How do you reference brightness and gamma in UV?

[Kai]

Since many substances absorb plenty of UV, the world often appears quite dark in UV. Unless the sky is part of the picture. The Skie likes to appear very bright. Sure, it also scatters a lot of UV light. This is certainly one of the reasons why daylight photos with a cloudless sky often appear softly lit.
When I was still doing the WB with anodized aluminum plates, I particularly noticed the darkness in the UV. Because with the aluminum I also had a kind of reference gray. PTFE is less suitable because our perception of brightness is not linear and we cannot differentiate light substances very well.

The following pictures illustrate my problem:

 

If I have PTFE in the picture, I can adjust a light value to it, e.g. 90% luminance. A point in the shadow defines black. Then the flowers look structureless black:

 

If I remove the PTFE, the whole picture looks much too dark:

 

Now I lighten up: I can see structures in the flowers and in the wood. However, this brightness does not correspond to the first picture with the Teflon. So it is totally overexposed:

 

If I increase the local contrast in the flowers, the result looks pretty nice. But now we are in the field of fine art ...

 

Now to my question:

How do you do it with the brightness and the gradation or the gamma?
Do you use standards (like PTFE for the WB) or do you just decide based on the visual result? 

 

 

 

[Andy Perrin]

I have always just adjusted based on what looks nice. As you demonstrated, the dynamic range is very far outside the usual range of what humans and cameras are accustomed to, so I think it’s permissible to curve the dark and light parts or use HDR techniques. 

[Fandyus]

...or try something like HP5 with it's huge dynamic range.

[Doug A]

@Kai great question. I struggle when deciding exposure on the camera monitor or computer screen. So far, I go for what looks good. 

Thanks,

Doug A

[colinbm]

UVA flash lighting.

[Cadmium]

I generally shoot outdoors, and I generally set the camera to aperture priority for exposure.

When I want a PTFE reference for the white balance, then I shoot a separate shot of the PTFE alone, using aperture priority and the same light as the other shots,

to white balance from separately, and then copy that white balance to the other shots in post.

 

[Kai]

Thank you all for your honesty. I suspected something like that - with a look at the pictures posted here.
With purely artistic aspirations, that's certainly a good thing. We don't do it any other way in the VIS or NIR range, right up to HDR or local adjustments with hand-drawn masks.

But for documentation purposes, where comparisons should be possible, that's not enough in my opinion.

 

What could be put into the picture besides the PTFE, which is color-neutral (at least in the range of 320-400 nm) and has a reflectivity of (ideally) 18% (or alternatively also 20% to 50%)?

I think that aluminum is less suitable because of the large angle dependence of the degree of reflection...

[Andy Perrin]

For documentation work, it should be done with Spectralon, which is a specially made PTFE material designed for calibration outside visible light. This is what Andrea and Birna use, I believe. For the display, you have no options if you don’t want to adjust contrast via HDR or curving. The dynamic range of the scene is simply outside the dynamic range of the display devices (if digital). Your camera is recording the information on the RAW, which is why it can be brought out by curving, but your eyes cannot see it and your screen may not be able to show it. Birna has been doing research for years in this area, so see what she says about her practices. 

[Kai]

2 hours ago, Andy Perrin said:

For documentation work, it should be done with Spectralon, which is a specially made PTFE material designed for calibration outside visible light. This is what Andrea and Birna use, I believe. For the display, you have no options if you don’t want to adjust contrast via HDR or curving. The dynamic range of the scene is simply outside the dynamic range of the display devices (if digital). Your camera is recording the information on the RAW, which is why it can be brought out by curving, but your eyes cannot see it and your screen may not be able to show it. Birna has been doing research for years in this area, so see what she says about her practices. 

Thank you for the tips and information, Andy!

 

I wasn't aware that Spectralon also offers gray standards. Investing in these reflection standards is out of the question for me - I also have other hobbies (and a family;)

 

So I can't avoid my own experiments. Let's see, I'll print transparencies with the laser printer one day. The result is also matt and the black pigments seem to be sufficiently color-neutral in the UV (at least in the range that is available to me). I then put that on my hand-sanded Teflon and then take test photos. If something comes out of it, I will report.

