• Ultraviolet Photography
  •  

Candle Flame with UV Imager

31 replies to this topic

#1 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 23 October 2020 - 20:10

Below are my series taken using the KSS UV Imager with 60mm lens at F4 with Olympus 30mm Macro at F4 on a Panasonic GM5 positioned on the back imager.

No filter:
Attached Image: NoFilter.jpg

First image 254.3bp25 UVC image:
Attached Image: 254_UVC_first.jpg

303bp10 with U340 2mm to block all IR:
Attached Image: 303bp10_U340.jpg

313bp25 with U330WB80improved to block all IR:
Attached Image: 313bp25_U330WB80.jpg

313bp25 only (Some IR leak even with this):
Attached Image: 313bp25Only.jpg

335bp10 with U330WB80improved to block all IR:
Attached Image: 335bp10_330WB80.jpg

370bp15 with U330WB80improved to block all iR:
Attached Image: 370bp15_U330WB80.jpg

395bp25 fitler:
Attached Image: 395bp25.jpg

Lp 720nm filter only:
Attached Image: LP720nm.jpg

Last image 254.3bp25 UVC image (incase I could see hot spot in wick):
Attached Image: 254_UVC_last.jpg

#2 Stefano

    Member

  • Members(+)
  • 1,719 posts
  • Location: Italy

Posted 23 October 2020 - 20:25

Two things:
It seems the candle is inside a casing. Can its material influence UV images, especially at shorter wavelengths?

Your UVC images are similar to those taken by Wilhelm, but not "equal". In particular, the tip of the flame is not bright.

#3 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 23 October 2020 - 20:34

No the imager looks above the glass dish, so no influence or glass restrictions.

Sadly my images are not the same as Wilhelm's. I now thik his are IR leakage at 254nm. Mine are the same as here with UVC on the edges of the flame:
http://photographyof...-light.html?m=1

I wonder now if the 2020BSI is too IR sensitive and not enough UVC sensitive. Jonathan showed that BSI sensors were 2x to 3x more IR sensitive than FSI sensors. So I don't know if its the sensor being more sensitive to IR than UV and really bad IR leak on his 250bp filters or if the sensor isn't strong in UVC sensitivity.

Also he had dark water at 313nm. Thats not right. That is clearly IR high band leakage.

Edited by dabateman, 23 October 2020 - 20:35.


#4 Andy Perrin

    Member

  • Members+G
  • 3,995 posts
  • Location: United States

Posted 23 October 2020 - 20:43

Yeah, I was beginning to suspect as much when his results didn't match Klaus's.

#5 Stefano

    Member

  • Members(+)
  • 1,719 posts
  • Location: Italy

Posted 23 October 2020 - 21:36

Wilhelm, if you are reading this, try to put an IR longpass filter (such as Hoya R72) on the lens, with your UV filters, and see if you can see something and in case you can, how much leakage you have. As Steve (Cadmium) says, if you see something with the same settings, you have a leak.

#6 WiSi-Testpilot

    Member

  • Members(+)
  • 105 posts
  • Location: Germany

Posted 23 October 2020 - 21:47

David, please take the 250 nm image again, but in front of the sky. At 254/25 nm the sky is dark. Maybe there is some leak. I have documented the full spectrum leak and think it’s ok.
Stefano, I will try it tomorrow.
Best regards,
Wilhelm

Leak test at high gain.
Attached Image: Lighter_Mod_250nm.jpg

#7 Andy Perrin

    Member

  • Members+G
  • 3,995 posts
  • Location: United States

Posted 23 October 2020 - 22:03

WiSi-Testpilot, the thing is that the ratio of IR to UV of the candle matters a great deal here. If there is only a little UV and a lot of IR (as one might expect), the IR could dominate, regardless of what the tree/sky/roof look like. The IR from the lighter might be much stronger than all of the above, and the UVC negligible in the relative sense.

#8 WiSi-Testpilot

    Member

  • Members(+)
  • 105 posts
  • Location: Germany

Posted 23 October 2020 - 22:19

Andy, ok, I understand. Nevertheless I would like to see a flame infront of the sky.
Best regards,
Wilhelm

#9 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 23 October 2020 - 22:32

I will see if I can squeeze that in tomorrow. Are you using a butane lighter? I think I only have matches and candles. I may have some butane that I could light though. Will have to look through my chemicals.

