• Ultraviolet Photography
  •  

SWIR camera setup and some pics

SWIR
34 replies to this topic

#21 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 18 February 2017 - 07:22

As observed in the last pic, many plastics seem to go dark in SWIR. Look what happens to the plastic tensile test specimen (I think it was HDPE?) in the mug here:

Attached Image: IMG_1407.JPG

Edited by Andy Perrin, 18 February 2017 - 07:58.


#22 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 22 February 2017 - 17:12

Heh, yes, if I had a piece of silicon!

What about sand? Or a chunk of quartz SiO2? Gets you a bit closer to silicon.
Andrea G. Blum
Often found hanging out with flowers & bees.

#23 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 22 February 2017 - 17:17

Unfortunately, chemical compounds are nothing like their constituent elements. Here is what quartz looks like versus silicon (images from wiki):
Attached Image: 240px-Quartz,_Tibet.jpgAttached Image: 220px-SiliconCroda.jpg

Silicon isn't that expensive, though, so I might just buy a chunk.

Edited by Andy Perrin, 22 February 2017 - 17:18.


#24 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 22 February 2017 - 17:33

http://www.ebay.it/i...wQAAOSwzaJYBpbM

There you go! On Ebay for only US$8.80 + shipping US$5.15 = US$13.95.
That's an Italian company. There is probably a US company somewhere who also sells samples of elements.
Andrea G. Blum
Often found hanging out with flowers & bees.

#25 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 22 February 2017 - 17:50

Yep, I'll probably get around to it!

#26 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 22 February 2017 - 18:11

I have a question....

If 900-1700 nm is referred to as Shortwave Infrared (SWIR), then what do they/you/we/whoever call the part from 700-900 nm?? Extra-short IR? Short-short IR? :lol:
Andrea G. Blum
Often found hanging out with flowers & bees.

#27 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 22 February 2017 - 18:54

Short answer: For me, NIR is between visible and SWIR. But some people call 700 (or 750...) to 3um all NIR. See below.

Longer version:
So the whole IR nomenclature issue is a mess! Depending on what you're reading, you can find many ways of dicing up the IR.

Edmund Optics: NIR=700 to 1100nm (in one part of their glossary) and "IR radiation is often divided into three smaller regions: 0.750 - 3μm, 3 - 30μm, and 30 - 1000μm – defined as near-infrared (NIR), mid-wave infrared (MWIR), and far-infrared (FIR), respectively," in ANOTHER part of the SAME Glossary!

Wikipedia's rundown (I use the Wiki cutoff table labeled "Commonly used sub-division scheme")
https://en.wikipedia...in_the_infrared

Edited by Andy Perrin, 22 February 2017 - 18:57.


#28 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 27 February 2017 - 17:35

So the whole IR nomenclature issue is a mess!

Yep!!
Andrea G. Blum
Often found hanging out with flowers & bees.

#29 Hornblende

    Member

  • Members
  • 209 posts
  • Location: Canada

Posted 15 March 2017 - 20:18

Andy, I was thinking about how to improve your SWIR pictures: you could use the well known stacking method used is astrophotography!

You have to:

- Disable every noise reduction settings on your camera
- Take at least 10 exposures of you subject (I think 30 should be far enough). These are the "light" frames.
- Take the same amount of exposures with same settings but without any light reaching the sensor (lens cap on, no light source). These are the "dark" frames.
- Use the freeware Deepskystacker to stack the images
- Get noise free image out of it.
- Enjoy :)

Edited by Hornblende, 15 March 2017 - 20:19.


#30 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 15 March 2017 - 20:21

I didn't realize until just now that it was necessary to take the same amount of dark exposures too.
Andrea G. Blum
Often found hanging out with flowers & bees.

#31 Hornblende

    Member

  • Members
  • 209 posts
  • Location: Canada

Posted 15 March 2017 - 20:52

In this particular case, the amount of light frame is low so it is better to take the same amount of dark frames.
If you take a lot of light frames (more than 100 for example) you don't necesseraly have to take the same amount of dark frames, it would take forever (some people do). I usualy take 50 dark frames max when I am shooting the night sky.

#32 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 15 March 2017 - 22:48

Andrea and Hornblende: unfortunately you are mistaking the source of the noise! This is not ISO noise for the most part. It doesn't change in time. It is the pattern of phosphor grains on the screen, which I am photographing from the rear. I could stack an infinite number of frames and it would make no difference because (from the camera's point of view) the screen is the SUBJECT of the photo. What I have thought of doing is trying to SUBTRACT the fixed pattern of grains from the image. If you want an analogy, this is more like the line noise that you see from the sensor.

Edit: here is a diagram of the setup.
Attached Image: SWIR setup.jpg

Edited by Andy Perrin, 15 March 2017 - 23:06.


#33 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 7,403 posts
  • Location: USA

Posted 17 March 2017 - 16:18

Makes sense! Thanks, Andy.
Andrea G. Blum
Often found hanging out with flowers & bees.

#34 Hornblende

    Member

  • Members
  • 209 posts
  • Location: Canada

Posted 21 December 2017 - 14:39

Andy, you could substract the pattern of the screen in photoshop. Just take multiple pictures in SWIR of a blanc screen, merge them in photoshop (that way you reduce the noise a little bit) and then you can substract it from other SWIR pictures. It should give you a much more "polished" image.

#35 Andy Perrin

    Member

  • Members(+)
  • 2,459 posts
  • Location: United States

Posted 21 December 2017 - 16:32

Hornblende, sadly that sounds like it should work but I have already tried it many times and failed. The issue is that the image does not come out of the camera looking like the above; instead the contrast is nearly zero and I am digitally stretching it quite a bit. Also the screen itself seems to be quite sensitive to the angle the light hits it at, and it doesn't fluoresce in quite the same pattern each time (although it is similar). Think of a rough surface that is illuminated at different angles -- different surfaces catch the light each time.

In fact, I just noticed that in post 32 above, I mentioned trying this.

Edited by Andy Perrin, 21 December 2017 - 16:35.