 

Mostly "pictorial" photography is enough for me. None of that matters.

But sometimes I just want to take factual information from the picture, then it takes more than just the WB...

[photoni]

3 hours ago, Kai said:

Mostly "pictorial" photography is enough for me. None of that matters.

one of the most disconcerting things that I realized reading this site is that the naked sensor (with Bayer filter) has a senseless UV and IR sensitivity.

There are some parameters that can only be called "standard" using the original Baader and Schott filters

which are too expensive for me. For the moment I am satisfied with what I like.

 

[Andrea B.]

Visible digital files are adjusted for dark, medium and bright tones by automatic application of a curve to the raw data which matches (as best it can) how our human eyes see. Then after conversion we adjust tones as we choose for artistic effect.

 

For reflected ultraviolet digital files, the same curve would be applied. But it is not always quite accurate for UV tones. The best we can do following conversion is to use some black, mid-gray and white standards which are stable under UV light (in the sense that they reflect black, mid-gray and white tones in UV) to set the black, mid-gray and white tones in the reflected UV photo. There are such standards made from Spectralon, a Lambertian diffuse reflectance material. (I don't recall exactly what is added to the Spectralon to make the grey and black tones.) You can also make such standards using certain materials. Jonathan, JMC here on UVP, has an excellent post showing the standards he made. 

 

I will go get the links to post here.

 

Labsphere Spectralon Calibrated Standards

These are ridiculously expensive!!!

https://www.labspherestore.com/product-p/as-0116x-x60.htm

 

Jonathan Crowther (JMC):  DIY Standards discussed in this topic.

A Calibrated Gray Scale for Forensic Ultraviolet Photography

Scroll down to JMC's first post.

 

 

 

 

And let me also post a link and a reflected UV photo showing how I have used the Spectralon standards for a Visible photo.

https://www.ultravioletphotography.com/content/index.php?/topic/844-lumix-gh1-broadband-white-balance-colour-correction-in-photo-ninja-midtone-curve-discussion/&tab=comments#comment-3370

(The discussion in this linked topic is applicable to any camera.)

 

The first part of that linked topic was about using Spectralon standards for obtaining a good white balance under different filters. The second part of that topic discusses using the 5 different standards for obtaining a visible tone curve. After conversion of the raw visible file, you can see in the first photo that the brightness of the standards in the photo does not "match" their calibrated reflectiveness. {Note that I am making an assumption that it is OK to do that == i.e., assume reflectance and brightness are correlated.] The "linear" tone curve made from the 5 standards (last visible photo) does not look quite right because our eyes do not see that way.

 

 

However, in both visible light and reflected UV light, I think you can use only the white and black and mid-grey standards  to create a nice view with good contrast. (Looking for that link now.)

LINK_GOES_HERE: I couldn't find a good one. So I'll just put a new version in the next post.

***********

 

But before I go get those links, let me add that if you don't have a set of Standards to work with for reflected UV, then remember that our cameras, both stock and converted, have a narrower dynamic range than our eyes. So converted cameras are not quite capable of capturing in one photo both the brightest UV-reflecting tones together with the darkest UV-absorbing tones. Thus you are always going to have to either pull back your highlights or pull up your shadows in a reflected UV photo, unless you shoot 2 or 3 frames to combine in an HDR manner.

 

My personal preference for a UV-absorbing, dark subject in reflected UV is to bring it up enough to show some details. Else what is the point of having a reflected UV photograph of that subject?

 

[Andrea B.]

Using black and white standards only in a reflected UV photo.

 

Photo as shot.

 

 

White balanced version. (Needs improved contrast.)

This white balanced conversion made in Photo Ninja applies the usual "gamma" curve intended for visible light. It does not bring the black standard to a 2% brightness nor the white standard to a 99% brightness.

 

 

Black & white point version. (This is the best version.)

I took the white balanced photo into NX2 and set the black and white points (on the Luminance layer). The black standard remained at 10/11% brightness while the white standard came up to a better 98% brightness. This photo looks better than the preceding one due to the improved contrast. Notice, however, that the 50% reflective grey standard has become even brighter.

In the next photo I'll show you that the photo "regresses" when I attempt to bring the 3 standards to 2%, 99% and 50% brightnesses.

