• Ultraviolet Photography
  •  

Anybody have some alabaster? WB in fluor photos?

87 replies to this topic

#21 UlfW

    Ulf W

  • Members+G
  • 1,830 posts
  • Location: Sweden, Malmö

Posted 26 February 2020 - 12:56

I am confident that it would be rather easy to design a tuneable light source for white balance for UVIVF-exposures.
There are LEDs (RGB) with wavelength-peaks in the same range as the much wider R-, G- and B-peaks of common bayer matrixes.

I would suggest two or three controls.
  • Intensity balance between red and blue LED.
  • Intensity control of green LED (independent of (1)).
  • General intensity setting of all three LEDs
If such a LED cluster is placed reasonably far away from a diffusor the light will be mixed evenly.

As the device would be used just temporarily it can be battery operated.

The device would just be for creating a small patch glowing with a controlled white light, not for illumination of other objects.

Edited by UlfW, 26 February 2020 - 12:59.

Ulf Wilhelmson
Curious and trying to see the invisible.

#22 Stefano

    Member

  • Members(+)
  • 2,174 posts
  • Location: Italy

Posted 26 February 2020 - 13:40

View PostAndy Perrin, on 26 February 2020 - 07:05, said:


I am not sure what is meant by "the monochrome spectrum" here. Most fluorescent photos that we take involve a wide variety of substances in the same picture each with their own fluorescence spectrum. The results are anything but monochrome.

That said, I suspect not all the light is within the gamut of computer monitors. I think it was Mark who mentioned that there was a particular color of red fluorescence he could see in real life but couldn't reproduce with his photos?
When I tried to find the approx. RGB coordinates of a 600 nm orange LED and a ~500 nm cyan one, I noticed (especially with the orange) that the color given off by the LED had at least twice the saturation of the orange on my LCD computer monitor. Also, deep red (660 nm) and far red (736 nm) are RED, the reddest red you will ever see. On my phone, 255, 0, 0 looks orangish-red in comparison.

#23 Akira

    Homo jezoensis

  • Members
  • 105 posts
  • Location: Tokyo, Japan

Posted 26 February 2020 - 13:43

View PostAndy Perrin, on 26 February 2020 - 07:57, said:

Yeah, I don't think fluorescence is usually a line spectrum like you get with a gas. There are examples here:
https://www.thermofi...ce-spectra.html

Thank you for the link! That is very helpful, and in Japanese!

#24 Andy Perrin

    Member

  • Members+G
  • 4,416 posts
  • Location: United States

Posted 26 February 2020 - 15:50

View PostUlfW, on 26 February 2020 - 12:56, said:

I am confident that it would be rather easy to design a tuneable light source for white balance for UVIVF-exposures.
There are LEDs (RGB) with wavelength-peaks in the same range as the much wider R-, G- and B-peaks of common bayer matrixes.

I would suggest two or three controls.
  • Intensity balance between red and blue LED.
  • Intensity control of green LED (independent of (1)).
  • General intensity setting of all three LEDs
If such a LED cluster is placed reasonably far away from a diffusor the light will be mixed evenly.

As the device would be used just temporarily it can be battery operated.

The device would just be for creating a small patch glowing with a controlled white light, not for illumination of other objects.
So, like, you could shine it on some PTFE or a color-checker and snap a photo for white balance afterwards?

#25 UlfW

    Ulf W

  • Members+G
  • 1,830 posts
  • Location: Sweden, Malmö

Posted 26 February 2020 - 17:32

View PostAndy Perrin, on 26 February 2020 - 15:50, said:

So, like, you could shine it on some PTFE or a color-checker and snap a photo for white balance afterwards?
Either that or shine it through some suitable diffusing material. I think that would be more efficient, but either method will work.
You do not need a neutral reflective surface to make it work.

As the LEDs have a narrow peak and the bayer dyes different sensitivity bumps have slopes the system will be slightly sensitive to wavelength shifts.
If the LEDs run at reasonably low stable current their peaks will be relatively stable too, as they will not get much heated.

