NatGeo and Fluorescence

[Damon]

Check it out. If you have already seen this I apologize.

 

https://www.national...raviolet-light/

[Andrea B.]

Those are beautiful!

 

(And of course I wonder if they are color accurate? :lol: :lol: )

[otoien]

Nice, thanks for the link Daemon.

 

..(And of course I wonder if they are color accurate? :lol: :lol: )

Andrea, wIth UVIVF, would not simply setting the white balance of the body to daylight provide accurate color, assuming blue emission from the UV light source has been properly filtered away ? (After all we are dealing with visual light). In astro-photography that is common practice to get accurate star colors with a non-modified body, together with color preserving processing steps.

[Guest]

C. Burrows does indeed make excellent UVIVF images. Admittedly, my admiration of his work is mixed with a dash of envy (in a healthy way, I like to think). Some time ago a friend of mine stumbled upon an article about him (different from the link above), and shared with me given my own interest in UVIVF imagery. I read that, and several other articles I managed to google about him (does that make me a creeper?! ;)). All in all, what I could gather is that he uses a homemade flat panel UV source; which he made himself with some number of LEDs (the type of which I do not recall at the moment). For prepping his subjects, he positions and stabilizes them by wrapping the stem in a heavy, bendable copper wire. As for filters, I do not know. Another thing I do not know is how he gets such even illumination, darkened background, and fantastic light balance (highlights are almost always detailed (not blown out)). On the other hand, one thing to note is that his images are not all strictly UVIVF - as many have ample amounts of VIS. This is not necessarily a bad thing - quite to the contrary, it was after viewing many of his images when I decided to try to push myself to start accepting VIS "contamination" in my own UVIVF images. I still have a long way to go with my own photography, and images like his just help to motivate me. I wouldn't mind sitting in on shooting session with that guy though, to see what I could learn from his techniques.

[Shane]

wIth UVIVF, would not simply setting the white balance of the body to daylight provide accurate color, assuming blue emission from the UV light source has been properly filtered away ?

and that's the problem .... define "accurate" and "assuming blue emission from the UV light source has been properly filtered away". Particularly difficult if excitation wavelengths partially overlap emission wavelengths.

​I shot fluorescence of gemstones for years, transitioning from daylight balanced film to digital using a "tweaked" Daylight WB but correction for excitation/emission overlap issues is tricky.

[otoien]

and that's the problem .... define "accurate" and "assuming blue emission from the UV light source has been properly filtered away". Particularly difficult if excitation wavelengths partially overlap emission wavelengths.

​I shot fluorescence of gemstones for years, transitioning from daylight balanced film to digital using a "tweaked" Daylight WB but correction for excitation/emission overlap issues is tricky.

 

In astro-photography the correct way to deal with light pollution is by subtraction, not multiplication (as when modifying white balance coefficients). I am not sure if this is directly transferable although It has some parallels: The light of interest is in both cases emitted. In astro-photography the unwanted light is directly hitting the sensor or more correctly reflected/dispersed by the atmosphere, while in UVIVF it is reflected off the object of interest. On the other hand if specles (for instance from ball bearings) hardly show up, the unwanted blue-violet light is unlikely intense enough to be reflected much. Fluorescence from surrounding objects being reflected by the object of interest has been a much bigger problem in my limited experience.

[Shane]

A tweaked Daylight WB was not used to correct the excitation/emission overlap but to fix an incorrect daylight WB implemented in the camera by the manufacturer. Correcting for/preventing excitation/emission overlap with filters alone is next to impossible. Correcting for excitation (as "light pollution") by subtraction is also difficult to implement because reflection differs subject to subject and area to area......in other words if excitation and emission overlap then how do you determine what percentage is contributed by excitation vs emission. This is much easier to perform using a luminescence spectrophotometer where excitation can be stepped nm by nm and the excitation contribution is quantified and subtracted from the emission.

[Alaun]

I think a normal WB is not applicable on fluorescence by definition.

WB assumes, you have a light source with a spectrum, that has deficits compared to a white light. You look (with the sensor) at surfaces reflecting part of that spectrum. With the WB, you correct that spectrum, that it fits to a reflected spectrum you would see with a white light source.

With fluorescence, your object emits light, probably with a very narrow spectrum. For simplicity, you could assume, it emits a monochrome color. Applying a conventional WB to that object, it would turn out grey. That is not, what you want! You need a “color measuring” WB setting, that the R G Bs represent the color of your active light source. And that is probably best done with a WB setting close to a setting you get with white light.

[Andrea B.]

EDITOR's NOTE: I needed to add the name of the other fluorescence standard, the UV Grey, to mention of the Target UV standard. Left it out the first time around.

 

 

from Øivind: Andrea, wIth UVIVF, would not simply setting the white balance of the body to daylight provide accurate color, assuming blue emission from the UV light source has been properly filtered away ? (After all we are dealing with visual light).

