• Ultraviolet Photography
  •  

And yet, another rainbow

Multispectral
34 replies to this topic

#1 Stefano

    Member

  • Members(+)
  • 2,080 posts
  • Location: Italy

Posted 05 June 2021 - 17:30

We had a rainbow today, and I had to try some invisible light photography on it. My camera caught some drops of water, but nothing to worry about.

Cameras:
Samsung Galaxy A40, front camera (visible);
Full-spectrum Canon EOS M (UV, IR, full-spectrum);
Lens: Soligor 35 mm f/3.5.

Filters:
UV: ZWB2 (2 mm) + Chinese BG39 (2 mm);
IR: Hoya R72;
No filters for the full-spectrum image.

Visible light (f/1.7, ISO 40, 1/2370 s exposure). Image cropped to approximately match the others:
Attached Image: 20210605_190115.jpg

Infrared (f/8, ISO 100, 1/250 s exposure):
Attached Image: IMG_2830.JPG

Ultraviolet (f/8, ISO 100, 1 s exposure):
Attached Image: IMG_2824.JPG
I have an image with more exposure, but the rainbow is less visible.

Full-spectrum (f/8, ISO 100, 1/1000 s exposure):
Attached Image: IMG_2839.JPG
The camera is very sensitive in full-spectrum. If I had the lens wide open, I would have overexposed at ISO 100. Also, the lens has quite a bit of chromatic aberration with a band this wide.

#2 colinbm

    Member

  • Members+G
  • 2,625 posts
  • Location: Australia

Posted 06 June 2021 - 00:05

Fantastic Stefano, that is a great capture & the wavelength spread is clear to me, in both rainbows.

#3 dabateman

    Da Bateman

  • Members+G
  • 2,933 posts
  • Location: Maryland

Posted 06 June 2021 - 00:15

I love multi spectral rainbows. You have visible leak in your R72 filter.

#4 Stefano

    Member

  • Members(+)
  • 2,080 posts
  • Location: Italy

Posted 06 June 2021 - 00:22

Thanks both. Yes, I noticed too that the red bow in the visible image overlaps the infrared rainbow. I don't know if my phone camera is seeing some far red or the R72 filter leaks below 700 nm (it should cut there). I should investigate this further.

#5 Andy Perrin

    Member

  • Members+G
  • 4,313 posts
  • Location: United States

Posted 06 June 2021 - 02:25

I would suspect the phone cam of leaking unless you are using an no-name branded 720nm filter. We know the (real) Hoyas are okay! :grin:

Edited by Andy Perrin, 06 June 2021 - 02:26.


#6 colinbm

    Member

  • Members+G
  • 2,625 posts
  • Location: Australia

Posted 06 June 2021 - 02:32

He says he only used the phone for Visible light.
" Full-spectrum Canon EOS M (UV, IR, full-spectrum); "

#7 Andy Perrin

    Member

  • Members+G
  • 4,313 posts
  • Location: United States

Posted 06 June 2021 - 02:35

Right, the phone cam is leaking INFRARED (or very far red).

#8 Cadmium

    Member

  • Members+G
  • 4,202 posts

Posted 06 June 2021 - 03:44

Phone cam, the first photo? Why do you say it is leaking IR?

#9 Andy Perrin

    Member

  • Members+G
  • 4,313 posts
  • Location: United States

Posted 06 June 2021 - 03:47

Cadmium, because the red part of the rainbow overlaps in the visible and IR photos. I tried aligning the two images and confirmed it. So probably the phone is going a bit into the 700nm's. It seems unlikely the Hoya is transmitting below 700nm.

Edited by Andy Perrin, 06 June 2021 - 03:47.


#10 Cadmium

    Member

  • Members+G
  • 4,202 posts

Posted 06 June 2021 - 04:15

Perhaps not, but even 720nm produced a good amount of false color, just think of all the 720nm images, swapped with blue skies, etc..

#11 Andy Perrin

    Member

  • Members+G
  • 4,313 posts
  • Location: United States

Posted 06 June 2021 - 04:32

The issue is that the rainbow is splitting up the colors like a grating would, so IR and red shouldn’t be in the same physical location. This has nothing to do with false color, the same issue would be there in a monochrome sensor. I don’t think the problem is the Hoya, I think that phone cam is not cutting off. Easy test would be to shine an 700nm LED at some PTFE and see if the phone can see it.