I can happily build something like this, but lack a suitable setup to test it in, as I do not have any space available for a UVIVF shoot.
My apartment is filled to the brim. It almost looks like Birna's famous car trunk filled with photo stuff.

Edited by UlfW, 26 February 2020 - 17:37.

Ulf Wilhelmson
Curious and trying to see the invisible.

#26 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,115 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 27 February 2020 - 08:10

I thought 'white' was equal amounts of RGB, but then we add near UV & near IR, so how does that change things.....?

Those who enjoy UV and/or IR photography in any combination typically take "white" to be the reflection from a material such as Spectralon which diffusely reflects all light equally between about 300 - 1000 nm. (Look up the specs if you want a precise range there. "-) )

Using the term "white balance" for white-click processing on a neutral area of a UV photo is, of course, not really an accurate use of the term because there is no "colour temperature" for false colours. But the method works to produce a standardized outcome which is desireable in documentary work.

I've always been very careful to stress that white balancing a UV (or an IR) photo is simply for standardization across different gear and processing platforms. It is certainly not mandatory.



As a white reference for UVIVF, do the reference itself have to have fluorescence or can a controllable very weak white light source work?

My personal preference is to have a reference which itself fluoresces white. But if a very weak white light source works, then why not? I have not experimented with that. So if you make such an experiment, we will all enjoy reading about it!



I'm not sure if the accurate color of the fluorescence can be recreated on a computer screen regardless of the accurate calibration.

That statement can be generalized: not every reflected visible colour can be recreated on a computer screen. Monitors have become much more sophisticated in recent years and now have much wider gamuts (colour spaces), but every colour space leaves some colour out. We usually do not notice that some colour is "missing" from the typical JPG viewed on a browser. And not to forget, your camera also has a gamut which cannot produce every visible reflected colour.



I think it was Mark who mentioned that there was a particular color of red fluorescence he could see in real life but couldn't reproduce with his photos?

And I have also seen this: some olive oil which I was fluorescing under a Nichia 365 nm torch produced a very intense green which I could not match during processing. It was really pretty.



The other problem is white is different to different people.

Subjectively, perhaps. But -- when using a reflective standard, we don't have to rely on the subjectivity of our eyes & brain. What we are aiming for with reflective white standards of any type is to eventually produce on the monitor an RGB "colour" for which the values R=G=B. Like, (0,0,0), (128,128,128), etc.



But this is the issue: are we aiming to duplicate the colors that we see with our eyes when the objects fluoresce? Or are we trying to match a daylight white balance? These aren't the same. Or do we find a material that fluoresces in a way that looks "white" to humans and use that to white balance with?

My take on this is the same one that UV Innovations has produced. We want to standardize the outcome of a visible fluorescence photo by using a standard-which-fluoresces-white to set an in-camera white balance under the conditions in which we are shooting the fluorescence. Or, alternately, we standardize the outcome of a visible fluorescence photo by photographing the standard to use for a white-click area when converting the raw file.

My experience with using this standard-which-fluoresces-white is that it it does indeed produce a reasonable match of the fluorescence I'm seeing while shooting. But do remember that the UV Innovations standard-which-fluoresces-white is tuned to a particular set of filters and a particular area of UV light (365-370 nm, I think?) It will not be entirely accurate under other filters or UV ranges.



Can I make one more small speculative statement? I don't think our vision goes entirely to scotopic when sitting in the dark closet while photographing visible fluorescence. Because we are looking intently at the visible light from the fluorescing subject. What do you think?



Alternative opinions, corrections or even agreements(!) are welcomed for the preceding attempt at explaining the extension of the concept of white balance to UV or UVIVF to produce standardized outcomes.
Andrea G. Blum
Often found hanging out with flowers & bees.