 

^{(Not sure I'll get all this right, but will abs make corrs if needed.)}

 

For starters, we do not know whether the temperature of visible fluorescence under, say, a 365 UV Led matches an in-camera Daylight temperature setting. (Or do we? I've never seen anything saying that visible fluor under 365 UV Led has a Daylight temp, but I don't claim to have read everything there is to know about visible fluor.)

 

So, just as the Daylight setting does not always work for accurate colour when shooting in shade or during early evening sunsets when making ordinary visible photographs, that Dayight setting might not always be the correct one for visible fluorescence photos. If anyone know differently and can provide a reference link, that would be way cool to learn about.

 

However, there is a way to standardize our visible fluorescence photographs so that we have at least a relative "white balance" for them. That is by using the Target UV or UV Grey grey fluor standards (which are quite possibly absolute standards, but I need to review that) OR any other fluor standard for setting WB that you might know about.

 

I buy these things and test them and write it all up, but you know -- I really don't know if anybody ever reads this stuff because this is about the 14th or 23rd time I've had to mention that there *is* a way to deal with standardizing visible fluorescence white balance. :lol:

 

But whatevs........sorry for the minor rant! Let's just say that I'm really really happy with the Target UV & UV Grey vis fluor standards. They are easy to use. They works great. And in my little head at least, I've solved the problem of visible fluorescence white balance. YMMV. :rolleyes:

 

Things I need to look up:

• What range of UV light can the Target UV and UV Grey handle?
  ^{Not that the UV range our cameras can handle is particularly wide being between about 320-400nm at best.}
  ADDED: See Post #14 below for the answer: http://www.ultraviol...dpost__p__19683
  

• How might visible light leakage from a UV illumination source affect the use of the Target UV or UV Grey?
  ADDED: See FAQ page: http://www.uvinnovations.com/faq They suggest use of a Wratten 2E to cut blue from illumination.
  

• What factors affect the temperature of visible fluorescence (assuming temperature of vis fluor can actually be defined)?
  

Target UV & UV-Grey Test References:

• Exp 1: First Look at Target-UV & UV-Grey for UVIVF [see Post 7]
  
• Exp 2: Target-UV with Stock Cam, Unfiltered Lens & Unfiltered UV-LED
  
• Exp 3: Target-UV with Stock Cam, Filtered Lens & Filtered UV-LED
  
• Exp 4: UVIVF White Balance with the UV-Grey Target, Stock Cam/Lens
  

*******************

 

from Werner: I think a normal WB is not applicable on fluorescence by definition. WB assumes, you have a light source with a spectrum, that has deficits compared to a white light. You look (with the sensor) at surfaces reflecting part of that spectrum. With the WB, you correct that spectrum, that it fits to a reflected spectrum you would see with a white light source.

 

Well, hmmm........cannot you also compare the visible fluorescence output to a white light? The visible fluorescence is also a "light source" in a sense?

Werner's comment is connected with my preceding comment about whether the Target UV or UV Grey are "absolute" standards. Are they giving us a WB setting in our converters which is equivalent to white light?

Even if not quite adjusting to a white light value, these Targets still do permit us to standardize our visible fluorescence photographs and make good relative comparisons.

 

 

**********

 

Try This Test?

Shoot some UV-induced visible red fluorescence output by the chlorophyll in a plant. It peaks between 650-700 nm. So let's say the peak is about 665 nm (looks close to that on most charts). The 665 nm wavelength can be correlated to the RGB color (255,0,0) which is at 0° on the color wheel. (Yes, there are some approximations happening to make that correlation.) After demosaicing your photo and applying your WB setting of choice, sample the red regions to test how close they are to 0°. The brightness and saturation of the red regions may vary, of course.

 

^{{{Andrea goes to dig out a red chlor fluor foto to see where the reds are on the colour wheel........}}}

 

 

***************

 

ADDED:

White light is 4000K? Daylight is 5600K, bluer. Do we have a firm definition in photography? I'm not sure how much to trust the Light Bulb people.

[Shane]

I think a normal WB is not applicable on fluorescence by definition.

 

I assume what you mean by normal WB is one that directly references the colour temperature of the emission source. In the case of fluorescence emission .... you are correct, a custom WB based on fluorescence would not work of course. I am suggesting using a Daylight WB.

 

WB assumes, you have a light source with a spectrum, that has deficits compared to a white light.

 

This infers an unbiased white source would not need WB, which is incorrect.

If that was so, then the white balance multipliers for white light would all be 1 (unity) and they are not. In fact the WB multipliers typically only attain a value of 1 when the image is slightly magenta in colour. In other words, only one specific colour biased target (slightly magenta, which differs for all camera models) would not need WB. This is the basis for creating a UniWB (which effectively is no WB because multipliers are all 1).

WB corrects the colour temperature of the source with respect to the overall sensor response, which is why pure white light would still require a WB to be applied.

 

With fluorescence, your object emits light, probably with a very narrow spectrum.