 

 

Curve version. (Too dark!)

Using a curve (on the Luminance layer) in the NX2 app, I tried to wrangle the standards into a 2%, 99% and 50% brightness. It is impossible to attain that precisely, but I got close enough. However, this would not be a very good reflected UV photo because its tones have a linear adjustment, but our eyes do not see that way. In my opinion, using just the black and white standards to set the black & white points is enough.

 

 

 

 

Comment

Someone might have other method or ideas for dealing with tones in reflected UV photography. I certainly don't claim to have entirely figured out what is the best approach to this which will maintain good contrast and reveal details in both the dark and bright areas.

 

General Remark about Reflectance Standards.

These are intended for calibration of various optical instruments. They can also be used to determine the approximate amount of UV-absorption or UV-reflection of a particular subject in a reflected UV photo by comparing the subject to a standard. But only use the black and white standards to deal with reflected UV tones.

 

 

 

 

 

[Andy Perrin]

Didn’t Jonathan make a gray gradient using alabaster or something at one point? That won’t work in the field but might do in a lab setting. And not expensive. 

[Kai]

11 hours ago, Andrea B. said:

 

 

 

Curve version. (Too dark!)

Using a curve (on the Luminance layer) in the NX2 app, I tried to wrangle the standards into a 2%, 99% and 50% brightness. It is impossible to attain that precisely, but I got close enough. However, this would not be a very good reflected UV photo because its tones have a linear adjustment, but our eyes do not see that way. In my opinion, using just the black and white standards to set the black & white points is enough.

 

 

All of this is very interesting and also very helpful to me. Thank you very much!

To me, it looks like the world in the UV is an essentially dark world. A low-key world. Just a bright sky, bright ice, some bright stones. In addition, some strong reflections and highlights. The rest more or less dark.

That's not surprising when you look at how light interacts with matter depending on the wavelength. Extended electron systems that can be excited with visible light are also rather special or even rarer in the world of molecules.
However, a vast number of molecules can be excited with UV light, especially the widespread aromatic systems with UV-A. In other words: a lot of organic substances, especially those coming from living systems, absorb UV light relatively strongly. On the other hand, many mineral substances reflect UV light comparatively well.

If we now want to recognize the world in UV and understand how it looks in UV, maybe we have to accept dark images.

 

Is this how the world looks in UV? 

 

 

 

 

Changing the gradation or even HDR recordings may lead to an "alienation" in the overall impression. But this is of course helpful to show differences in struktures - as shown in my examples of this post.

But the transformation should be verifiable. For example, the three Spectralon (R) objects could simply be included in a picture as internal standard.

 

I am mainly concerned with "artistic" photography. I can do what I want there.
Sometimes, however, I am more interested in scientific investigation and observation. Then - I'm afraid - I cannot avoid the procedure that you have shown here ("curved version"). So I need internal standards for comparison. However, I have to find cheaper alternatives to the Spectralon (R) reference plates...  ;)

 

 

 

[Andrea B.]

I'm very happy to know that was helpful! I love to share what I've learned and then get refinements and feedback from other members.

 

Please do look at the link for the do-it-yourself standards made by Jonathan (JMC). PTFE is very useful as a white standard as long as care is taken not to overexpose it. And JMC's magnesium oxide black standard is not difficult to make.

 

I do think you can easily set the black and white points for any photo without using standards because most converters have some kind of histogram tool which permits adjustments of the black end point and the white end point. (I set those only on the Luminance layer and not on the Chrominance/Color layer.)

 

Also check your particular converter to determine whether there is any manual setting for the minimum black point and the maximum white point. I don't use 100% (255,255,255) for white point settings. Bring it in a little bit to 99/98% (252/250,252/250,252/250), approximately. Similarly setting minimum black point to 0% (0,0,0) is too hard on the photo. I usually use 2% (5,5,5) as a minimum setting.

 

Yes, reflected UV is mostly dark, but there should be some detail in all shadowed areas for visual appeal. Just as blocked shadows are to be avoided in Black & White Photography, they should be avoided in reflected UV photography -- unless, of course, you are going for some kind of extreme artistic effect.

Anything I say like that is an opinion or a suggestion and not a "rule". {.....laughing.....}