#10 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 23 October 2020 - 22:40

Andy or Stefano,
How do we calculate the area signal?
Lets assume the quantum efficiency x fill factor plot to be correct for the 2020BSI sensor. Now lets assume OD3 blocking in IR for Wilhelm's filters, as thats common and 30% sinal as indicated in his plots.
How much IR is comming through over the sum of 700nm to 1100nm vs area under the 250nm peak?
I wonder if its very close

As in 400 x IR leak x QE vs 25 x 30%peak x QE.

Edited by dabateman, 23 October 2020 - 22:44.


#11 WiSi-Testpilot

    Member

  • Members(+)
  • 105 posts
  • Location: Germany

Posted 23 October 2020 - 22:41

View Postdabateman, on 23 October 2020 - 22:32, said:

I will see if I can squeeze that in tomorrow. Are you using a butane lighter? I think I only have matches and candles. I may have some butane that I could light though. Will have to look through my chemicals.
Thank you, I'm very excited. I think it doesn't matter if you take a candle or a lighter.
How long is your exposure time at 250 nm?

#12 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 23 October 2020 - 22:46

My exposure for the candle was 1/2.5 at ISO 200. My exposure for the LP 720 filter was 1/30 at ISO200. The sensitive is significant less at IR than UV for the imager.
I would assume about 6 stops more IR than UVC in the candle.

#13 Andy Perrin

    Member

  • Members+G
  • 3,995 posts
  • Location: United States

Posted 24 October 2020 - 00:21

View Postdabateman, on 23 October 2020 - 22:40, said:

Andy or Stefano,
How do we calculate the area signal?
Lets assume the quantum efficiency x fill factor plot to be correct for the 2020BSI sensor. Now lets assume OD3 blocking in IR for Wilhelm's filters, as thats common and 30% sinal as indicated in his plots.
How much IR is comming through over the sum of 700nm to 1100nm vs area under the 250nm peak?
I wonder if its very close

As in 400 x IR leak x QE vs 25 x 30%peak x QE.
I don't think that kind of calculation can be done without knowing the emission spectrum of the flame. Like, you would have to integrate (flame emission)x(lens transmission)x(filter transmission)x(fill factor)x(QE) I think, for each range?

Edited by Andy Perrin, 24 October 2020 - 00:21.


#14 Stefano

    Member

  • Members(+)
  • 1,719 posts
  • Location: Italy

Posted 24 October 2020 - 05:15

We don't know the exact spectrum of the flame, but Wilhelm linked one in his topic: https://www.dguv.de/..._gasbrenner.pdf

The lens shouldn't matter much, as it should have a flat transmission curve. Anyway it is quite a complex calculation, you have to do it for every point (like every nanometer) and then summing all points together to get the area under the curve.

#15 colinbm

    Member

  • Members+G
  • 2,287 posts
  • Location: Australia

Posted 24 October 2020 - 05:20

Shouldn't the flame spectrum represent the atoms being consumed ?

#16 colinbm

    Member

  • Members+G
  • 2,287 posts
  • Location: Australia

Posted 24 October 2020 - 06:44

I have taken a spectrum of a household domestic candle flame.....
Attached Image: 20201024CandleFlame-Ref.png

#17 colinbm

    Member

  • Members+G
  • 2,287 posts
  • Location: Australia

Posted 24 October 2020 - 07:24

Household domestic candle flame with S8612 on spectrometer...
Attached Image: 20201024CandleFlame+S8612-Ref.png

#18 WiSi-Testpilot

    Member

  • Members(+)
  • 105 posts
  • Location: Germany

Posted 24 October 2020 - 07:32

Colinbm, thank you, that is very interesting.
I made images with and without a 5 mm glass plate between the camera and the candle. Before opening the other thread I had tested it with the lighter, but not with the candle.
Later I will repeat it also with the lighter and with an IR Filter.

I need an additional short pass filter. Which one can you recommend?
Best regards,
Wilhelm

270 nm, 500 ms, with and without a 5 mm glass plate
Attached Image: with&without_glass_plate.jpg

Edited by WiSi-Testpilot, 24 October 2020 - 07:38.


#19 dabateman

    Da Bateman

  • Members+G
  • 2,699 posts
  • Location: Maryland

Posted 24 October 2020 - 07:52

That's the problem. There isn't a good short pass filter.
At 300nm and 313nm you can use U340 on its own. That helps cut out a lot of IR, but can't use U340 below 300nm.

U330 allows the lower UV. But also the higher IR. Excellent IR blocking filters are hard to get. I gone back and forth with a lot.

#20 colinbm

    Member

  • Members+G
  • 2,287 posts
  • Location: Australia

Posted 24 October 2020 - 08:24

Short pass filter is a dodo
It has been on my wish list a long time....