Edited by Andy Perrin, 06 June 2021 - 04:36.


#12 colinbm

    Member

  • Members+G
  • 2,625 posts
  • Location: Australia

Posted 06 June 2021 - 04:55

If the phone has Chinese QB39 it will leak IR

#13 Cadmium

    Member

  • Members+G
  • 4,202 posts

Posted 06 June 2021 - 05:21

...or just compare a shot with S8612 2mm in front of it to a shot with nothing in front of it (Schott BG39 2mm should work the same as Schott S8612 2mm, it cuts off at the same place as S8612 2mm, no leak).

#14 Andy Perrin

    Member

  • Members+G
  • 4,313 posts
  • Location: United States

Posted 06 June 2021 - 05:30

At first I thought that made sense but now I don’t know.

How would we tell a small IR leak in a vis photo without IR of known wavelength to verify with? We can’t say the phone has any particular filter glass.

Edited by Andy Perrin, 06 June 2021 - 05:44.


#15 Stefano

    Member

  • Members(+)
  • 2,080 posts
  • Location: Italy

Posted 06 June 2021 - 08:01

So, a few points:
- I bought the Hoya filter on Amazon from Hoya, I really think it's genuine (and it was also quite expensive);
- I can try to stack my (Chinese, not Schott) BG39 on the phone and see what happens (but I no longer have a rainbow);
- I can use a diffraction grating to see how deep the phone sees with and without the R72 filter. I could even try to see the Fraunhofer lines to have a wavelength reference, but I don't know if I can do that.

So I will try those experiments, and post here.


#16 Fandyus

    František G.

  • Members
  • 122 posts
  • Location: Czech republic, Northern Bohemia

Posted 06 June 2021 - 09:52

I wanted to do this myself as an experiment, but you did it earlier. Thanks! :)

#17 Stefano

    Member

  • Members(+)
  • 2,080 posts
  • Location: Italy

Posted 06 June 2021 - 09:57

Thanks Fandyus. Actually I am not the first one to do that, there are many other examples here on UVP.

#18 dabateman

    Da Bateman

  • Members+G
  • 2,933 posts
  • Location: Maryland

Posted 06 June 2021 - 11:19

If you have a garden hose you can make a rainbow in your backyard. They just tend to be full circles and much smaller.

#19 Stefano

    Member

  • Members(+)
  • 2,080 posts
  • Location: Italy

Posted 06 June 2021 - 11:51

View Postdabateman, on 06 June 2021 - 11:19, said:

If you have a garden hose you can make a rainbow in your backyard. They just tend to be full circles and much smaller.
Yes, I have it, I will give it a try if I can.

Initial experiments:

Solar spectrum with the Canon EOS M + Soligor 35 mm f/3.5 + Hoya R72 + 1000 lines/mm diffraction grating. I can't identify the other Fraunhofer lines, I would really appreciate if someone could do that:

f/8, ISO 100, 1/4000 s exposure:
Attached Image: IMG_2885 - Copia.JPG

I tried it with my phone, but I could barely see a couple of lines, not very helpful.

My phone + R72 can see, so it does see a bit past 700 nm:

f/1.7, ISO 100, 1/100 s exposure (both images):
Attached Image: 20210606_104623.jpg

Attached Image: 20210606_104631.jpg

I then took two images with my phone, 1000 lines/mm diffraction grating, 12 V halogen lamp ran at 20 V, no filters and with the R72:

No filters (f/1.7, ISO 100, 1/8850 s exposure):
Attached Image: 20210606_110758.jpg

R72 (f/1.7, ISO 100, 1/100 s exposure):
Attached Image: 20210606_110808.jpg

BiColour stack (cyan: no filters, red: R72):
Attached Image: somma 319.jpg

Between the images there is a difference of 88.5 times (6.46 stops) of exposure, consider that.

A nice site to help find Fraunhofer lines: http://bass2000.obsp...solar_spect.php

#20 colinbm

    Member

  • Members+G
  • 2,625 posts
  • Location: Australia

Posted 06 June 2021 - 12:03

You could make a slit by sticking a needle in the floor or something a bit higher & get the lamp out of the picture So you don't over-expose the spectrum.