 

Most fluorescence emission in nature is broad band due to organic molecules. Fluorescence involving transition metals, rare earths, etc are typically narrow band but can also be broad band in some cases.

 

And that is probably best done with a WB setting close to a setting you get with white light.

On that we agree...... Daylight WB.

[Shane]

[[Off Topic]] EEK what the hell happened to my post?

[Andrea B.]

[[Off Topic]] wow! I don't know. It appears to have acquired some html formatting? I will try to fix it. Did you copy this from some other source??

[Andrea B.]

Shane did you look at the Target UV & UV Grey tests. When I compared the WB made using the Target UV - UV Grey to the Daylight WB setting, the Daylight setting was too blue.

[Andrea B.]

Found answer to one of my questions about the tuning of the Target UV or UV Grey: http://www.uvinnovations.com/faq

Excerpt:

The UV-Gray™ and Target-UV™ have been calibrated to appear neutral to the camera when excited by UVA radiation with a single peak of approximately 368nm. Using the UV-Gray™ card and Target-UV™ as suggested will create consistent images that mimic these radiation source conditions.

[Alaun]

Shane, I think we agree.

 

This infers an unbiased white source would not need WB, which is incorrect.

If that was so, then the white balance multipliers for white light would all be 1...

 

Yes and no. Off course white light is not like white light. As it depends on temperature by definition, it might need some different factors than 1, if you want it adjusted to a different temperature. (sorry, I thought, that was clear).

 

Most fluorescence emission in nature is broad band due to organic molecules.

 

My intention was to go a to an extreme case, to visualise/explain my point. (This is a principal I learned, when I studied physics. Often it is easier to understand something, when you look at the extreme cases :) ).

[Shane]

The UV-Gray™ and Target-UV™ have been calibrated to appear neutral to the camera when excited by UVA radiation with a single peak of approximately 368nm. Using the UV-Gray™ card and Target-UV™ as suggested will create consistent images that mimic these radiation source conditions.

I saw this back when they first published but have not had a chance to see their recent info.

when excited by UVA radiation with a single peak of approximately 368nm.

This procedure would appear to work fine for a narrow band UV source .i.e. the suggested 368nm but there are some issues here.

If a UV LED with an emission at 368nm is used (instead of broad band with narrow band excitation filter) then care must be taken with the power supply as it is easy to lower or increase the emission wavelength by incorrect power application.

The other issue I have, this assumes that everything you want to test has a 368nm excitation center, and although this is common excitation center, it is not necessarily found in all subjects that fluoresce. They may for instance exhibit an excitation center at 385nm and not 368nm and therefore no fluorescence would be detected. So here is where a compromise is made using broad band UV source for fluorescence, ....the broader the excitation source, the greater the chance of finding an excitation center but the harder it is to isolate the excitation from the emission when photographing.

[Shane]

Werner's comment is connected with my preceding comment about whether the Target UV or UV Grey are "absolute" standards. Are they giving us a WB setting in our converters which is equivalent to white light?

Sorry....did miss that connection to the earlier post. (Too many distractions at this end).

[Shane]

[[Off Topic]] Did you copy this from some other source?? I normally type in word or similar, in case there is not time to finish, then copy and paste. Never been a problem before but I did use MacbookPro Notes, so possibly the source of problem. Thanks for cleaning it up Andrea.

[Shane]

They suggest use of a Wratten 2E to cut blue from illumination

Again, sample dependent as the 2E cuts off some blue

http://www.beyondvisible.com/BV3-filter.html fig 10

[Shane]

Going through my old notes I noticed that in cases where I couldn't get satisfaction using filters and Daylight WB to remove excess blue, I often tweaked a Shade WB.

[Andrea B.]

Shane, thanks for all comments. Much appreciated.

 

 

I'm approaching this somewhat practically (I hope) in that 368 nm is "close enough" to the typical 365 nm torches (which we all use) which are most likely not really peaking at 365 nm but a couple of nanometers to either side. Thus the Target UV under 365nm torches can standardize the photographic output and bring it closer to an "ideal WB in visible fluorescence" (NOT well-defined, I really do know) than can a Daylight WB.

 

Whew.

 

Questions:

• Is there any point to attempting to standardize WB in vis fluor photography??
  
• I think it should be done. We should try, yes?
  
• But will it be an absolute standard or a relative standard?
  
• Or should we always present our vis fluor work in several different white balance settings?
  

Given that I have the Target UV and several UV-Led torches, I would be happy to try to perform any experiments which Shane, Ulf, Werner and other members suggest in order to determine the best WB setting for visible fluorescence work. That is why I got the thing in the first place and make all those initial experiments.

 

In an ideal world I could possibly measure the visible fluorescence output wavelength and then attempt to "translate" that into an RGB color and then try to set WB to produce that color. But I do not have the means to do so.

 

[That reminds me to look up the output wavelengths for the R, G and B patches on the companion Target UV patch card.