MWIR startup

[lukaszgryglicki]

Hi - very simple question:

- What is the best option to start MWIR right now?

- MWIR is considered to be between "emission" and "reflection" - how it really works? Simple question: having a MWIR device in a normal sunny day - I'm making a photo - what will I get?

- If I point a MWIR camera on sun - will it emit enough "light" (IR radiation) to make a photo?

 

And then:

- What device can I buy + what lens?

 

[Stefano]

I think only Andy tried MWIR here, you may have read his topic: 

As for the reflection and emission thing, you get both. Here's a nice example of an optical+thermal shadow I already posted elsewhere:

 

I don't know if the video was taken in MWIR, but it could be.

[Andy Perrin]

I think the HgCdTe cameras are the most affordable option for MWIR. I have to say, one reason you haven't seen much followup on this topic from me is that MWIR filters are costly and the camera itself is something like 15kg and LOUD, in fact it sounds like an airplane taking off when you turn it on. Both of those make it a little uncomfortable to experiment with, although I do intend to get around to it one of these days! 

 

Lukas, you are much richer than I am, you may be able to afford some of the other MWIR options. FLIR makes some, but they are in the range of the UV Nikkor and more for price...

[Andy Perrin]

1 hour ago, lukaszgryglicki said:

- MWIR is considered to be between "emission" and "reflection" - how it really works? Simple question: having a MWIR device in a normal sunny day - I'm making a photo - what will I get?

Just like LWIR, you get get both emission and reflection at the same time. All optics behaves the same way, regardless of wavelength. If you have a glowing object in room with not much other light, you see primarily the self-glow. If you have a glowing object in a room with a very bright external source like sunlight, the glow might not be visible even, and you see mainly the reflected light and shadows cast by the external source. We all have everyday experience with this, and so you should be able to extend that experience in your imagination to the MWIR case. The sun is definitely a good MWIR source.

 

In the "frozen shadow" case, you can see that the object glow and the sunlight are roughly equal to each other, or else the frozen shadow and regular shadow would not be visible at once.

[Stefano]

Following this discussion, could he simply put a MWIR shortpass filter on a typical microbolometer LWIR camera? Would that even work?

[Andy Perrin]

Not with any certainty - as I said, I'm not sure what additional filtration may be in the FLIR camera, but in addition your speculation in that thread was flawed. You cannot simply replace the lens because the way the camera works is that it has to be electronically calibrated to adjust for the light generated by the lens itself. FLIR cameras are really miracles of engineering, because they have to compensate for SOO many things to get an image out. And also there is all the usual issues with lenses focussing at different locations for different materials (refractive index is BIG in LWIR/MWIR, it is not close to 1 like glass).

 

In other words, it is possible in principle to make a microbolometer based camera in other wavelengths, but it really needs to be designed that way from the beginning. There's no way you are gonna adapt one. It's a harder problem than lens design, and nobody here has even done that much, despite much talk and playing with software.

[Andy Perrin]

oops, wrong place.

[Stefano]

Thanks. I didn't know it was that complicated. If the calibration is about temperature accuracy or lens vignetting it shouldn't be a problem, if it's deeper instead (as I understand it is) it probably makes it impossibile to convert a camera.

[Andy Perrin]

It's about just getting an image at all, Stefano. The most basic requirement to build a camera is to have a box that's dark inside.  Imagine if you made a camera out of glowing material. You have to come up with a complicated scheme to subtract off the ambient glow of everything inside the box, and on top of that, the glow isn't static, it's actually changing with temperature. This is why FLIR cameras make that clicking-sound every few seconds. It is called the NUC -- the "non-uniformity correction" -- and it involves closing a shutter and taking a picture to subtract off the uneven glow from the camera walls and lens, as well as sensitivity differences between pixels. 

[Stefano]

Thanks, it makes sense.

[lukaszgryglicki]

On 2/25/2024 at 1:52 AM, Andy Perrin said:

Just like LWIR, you get get both emission and reflection at the same time. All optics behaves the same way, regardless of wavelength. If you have a glowing object in room with not much other light, you see primarily the self-glow. If you have a glowing object in a room with a very bright external source like sunlight, the glow might not be visible even, and you see mainly the reflected light and shadows cast by the external source. We all have everyday experience with this, and so you should be able to extend that experience in your imagination to the MWIR case. The sun is definitely a good MWIR source.

 

In the "frozen shadow" case, you can see that the object glow and the sunlight are roughly equal to each other, or else the frozen shadow and regular shadow would not be visible at once.

 

Hi - I understand it all, but my question is (I'll try to explain).

If I just go outside on a normal sunny/half-overcast day, temperature around +15C, and shoot an average city scene - what would be registered? Will sunlight be enough to register data, will resulting image be mostly reflected MWIR or emission MWIR for hotter objects? At which temperature emission will be about the same as reflected light?

 

I'm interested in proportions in an average scene, I guess MWIR is the place where they will "meet", while LWIR is mostly emission (my FLIR 8-14 um camera) and SWIR/NIR are mostly reflection (I don't have any SWIR device, but I tried to make 1um+ on my silicon based fs cameras and they were juest fully reflected light images with slightly darker water.

 

[Andy Perrin]

Yeah, that’s what I’m telling you — the 2-5 micron MWIR range, taken as a whole (no filtration) would give you something like a 50-50% mix of reflected and emitted light on a sunny day at “normal” earthly temperatures. That is a massive approximation, however, because the details depend on so many factors. 
 

If you wanted to analyze it, assume you have an opaque surface insulated on the bottom with emissivity eps and unknown temperature Tsurf. Assume you have full sunlight perpendicular to the surface with irradiance G and absorption coefficient alpha. Assume the surface faces up to the sky, which is at temperature Tsky. Let h be the heat transfer coefficient due to air cooling, and Tair be the air temperature. Sigma is the Stefan-Boltzmann constant (5.67e-8 W/m^2K^4), which is a constant of nature and never changes. All temperatures must be in Kelvin. The eps and alpha are both functions of wavelength. You should pick alpha based on the sun’s dominant wavelength (0.5 micron) and epsilon at around 9 micron when you solve for the surface temperature. 
 

The energy balance would be:

alpha*G = eps*sigma*(Tsurf^4-Tsky^4) + h*(Tsurf-Tair), 

which in English says,

”absorbed radiation from the sun in visible light = emitted radiation from the surface in LWIR + heat lost to the air.”

 

You would solve that equation for Tsurf, then the reflected solar MWIR would be

   q_reflected_sun = (1-alpha_in_MWIR)*G_in_MWIR +  (1-eps_in_MWIR)*sigma*(T_surf^4-T_sky^4)*(fraction of sunlight in MWIR)

 

and the emitted light is

  q_emitted = eps_in_MWIR*sigma*(T_surf^4 - T_sky^4)*(fraction of blackbody light in MWIR)

 

In English those would say,

“MWIR light reflected from the surface =  reflected MWIR light from the sun + reflected MWIR light from the sky”

 

and

 

”MWIR emitted from the surface = emissivity * fraction of the blackbody radiation in MWIR”.

 

The final answer would be the ratio of q_reflected_sun to the total light. 

 

It is VERY complicated and that is the real point here, as well as hopefully aiding your understanding of how the problem would be solved. 

[lukaszgryglicki]

Thanks, I guess I would need to see some images from very different every-day scenes.