• Ultraviolet Photography
  •  

Time lapse of melting ice cube in SWIR

Infrared SWIR
21 replies to this topic

#1 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 24 February 2021 - 06:27

After a great deal of work, I have figured out how to access the video feed from my TriWave in MATLAB, and I've written a program to capture time lapses and process the resulting frames. Each frame of video is actually 120 frames from the camera (4 seconds' worth) averaged together, and this is repeated once a minute. This video took 90 minutes in real life, but's 13 seconds of video, or about 415 times faster than reality.

Lens: Wollensak Velostigmat 25mm/1.4
Filter: Thorlabs FB1450-12 (1450nm, 12nm FWHM)



YouTube's compression really degrades it. :wacko:
----

Update: I repeated the experiment at 1250nm. For some reason the Triwave is significantly less sensitive at 1250nm. (Or my lens has lower transmission that at 1450nm? Unlikely.) I don't THINK it's the filter, since the peak transmission is 50%-ish for the 1250nm but 30%-ish for the 1450nm, and the bandwidth is almost the same. ANYway, excuse my noise.
Filter: Thorlabs FB1250-10.


Edited by Andy Perrin, 25 February 2021 - 08:17.


#2 colinbm

    Member

  • Members+G
  • 2,405 posts
  • Location: Australia

Posted 24 February 2021 - 06:55

Funny, it looks like a block of ink, thanks for showing.

#3 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 24 February 2021 - 07:59

I’ve never seen a block of ink!

#4 UlfW

    Ulf W

  • Members+G
  • 1,607 posts
  • Location: Sweden, Malmö

Posted 24 February 2021 - 08:58

You could make an ice cube with black ink.

Blocks of ink are often used for asian calligraphy.
https://www.ebay.com...+block&_sacat=0
Ulf Wilhelmson
Curious and trying to see the invisible.

#5 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 24 February 2021 - 09:51

Really nice Andy.

#6 dabateman

    Da Bateman

  • Members+G
  • 2,767 posts
  • Location: Maryland

Posted 24 February 2021 - 11:54

Only took you ten days to post your black valentine's day heart. But you need to stop the video a couple seconds earlier. It just goes to mush at the end.

Looks like you can have some fun with this. What happens with a piece of paper, should see it darken up, but if you use color paper, I wonder if the dyes would interfere with the water dark signal.

Edited by dabateman, 24 February 2021 - 11:55.


#7 JMC

    Member

  • Members+G
  • 1,270 posts
  • Location: London, UK

Posted 24 February 2021 - 14:45

Wow, funky. It really does look dark at that wavelength.
Jonathan M. Crowther

http://jmcscientificconsulting.com

#8 Bernard Foot

    Bernard Foot

  • Members+G
  • 718 posts
  • Location: UK

Posted 24 February 2021 - 17:01

Excellent, Andy.
Bernard Foot

#9 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 25 February 2021 - 08:17

Bumping to mention that I just added a redo at 1250nm to the original post.

#10 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 25 February 2021 - 10:01

Very interesting, there is clearly a difference. The ice looks darker than the water at 1250 nm, I should check the curves but maybe it is just because it is thicker.

Found the graph: https://www.ultravio...attach_id=17341

Water peaks just after 1200 nm, ice peaks a bit further and absorbs a bit more than water at 1250 nm. I don't know if this little difference can be seen in the video, thickness probably plays a bigger role.

Also, why do you have more noise at 1250 nm than at 1450 nm? I guess you used a halogen bulb as a SWIR source. Also, the camera's sensitivity at both wavelengths is similar: https://www.ultravio...attach_id=15251

Maybe the 1250 nm filter is tighter?

Oh, nevermind, I just noticed you already wrote about this.

Edited by Stefano, 25 February 2021 - 10:16.


#11 dabateman

    Da Bateman

  • Members+G
  • 2,767 posts
  • Location: Maryland

Posted 25 February 2021 - 13:00

You may not have a cold plate, but if you did then you could really do some creative mixed icing and water layers looking at these wavelengths.
I know they aren't common, we had one in our lab and many people had never seen one before. But its fun to rapidly heat or cool something.

These are great. I wonder how dyes and salts or sugars would effect it.

