Jump to content
UltravioletPhotography

Non-polarized and cross polarized UV face photo


Recommended Posts

As some of you know I've been playing around with cross polarization in the UV for a couple of years now. When I started the work, my eventual aim was to do cross polarized in-vivo photography, as my main research is on skin. The big issue of course is getting polarizers with good UV transmission in decent sizes, which has limited my work to small light sources and small areas.

 

Yesterday though I was playing around with a studio flash (Bowens 600) and I decided to mount one of my old linear polarizers to the front of it, and try some cross polarized shots using the Nikon d810 modified for UV by ACS in the UK, a Rayfact 105mm lens and another linear polarizer on the front. SPF50 sunscreen on both cheeks. Flash was about 30cm from my face. Just me in he house, so focussing was guessed, and I don't have a remote release for the camera so I had to hit the shutter and rush to my seat.

 

I ended up with the following images.

 

Firstly, no polarizers

post-148-0-91079100-1560589117.jpg

 

And cross polarized

post-148-0-91745800-1560589125.jpg

 

I boosted the brightness and contrast a bit with the cross polarized image. Flash settings were waaaayyy apart - maximum for the cross polarized and minimum for the non polarized. Even then I needed ISO12800 for the cross polarized image. These basic polarizers will work, but are not ideal.

 

Even with this non ideal polarizer on the flash (it was a 72mm linear polarizer in a cardboard mount) it did a good job of cutting down the specular reflections from the cream and my skin. I could image with a bigger diameter filter, and 2 flashes it would be pretty good.

Link to comment
Andy Perrin
Interesting how it looks like you have a diffuser on the cross-polarizing one. I guess all you see is the diffuse reflection and the specular is cut out.
Link to comment

I think there is an element of that Andy, although I cannot rule out me being out of focus slightly when I sat down. A few mm each way will have a big effect. ISO was also higher for the cross polarised one - 12800 Vs 6400 for the non polarised one.

 

However I have seen this softening before under more controlled imaging too, which makes me think scattering in the skin is a big contributor to the observed effect.

Link to comment
I wonder... if your objective in using polarized/cross-polarized UV is to get a better view of the subject (by removing glares (specular reflections) that mask what lies beneath) then perhaps using a continuous source and light-painting could help with that? Or maybe that would make it worse? Just trying to think of something that could help (depending on your intent of course).
Link to comment

Mark, the eventual aim is a setup that can be used in a clinical test setting, for imaging panels of people in a reliable and reproducible way.

 

As for continuous sources or light painting, perhaps I just haven't worked through the logistics of using them. Off the top of my head though, reproducibility with light painting would be an issue. Potential exposure risks with continuous sources are also a big problem. Fluorescence can also have it's place - sunscreens darken the underlaying skin.

 

As we all know, UV imaging has plenty of 'quirks' and considerations above and beyond standard visible imaging. I have a way to go before I start with panels of people.

Link to comment
Andy Perrin
Polarization of the reflected light also depends on angle so light painting would probably negate the benefit you could get by putting the light source at Brewster’s angle.
Link to comment
Wouldn't it be practically impossible to place the light source at Brewster's angle... for a surface which is not constant (i.e., the curvature of the subject's face)? I wonder if employing coherent light would help? Though I suppose it would certainly be time consuming to illuminate such a subject with a laser (and also runs the risk of burning flesh!).
Link to comment
However I have seen this softening before under more controlled imaging too, which makes me think scattering in the skin is a big contributor to the observed effect.

 

The stuble growth seems to suggest slightly out of focus however something additional appears to be occurring as there seems to be some loss of UV contrast.

 

I would imagine that scattering "in" the skin plays an insignificant part since the short wave nature of UV light doesn't really permit transmission to any depth within the skin, unlike visible and IR. I am sure you have measured the transmission spectrum for your polarizers and know their performance in the UV as well as in the visible and IR. In combination with the internal UV modification of the camera and it's transmission characteristics in both the UV and IR, while under a strong presence of IR output from your source. I'm wondering if, when cross polarized, the UV content is reduced significantly with respect to any IR present and the internal camera conversion allows some IR transmission. The end result is a loss of UV contrast due to IR..... some loss of skin pore detail and your sunscreen is grey and not black.

Link to comment

Yes, I know the transmission of the filters, at least up to 850nm where my spectrometer gives out. I hadn't thought about IR leakage and that could defintely be a problem especially with the studio flashes. Thankfully something I can test in future.

 

However I still think scattering within the skin could be a factor, and I do not agree that UV doesn't really penetrate into the skin. These images are likely to be longer wavelength end of UVA (just becuase of the nature of the light source and camera sensitivity). UVA especially will penetrate down into the dermis, and on the face that can be a depth of between about 70microns up to 2mm. Even penetrating 100 microns and scattering would result in a loss of sharpness in the final image, although I get your point about the hair.

 

Early days, and I need to do more with it, to assess poor focus, IR bleed etc etc.

Link to comment

I am aware of the transmission to the dermis layer and was not saying it doesn't penetrate, I was making a comparison to the visible and IR penetration, which perhaps I didn't make clear enough. It is obvious from the greater resolution images obtained of the skin using UV vs IR, that scattering appears relatively insignificant.

 

These images are likely to be longer wavelength end of UVA (just becuase of the nature of the light source and camera sensitivity).

 

I can see there could be some greater penetration at the long end however, if this was the case and it was significant, how do you account for the sharpness and high UV contrast of the unpolarized image? Assuming the image is in focus, then perhaps your polarizers are passing the longer wavelengths more efficiently than the shorter wavelengths of the UVA-I range? or there is an IR leakage problem.

Link to comment
The polarisers drop steadily in transmission from 400nm to about 350nm, which is why I think this is skewing towards the long wavelength end. These aren't UV specific polarisers.
Link to comment

Jonathan,

I think you need a consistent reproducible subject, before we start to argue the fine details.

I also don't think its safe for you to hit yourself, your wife or a model constantly with UV to work out the variables.

I think Hog/pig skin is closest to human. See if you can go to local butcher and ask for a hock, or side with some skin or a little hair on it. Best would be if you could get a pigs head. That would most likely be the best subject to work on and optimize.

Then you don't have to worry about the UV while you get the test conditions correct. Also if you get a ham, you could eat it later.

Link to comment
Andy Perrin
If you use a pig’s head, you had better warn people before they walk into your laboratory (MWHAHAHA!) and are face to face with a decapitated pig head. It might be distressing.
Link to comment

Like Shane I also noticed the decreased contrast in the polarized photo.

 

But photographically speaking, perhaps just a reset of the black point was needed in the 2nd photo? Or maybe the 2nd photo was just overexposed a bit?

 

If you include a diffuse white standard and diffuse black standard in such a test, perhaps that would help determine what is happening and ensure proper exposure? A diffuse standard should look the same with/without polarizers.

 

In the polarized photo the UV-dark spots are more easily seen. So I think you're on the right track here to use a polarizer.

 

I'm wondering about converting the photos to B&W or greyscale before making comparisons?

 

Me, I'd get a real UV polarizer. Who sells them?

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...