#27 colinbm

    Member

  • Members+G
  • 2,741 posts
  • Location: Australia

Posted 27 February 2020 - 08:49

All very interesting & especially when all put together like this....
Equal amounts of RGB give white light, but so does over saturation of one colour on the receiver / sensor.
Because UVIVF is a low intensity light, anyway, I wonder if there is a 'colour' under the white at a lower threshold / intensity.
I am not properly set up yet, but I will like to find out...?

Edited by colinbm, 27 February 2020 - 08:50.


#28 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,115 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 27 February 2020 - 19:57

Col, I'm not sure what you mean about over saturation of one colour producing "white"?

Because UVIVF is a low intensity light
Depends on what you are fluorescing. If you fluoresce those text highlighter markers, they can be quite bright. And fluoresced optically brightened paper is a very bright blue. Flowers, of course, not so much brightness.



Oh. That reminds me. Fluorescent text highlighters are good test subjects for getting your UVIVF "white balance" correct.
I just bought a pack of highlighters which has 10 colours. I've been meaning to test whether they are all fluorescent.
Andrea G. Blum
Often found hanging out with flowers & bees.

#29 Stefano

    Member

  • Members(+)
  • 2,174 posts
  • Location: Italy

Posted 27 February 2020 - 20:31

View Postcolinbm, on 27 February 2020 - 08:49, said:

Equal amounts of RGB give white light, but so does over saturation of one colour on the receiver / sensor.
...Yes, kind of. When you look at a bright monochromatic light source, or when you do that with a camera, you see white (everything "blows out") because the light is so intense that it actually oversaturates ALL channels (in a camera) or ALL cones (in your eyes). For example, in a camera sensor, as you know, there is the CFA. Those filters are not perfect, and let in a small quantity of unwanted wavelengths (for example, the blue channel filter will be slightly transparent to red and green light). If those filters were perfect, you would just see a strong, saturated color. Using red light as an example, your camera will probably see red at first, then orange and yellow (the green channel is the first one to be affected) and then white.

#30 Alaun

    Member

  • Members(+)
  • 324 posts

Posted 27 February 2020 - 23:04

Stefano, if you use a monochromatic light to illuminate your scene, you will always get a BW picture, when you do a white balance (assuming you do not have fluorescence)
Werner

#31 Stefano

    Member

  • Members(+)
  • 2,174 posts
  • Location: Italy

Posted 27 February 2020 - 23:43

View PostAlaun, on 27 February 2020 - 23:04, said:

Stefano, if you use a monochromatic light to illuminate your scene, you will always get a BW picture, when you do a white balance (assuming you do not have fluorescence)
Yes, I agree. You wouldn't have color information. I used that example to explain the concept Col talked about. I was thinking to say that any light which isn't white can become white if the sensor is saturated, but we are having a hard time defining "white" in the first place, so it would have been misleading.

#32 Alaun

    Member

  • Members(+)
  • 324 posts

Posted 27 February 2020 - 23:51

Actually you have a color information, when your sensor is white balanced to „normal“ white light.
Werner

#33 Stefano

    Member

  • Members(+)
  • 2,174 posts
  • Location: Italy

Posted 27 February 2020 - 23:57

View PostAlaun, on 27 February 2020 - 23:51, said:

Actually you have a color information, when your sensor is white balanced to „normal“ white light.
Yes, you have information about what color it is, but once you WB that image, it becomes B&W (as you said), and you will not be able to distinguish any hues.

#34 Alaun

    Member

  • Members(+)
  • 324 posts

Posted 28 February 2020 - 00:04

And that is exactly the problem with fluorescence, as the only light you get is that from the fluorescence, which might be close to a monochromatic light source.
Werner

#35 enricosavazzi

    Member

  • Members
  • 602 posts
  • Location: Kalmar, Sweden

Posted 29 February 2020 - 09:09

View PostStefano, on 25 February 2020 - 19:50, said:

I can easily start to see my cyan LED as white at a high-enough brightness. I'm not talking about staring at the LED, but about illuminating white objects in a ~30 cm wide area, concentrating there all the (maybe) 3 W of radiation emitted by this 8-9 W LED. I really see it as white at some point. The same worked with a green LED, it looked like a mint white after minutes staring at the output concentrated in a ~15-20 cm wide circle.

