Jump to content
UltravioletPhotography

First steps in MWIR!


Recommended Posts

Andy Perrin

A week ago, Stefano PMed me, with the title, "MWIR camera at 'affordable' price?" I admit, I was skeptical, but the skepticism turned to amazement as I read the eBay listing. For sale was an Agema 470 Pro, at "Buy It Now" of $650, or best offer. The camera was of the HgCdTe (or MCT) type, which means it has a single pixel and a high speed rotating mirror that directs light onto the sensor, which is cooled via Peltier effect to -80C or so. Effective resolution was 100x140 pixels. The sensitivity is 2-5 microns, going from the long end of SWIR into the mid-MWIR. From 5-8 microns, air is absorbing, so no cameras are available in that range currently (nor likely ever to be). Beyond 8 microns is the usual LWIR window where my other thermal cameras work.

 

The seller had posted pictures of the camera operating, and a power supply was easily available, so I thought: why not? So I got the camera.

 

The camera is extremely large and heavy. It is about 50cm (22 inches) long, and it weights 7kg (15.4lb). It has a monochrome viewfinder and a floppy drive (3.5") for storage. I do not have any floppy disks alas.

 

post-94-0-62657500-1617081528.jpg

 

It took a few days to acquire a power supply and a light source. I bought a "Deep Heat Projector" from Arcadia Reptile. Arcadia had this to say when I asked about the spectrum last year in reference to my TriWave:

Hi Andy,

I haven’t got an image that I am able to share as yet, but I can tell you how the energy flows.

 

Energy starts to increase from very little at all at 1000nm. Between 1000 and 2000nm The energy gradually increases to approx. 50% relative energy. It then reaches a plateau through to around 2700nm from which is slowly starts to decrease again.

 

So, yes, energy is increasing between 1100-1600nm but flats off around 2000nm. It the starts to drop back down again from 2700nm approx..

 

Hope this helps.

 

John Courteney-Smith MRSB

Arcadia Reptile; Head of Science and Innovation.

I didn't buy it last year (didn't get around to it) but with the MWIR camera it was too handy to resist: better SWIR light and one that worked for the short end of the MWIR. So I got that along with a socket for it.

 

Today all the stuff arrived and I put it together. The camera makes a revving up noise like a jet engine: a slowly building whirrrrrrRRRRRR!R!R!R!R!!!!!! as that mirror goes faster and faster. The electronics turns on and a boot-up screen appears, showing the software dates to May 1, 1989.

 

 

 

The camera originally came with a variety of lenses, so the lens on the front is detachable. The one it came with was a 20 deg FOV lens — in IR, camera lenses are described by field of view (FOV) rather than focal length. With the lens removed, there is another (concave) lens behind it, and according to the ancient manual, which is available still from FLIR's website since FLIR bought Agema eons ago, you can use it in macro mode if you leave the outer lens off.

 

 

Showing how highlights on my hand vanish when the reptile light is removed from my hand:

https://youtube.com/Kz6nH_u0Teo

 

Showing teeth changing temperature as I breathe:

https://youtube.com/0exOl11PJSk

Link to comment

I'm so happy I helped you start exploring MWIR. This is just the beginning. Now you still have to fine tune things and learn how the camera behaves, but after that you will be set. There's so much to see there!

 

Colin, I think not much since skin should behave like in SWIR, water is dark and the skin contains water (at least, this should be the main reason for dark skin in SWIR), but I can be wrong.

Link to comment
Andy Perrin
Colin, Stefano is probably right, at least near the 2 micron (2000nm) end of the range. The water absorption does stay high in LWIR at 8000nm, so unless there's a reason for it to be transparent in between it probably isn't? I think skin is most transparent in the near infrared (where I took photos before), and also out in X-rays.
Link to comment
Andy Perrin

I'm so happy I helped you start exploring MWIR. This is just the beginning. Now you still have to fine tune things and learn how the camera behaves, but after that you will be set. There's so much to see there!

 

Colin, I think not much since skin should behave like in SWIR, water is dark and the skin contains water (at least, this should be the main reason for dark skin in SWIR), but I can be wrong.

Thank you again!

 

My first impressions are that taking reflected-light-only photos may be a bit challenging because the emitted thermal radiation is much stronger than I realized from room temperature objects. Essentially EVERYTHING wants to be glowing all the time. In principle, I can find a cool surface, put my objects to be imaged on it, and shine very bright MWIR on it while I shoot photos, drowning out the emitted light. In practice, you can see above with my hand that although the reflected light is visible, it competes head-to head with the emitted light from my hand.

Link to comment

Wow that is so cool.

I remember Sony digital cameras that took 3.5" floppy disks and remember how crazy that was. You should be able to get a box of 3M ones from somewhere.

