotting distance at night of around 1KM, meanwhile the M1A2s can see them from almost 4KMs.

Third tracks depicts the flash spotting.

T-72B3 AI Night Spotting Distance.trk.T-55 AI Night Spotting Distance.trkRussian AI Night.trk
 

Edited January 8, 2021 by Shadow KT

[Shadow KT]

Bump

[Fri13]

Hopefully new FLIR fix this, as they talked it not restircted to aircraft

[Shadow KT]

The FLIR, at least visually, is the same for all ground vehicles in DCS. All that has to be done are some numbers to be adjusted. Every other, NATO, modern MBT can spot at 4KM at night, yet the most modern Russian ones can't. T-72B3 has the highest gen thermal from all the models of MBTs we have in-game.

[Fri13]

4 minutes ago, Shadow KT said:

The FLIR, at least visually, is the same for all ground vehicles in DCS. All that has to be done are some numbers to be adjusted.

 

The FLIR should come more realistic in the near future. Far more blurrier images, meaning it will as well very effectively drop the ranges you can spot something or engage something as you can't make out what is the target. And when heat sources should start to get much smaller or even gone, the range and weather effects on it that your FLIR starts to be limited.

It is what should eventually happen in the ground units that your engagement ranges drop a lot.

[Shadow KT]

Cool, doesn't help the current situation. The problem is now, not in the future and can be adjusted. Also, seems like you haven't used much ground vehicle thermals. They are 10 times, if not more, shittier than TGP FLIR. It is super low resolution.

[Fri13]

7 hours ago, Shadow KT said:

Cool, doesn't help the current situation. The problem is now, not in the future and can be adjusted. Also, seems like you haven't used much ground vehicle thermals. They are 10 times, if not more, shittier than TGP FLIR. It is super low resolution.

 

The upcoming FLIR update is just around the corner.

 

And yes, I haven't used for a while because CA doesn't work in VR and I don't play DCS anymore without VR....Sure I could take it in mission editor for test run, but idea alone is.... As I am waiting the overhaul to come because the CA controls are so bad, and terrains are little unfit for proper tank warfare as there is not terrain shapes and covers etc.

 

 

[Northstar98]

On 1/14/2021 at 11:04 PM, Shadow KT said:

Cool, doesn't help the current situation. The problem is now, not in the future and can be adjusted. Also, seems like you haven't used much ground vehicle thermals. They are 10 times, if not more, shittier than TGP FLIR. It is super low resolution.

It also is super generic, though I can't imagine that changing until we get higher fidelity ground vehicles, or something more reminiscent of a tank simulator (which will probably mandate a large kick up the behind to get them fleshed out - I can go on for hours on end about ground vehicles).

On 1/15/2021 at 6:59 AM, Fri13 said:

The upcoming FLIR update is just around the corner.

I wish I shared your optimism.

We still haven't had many details about IR/FLIR modelling. They've said that they're redoing the maps and doing the 'underlying technology' but not many details apart from that (will barrels heat up as they fire for instance?).

Quote

As I am waiting the overhaul to come because the CA controls are so bad, and terrains are little unfit for proper tank warfare as there is not terrain shapes and covers etc.

True, direct CA would really mandate a dedicated map, that's maybe a lot smaller but more resolution. The Channel is getting there and is probably the best map so far. Don't forget that vehicle suspension modelling is non-existent, and if enabled (platform shake in settings) is completely backwards - simply randomly bobbing the hull around (regardless of what the tracks/wheels are actually doing), with magnitude being directly proportional to vehicle speed; and on older vehicles the animation isn't there either; with the hull moving independently to return rollers, idler wheels and drive sprockets; which looks very wrong. 

Edited January 12, 2022 by Northstar98
formatting, spelling

[Fri13]

42 minutes ago, Northstar98 said:

I wish I shared your optimism.

 

ED can anytime delay it to 2020, but hope is only thing that really helps that we see improvements now when everything is delayed to large updates that needs fixing in long delayed periods instead sooner smaller update.

 

We might not get all needed things at first, but maybe in 2-3 years then.

 

42 minutes ago, Northstar98 said:

We still haven't had many details about IR/FLIR modelling. They've said that they're redoing the maps and doing the 'underlying technology' but not many details apart from that (will barrels heat up as they fire for instance?).

 

There has been tidbits about those things. And if following argument "realistic FLIR" it as well means much lower resolution targeting systems if not though optic to eye.
 Like LITENING II to turn 512x512 px at both Wide and Narrow but digital zoom goes at 9x as small as 52x53 pixels. You barely see anything from 5-9 nmi range that is not a large burning building or other large area that you can try to interpret as "blob in blobs" becomes a thing.

 

On a MBT warfare it becomes even worse. Sure one can detect in optimal case something at 8 miles but that is around 6-7 nmi and not going to see for enemy vehicles really.

 

 

42 minutes ago, Northstar98 said:

True, direct CA would really mandate a dedicated map, that's maybe a lot smaller but more resolution. The Channel is getting there and is probably the best map so far. Don't forget that vehicle suspension modelling is non-existent,

 

I don't remember year, but maybe 2014-2016 Wags was searching a M1 Abrams personnel who had knowledge of the systems. They had a purpose to develop a M1A1 module.