You can have ice and sprinkle some salt on the top to get cold water.

#12 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 25 February 2021 - 14:02

Sugar is dark in SWIR, it would be cool to see transparent water and dark "sweet" water, but water is already black in SWIR.

#13 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 25 February 2021 - 15:49

I wish I had a way to measure the sensitivity of the TriWave at different wavelengths because I’m starting to think that curve shown in the link you put (Stefano) is not necessary accurate for my device. 1250nm definitely seems substantially less sensitive than 1450nm, and the filter transmission is the opposite pattern so it is not the filters. (I trust ThorLabs’ spectra.) The light should be stronger at 1250nm also, with halogen.

I also have a 1200nm filter and the signal is much stronger at 1200nm.

Edited by Andy Perrin, 25 February 2021 - 15:52.


#14 UlfW

    Ulf W

  • Members+G
  • 1,607 posts
  • Location: Sweden, Malmö

Posted 25 February 2021 - 15:50

View Postdabateman, on 25 February 2021 - 13:00, said:

You can have ice and sprinkle some salt on the top to get cold water.
It is amazing how low temperaturs you can reach with rather common cooling mixtures:
https://en.wikipedia...ki/Cooling_bath
Ulf Wilhelmson
Curious and trying to see the invisible.

#15 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 25 February 2021 - 15:58

For looking at sugar dissolving in water, I think my first choice would be crossed polarizers in visible light, actually. I may try that also...

#16 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 25 February 2021 - 17:53

View PostAndy Perrin, on 25 February 2021 - 15:49, said:

I wish I had a way to measure the sensitivity of the TriWave at different wavelengths because I’m starting to think that curve shown in the link you put (Stefano) is not necessary accurate for my device. 1250nm definitely seems substantially less sensitive than 1450nm, and the filter transmission is the opposite pattern so it is not the filters. (I trust ThorLabs’ spectra.) The light should be stronger at 1250nm also, with halogen.

I also have a 1200nm filter and the signal is much stronger at 1200nm.
If you have a diffraction grating with less than 1000 lines/mm you may try seeing a halogen bulb with it. It will not be accurate, but you should see something.

Since your camera was not a finished product, but something still under development, it is possible that the curve they provided was not accurate. If you can not trust it, and if you are a bit lucky, maybe your camera can go a bit beyond 1600 nm.

Something truly crazy to try would be to use upconversion phosphors on it to see MWIR. But I can hardly imagine a molecule capable of reacting to a ~0.5 eV photon. I am already impressed IR film is a thing.

#17 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 25 February 2021 - 20:19

Stefano, I don’t see the point of a diffraction grating without a scale to figure out wavelength. Also it’s a monochrome camera

#18 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 25 February 2021 - 21:58

You may see the "bands" since the sensitivity should be bumpy according to the (not so trustable) graph. About the wavelengths, you may use the Fraunhofer lines or a remote control as a rough reference (actually, the Fraunhofer lines are precise). But honestly, I don't know if it is a good idea.

#19 Stefano

    Member

  • Members(+)
  • 1,796 posts
  • Location: Italy

Posted 25 February 2021 - 22:13

I did image the Fraunhofer lines before, in UV, visible and IR. I even saw the calcium doublet (about 393.4 and 396.8 nm) with the naked eye (not really something to try, but I was curious). Once you figure out how "stretched" they are (they are not linear) you have a valid reference. But then your light source will not be a valid reference. Although halogen bulbs are blackbody radiators, their spectrum doesn't match exactly to that of a true blackbody. If you want something indicative only, you may try this.

The Sun must be narrowed down with a slit in order to achieve sufficient resolution to see the lines well.

#20 Andy Perrin

    Member

  • Members+G
  • 4,084 posts
  • Location: United States

Posted 25 February 2021 - 22:19

View PostStefano, on 25 February 2021 - 21:58, said:

You may see the "bands" since the sensitivity should be bumpy according to the (not so trustable) graph. About the wavelengths, you may use the Fraunhofer lines or a remote control as a rough reference (actually, the Fraunhofer lines are precise). But honestly, I don't know if it is a good idea.
The Fraunhofer lines are blotted out at the points 1450nm and 1250nm. Lots of water vapor in the atmosphere.