I think I have a box or two and an old Toshiba usb to 3.5 floppy drive somewhere. That came with my first laptop.

 

So now we can see IR, SWIR, MWIR, LWIR from you around town. Next will be microwaves.

 

 

 

Link to comment
Andy Perrin

So now we can see IR, SWIR, MWIR, LWIR from you around town. Next will be microwaves.

Hah, not for awhile. Current projects are:

1) Laser-induced Fluorescence project

2) Planning to try Schlieren imaging

3) Still want a camera that works between 1600-2000nm ("SWIR-B" is my made-up name).

Link to comment

Thank you again!

 

My first impressions are that taking reflected-light-only photos may be a bit challenging because the emitted thermal radiation is much stronger than I realized from room temperature objects. Essentially EVERYTHING wants to be glowing all the time. In principle, I can find a cool surface, put my objects to be imaged on it, and shine very bright MWIR on it while I shoot photos, drowning out the emitted light. In practice, you can see above with my hand that although the reflected light is visible, it competes head-to head with the emitted light from my hand.

You will surely need a good set of filters. MWIR is very interesting also because it's the land where reflected light images and thermal emitted light images merge, so you get something from both worlds. Objects glow, but also reflect (you can see reflections in LWIR too as you told me, but LWIR must be more "thermal" and less "reflected" than MWIR).

 

With your TriWave and a 1500 nm longpass filter you didn't see any emission at room temperatures, a soldering iron is a lightbuob in SWIR but people are black.

 

If you put a 3 μm or a 2.5 μm shortpass filter on the lens you should get rid of most thermal emission. Above about 3 μm objects should start to glow (but the cut-on is anything but sharp and very temperature-dependent). It will be challenging to produce reflected images in the higher 4-5 μm band. Either you subtract the thermal emission in post or you have to cool down the subject in some way. It's like reflected LWIR images, but a bit less challenging.

 

Next could be THz waves. My design with a thin helical antenna and Schottky diodes may not work at all, I will try to think about it more seriously in the next years. Microbolometers seem to be very promising to me since they can virtually sense anything that heats them up. You may be able to modify a LWIR camera to see UV, changing the lens and putting a UV-pass filter on it.

 

With the addition of MWIR, the IR spectrum is almost completely covered. There's a small hole between 1600 and 2000 nm (TriWave/Agema 470 Pro), and 5-8 μm. The "SWIR-B" band should be interesting as thermal emission shouldn't occur (I think, but one really never knows until trying).

 

If you will find a way to use your UVC imager, you will be able to see from 254 nm (or even 185 nm) to LWIR, passing through UVB, UVA, VIS, NIR, SWIR and MWIR. Not bad.

Link to comment
Andy Perrin

Yeah, but no need to cool things that much. The research literature I found said they could just shine a strong MWIR light source on a painting and the reflected light would be so much stronger than the emitted light that you would get a mostly-reflected photo. The majority of the emitted light is LWIR, not MWIR.

 

In fact people use MWIR now mostly for high temperature thermography (looking at furnaces and stuff). LWIR cameras are much better near room temperature.

Link to comment
I know, but as far as I know most things appear similarly bright (if you don't consider absolute brightness. MWIR and LWIR images are often similar). People glow, the sky is dark and so on. But some things are clearly different.
Link to comment
Andy Perrin
Sure, but why are you making this comment here when the discussion was about reflectography? The point is that if I keep the surfaces cool then I can do reflectography but they don’t have to be at dry ice temperatures. I just need more MWIR light from reflection than emission. I mean, I’m not trying to take a picture of a flower at 500C or something.
Link to comment

Yes, your point is valid. You want reflected MWIR images now. About this, one idea I had to "remove" emitted light is to try using normal glass lenses, if they work at all. Glass (depends on the type) should still be partially transparent in the lower end of the MWIR band. Above 3 μm they should be useless, but below they may still work. And also they will act as a shortpass filter. So if you are aiming at the 2-3 μm band, that can help you.

 

Or, subtracting the emitted MWIR image from the emitted + reflected MWIR image should give you a purely reflected MWIR image. Kind of like this: https://www.ultravioletphotography.com/content/index.php/topic/4203-tree-stump-uvivf-with-contaminating-light-removal-by-image-subtraction

Link to comment
Andy Perrin

I would think glass filters would be better because the lenses not only don’t mount to the camera easily (It uses custom mount tech) they would also have massive focal shift out in MWIR. I can put a filter on the front much more easily.

 

Image subtraction is unfortunately not a great option here because the analog signal has only 128 gray levels. You get poor subtraction results when the bit depth is low.

Link to comment

Andy flowers at 500 might be hard but entertaining.

Just stick some stuff in your oven, cook it up and look for the fun.