Regarding vision in the dark, rods peak in the blue-cyan region. I once played with a little blue LED (the epoxy type, but smaller than the 3 mm ones), and I saw blue near it. In the distance, its light appeared white, colorless. I could still see shadows, I could see my fingers, but the color blue disappeared.
From my darkroom work decades ago, I remember working several times with orthochromatic film under a red safety light. Within minutes my eyes adapted and the red looked white to my eyes. A red plastic tray of developer solution on the table looked just light gray, and a green tray a darker gray. Our vision system is a poor judge of color unless there is something else of another color (or a white standard) simultaneously in the field of vision to compare colors with, and performs even worse when the illumination source is not full-spectrum.
-- Enrico Savazzi

#36 Stefano

    Member

  • Members(+)
  • 2,174 posts
  • Location: Italy

Posted 29 February 2020 - 10:33

View Postenricosavazzi, on 29 February 2020 - 09:09, said:


From my darkroom work decades ago, I remember working several times with orthochromatic film under a red safety light. Within minutes my eyes adapted and the red looked white to my eyes. A red plastic tray of developer solution on the table looked just light gray, and a green tray a darker gray. Our vision system is a poor judge of color unless there is something else of another color (or a white standard) simultaneously in the field of vision to compare colors with, and performs even worse when the illumination source is not full-spectrum.
Also, think about the Bezold-Brücke shift. We can't even perceive the same color at different intensities. For me, at higher intensities, violet becomes bluish, green becomes also bluish (towards cyan), and red becomes orangish. Also, the only pure color we can perceive (in normal conditions) is red. My 660 nm deep red LED is almost too red, it's like looking through red wine, it's that kind of red. It almost appears purplish, like blackberry juice. My ~736 nm far red LED is even "stronger".

#37 Ninjin

    Member

  • Members(+)
  • 98 posts

Posted 03 March 2020 - 17:25

I have some samples of white pigments I use, and there is a sample of gypsum - №4.
(All pigment powders mixed in chicken yolk + vinegar.) On the samples: half on the left - 1 layer of paint, on the right - two layers.

This photo with MTE 301 filtered 340.
Complete darkness. UV block - ЖС4 (420) glass on the lens.

I do the white balance the way: in NX-D, in the gray area on the cotton watercolor paper (Canson), which is always the same color.
I use this paper for targets, and as background, stable light gray color.
And color of luminescence on this is very close as I see with my eyes. Almost exactly the same.

Such a balance for pigments is the best I've found so far.. Of course I would like to improve it yet,
but I don’t know which way.
I do not have a 'special' target, so I selected this paper as a target, from many materials I checked.

Attached Image: mte301_340_w.jpg

This paper also works well in reflected UV.

Attached Image: uv360_w.jpg

#38 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 9,115 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 04 March 2020 - 19:57

I use this paper for targets, and as background, stable light gray color.

Am I understanding correctly that Canson watercolor paper is useful as both a UV-induced visible fluorescence standard as well as a reflected UV standard??
Andrea G. Blum
Often found hanging out with flowers & bees.

#39 Alaun

    Member

  • Members(+)
  • 324 posts

Posted 04 March 2020 - 21:16

In many papers, CaO is used to improve properties e.g. like surface smoothness.
Werner

#40 Cadmium

    Member

  • Members+G
  • 4,247 posts

Posted 05 March 2020 - 06:10

View PostAndrea B., on 04 March 2020 - 19:57, said:

I use this paper for targets, and as background, stable light gray color.

Am I understanding correctly that Canson watercolor paper is useful as both a UV-induced visible fluorescence standard as well as a reflected UV standard??

Andria, I think the idea here was because you wanted a sample of alabaster, and the gypsum test sample #4 is related to that (?).