He even had for some time (now gone AFAIK) in youtube a new vehicle model with animations and physics, for the suspension modeling, acceleration, braking, turning etc.

It was impressive back then considering that we had DCS 1.2.x back then. And it felt stupid to have such features when one square kilometer was one flat surface.

 

It is as well sad that has gone silent since then and features hidden etc. Like they really noticed that they must get far more vehicles to it and first the terrain and all other support it.

 

42 minutes ago, Northstar98 said:

and if enabled (platform shake in settings) is completely backwards - simply randomly bobbing the hull around (regardless of what the tracks/wheels are actually doing), with magnitude being directly proportional to vehicle speed;

 

Never kept that shaking active more than a minute for testing. And that was as 1.2 series.

It basically made all engagement happen on stationary situations and couldn't shoot from the move at all (normal battle combat speeds) or perform the combat actions as any movement was bad.

 

42 minutes ago, Northstar98 said:

and on older vehicles the animation isn't there either; with the hull moving independently to return rollers, idler wheels and drive sprockets; which looks very wrong. 

 

Why I am interested to see if the Battlefield Studio can come up with the contract to make three era battleground units for DCS. As I years back wished to see someone start offering ground units packs like current WW2 assets. But as like 70's era for UK, USA, USSR, DDR etc.and then own for 80's and so on.

 

I would be happy for such details as current ZSU-23-4 offers, even when not something like just released BTR-82 or T-72BM3 that are something great, but first start for the ground-Air purposes and then get more detailed as for ground combat only once the damage modeling and RTS elements come.

 

But to fix the current spotting ranges, it is possible just to change the current values on files. Like get that spotting ranges to same level, but nothing goes better until we actually get better AI for spotting like what Mi-24 WSO is getting. As that is as well required for ground.

 

One reason why I stopped long time ago the CA was AI capability spot you and engage you before anything was seen. And in 1.2.x it was even worse as AI saw through trees and buildings and even terrain, as long you entered their vision circle they reacted to you and waited that LOS was acquired from terrain and buildings.

Like sit in T-72 ready to launch a missile at coming helicopter and missile was already coming at you on the moment the popped first time behind the hill.

 

Fighting against AI was like going against superman with his vision.

[Shadow KT]

Can you stop making every bug report into a feature request ? The problem is out there and it can be fixed without new features. We can talk NEW, when it gets here. Now we need fixes for what is already in. 

Creating any realistic scenario/battle at night is almost not possible, when russian units are blind as bats, even when not supposed to and their counterparts just outspot them, for no reason.

I hope I don't read another "as intended".

[BIGNEWY]

Hi Shadow 

 

so partially related to this report I have already made 

 

[Shadow KT]

19 hours ago, BIGNEWY said:

Hi Shadow 

 

so partially related to this report I have already made 

 

Yes, talking about the same end effect. I just figured out what was causing it. The popping in an out are, when the enemy fires, the AI detects that for a moment. I guess the flash of light, from the gun firing. If the enemy does not fire, the russian vehicle will not spot it, until close to a kilometer.

Russian vehicles can spot at night, but at absurdly low distances, especially compared to their counterparts, with no objective reasoning.

[BIGNEWY]

I will mark this reported, and add the information to the report I already have in our system. 

 

thanks

[BIGNEWY]

I have spoken with the dev, this is correct as is, the night vison and optics are not as good as the USA counterparts, the RU units will loose sight at max range and reacquire when an engagement starts.

 

thanks 

[Shadow KT]

Sigh...
 

Edited January 22, 2021 by Shadow KT

[Shadow KT]

1)A T-90 AI can only spot a static M1 at a distance of 1.5KM, while an M1 can spot a stationary T-90 at around 4.3KM. This is a huge difference.

2)The T-90 AI can spot an M1 at the same range an M1 can spot a T-90, only if the M1 is on the move (even a very slow movement), or when it fires. This leads me to the conclusion that the AI can "see" a thermal spot in the distance, but cannot "identify" it as a target, unless it fires/moves.... Problem with this is that the AI has no memory. The moment the M1 stop moving/stops firing, it disappears for the AI. At high ranges where the russian tanks like to use ATGMs, it won't even return fire, because it takes 12 seconds for some reason to "aim" an ATGM shot. The M1 "disappears" at around 50% that time.

3)Most of the time, you can even see the target unaided, but the AI still won't see it. You can identify a target, while using CA and driving the tank, way beyond what the AI can see as well. 

I've attached a track showing the difference in distances.

 

AI Night Spotting.trk

 

Edited January 22, 2021 by Shadow KT

[Shadow KT]

Russians made tanks which shoot ATGMs at 4-6KMs, but they can't see past a km, a km and a half ? Doesn't sound right, does it ?. I get that Russian sights are worse, but they are not that worse. Usually you can spot further away than you can engage. One can argue that, it should be easier to spot at night, with thermals, because ambient is lower


 

Edited January 22, 2021 by Shadow KT

[Shadow KT]

Here is a B3 engaging at 1925 meters. Almost twice the range at which the AI in DCS will spot at night.

They might be worse than on an M1, but good enough to engage at a distance. I am sure there can be other parameters, which can be adjusted, so this can be factored in DCS, such as reaction time, or increased time to spot, but completely cutting their capability is not right.


 

 

Edited January 22, 2021 by Shadow KT