Visible its on fire, but just look at that MWIR image!

Link to comment
Andy Perrin

Honestly, I am thinking that it might be better to stick things I want to image in the freezer first, though. Look how much the Planck spectrum changes between 70F (21.1C) and 20F (-6.66C):post-94-0-51234500-1617497040.png

 

If we regard the thermal radiation as the "noise" then that's like doubling the signal-to-noise ratio if the incoming MWIR is kept constant.

Link to comment

Honestly, I am thinking that it might be better to stick things I want to image in the freezer first, though. Look how much the Planck spectrum changes between 70F (21.1C) and 20F (-6.66C):post-94-0-51234500-1617497040.png

 

If we regard the thermal radiation as the "noise" then that's like doubling the signal-to-noise ratio if the incoming MWIR is kept constant.

Also look at the huge difference between 3 μm and 4/5 μm (in both lines). I bet if you put a 3 μm shortpass filter as I thought you will cut down like 99% of the blackbody radiation.

 

This also means that getting a purely reflected image becomes much more difficult very quickly as you increase the wavelength.

Link to comment

Some Schott glass filters you may be interested into (go here to see the full spectrums yourself at various thicknesses, I will report the approximate eyeballed values at a thickness of 2 mm):

 

Cut-off at ~2900 nm:

S8612, BG39;

 

Cut-off at ~2850 nm:

BG40, BG42, BG55, UG5;

 

Cut-off at ~2800 nm:

BG18 (1400-2800 nm), BG38, BG50;

 

Others:

 

BG36 (peaky spectrum, cuts below ~3450 nm);

 

BG60 (up to ~4100 nm, attenuates above ~2800 nm);

 

BG61 (same as BG60, but with stronger attenuation in the longer wavelengths);

 

BG62 (similar to BG60);

 

BG63 (similar to BG60, up to ~4050 nm);

 

BG64 (similar to BG63);

 

BG67 (up to ~4100 nm, but with less attenuation);

 

WG, GG, OG and RG glass (Schott longpass filters) cut in the high 4000s, at ~4500 nm or a bit above;

 

KG1 and KG2 cut at ~2800 and ~2900 nm respectively, but with low transmissions, KG3 and KG5 have little bumps in the 2000s;

 

NG filters have a peak at ~2700 and ~3300 nm;

 

UG1 transmits between ~2150 and ~4650 nm, but the transmission is low; UG11 is practically useless;

 

VG9 cuts at ~4600 nm;

 

VG20 cuts at ~4150 nm.

Link to comment

Andy flowers at 500 might be hard but entertaining.

I woke up today feeling a bit mean and could not resist.

Apologies in advance.

 

They say that water is a key factor for life and our bodies contain a lot of it. That is logical.

Water changes from solid to liquid and to gas (at a rather normal pressure) at two distinct temperatures.

You can use those temperatures as references and combine them with two decades of our decimal system.

 

Then you get a logical temperature scale:

Freezing : 0 (°C) and boiling 100 (°C)

Compare with

Freezing : 32 (°F) and boiling 212 (°F)

 

I guess that the 500 above is temperature in °F. => 260°C

 

Can someone point out the logics behind the °F :smile: :wink:

 

No offence, just teasing.

 

If looking for more strange American units take a look at

AWG where AWG-number goes higher when diameters decrease

https://en.wikipedia...ican_wire_gauge

and

NPT where the NTP-values are much smaller that the actual diameters of the threads

https://en.wikipedia...nal_pipe_thread

 

Strange and confusing units can sometimes be expensive.

Remember this:

http://edition.cnn.com/TECH/space/9909/30/mars.metric/#:~:text=(CNN)%20%2D%2D%20NASA%20lost%20a,a%20review%20finding%20released%20Thursday.

Link to comment

And what about the psi? I just can't use it. I am ok with "atmospheres", bars and so on. But "pound-force per square inch" is just no.

 

As of temperature, the SI unit is Kelvin. It is "perfect" as 0 K is 0. It doesn't go negative.

 

And also, you have to put a space between the number and the unit. 365 nm must be written with the space, not "365nm". I am saying this as a lot of people make that mistake here. I too make mistakes of course, but this is a rule I know and always follow.

 

See here: https://en.wikipedia...s#General_rules

Link to comment

SI rules, but temperature in Kelvin is not that practical for daily usage use.

The °C scale is more practical for that with its strong relation to the common behaviour of Water.

I have negative numbers in my rational number range. They are quite useful. I cannot imagine being without them or the number 0. :wink:

 

Thanks for the notice about the space by nm. I am sure I made that mistake many times, even if it is rather harmless. :smile:

I cannot find any bad consequence caused by the error except that it is not the correct way to write.

Can you?

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...