PTFE in UVIVF?

[Guest]

While testing an equipment setup for shooting UVIVF I saw something I don’t quite understand, yet. Hoping someone can help explain this.

 

I shot this series of images using the equipment set above, changing only the on-camera filter for each. There are two test subjects (both are gaskets used in sanitary piping connections): one is silicone (the larger, translucent outer disk), and the other is PTFE (the smaller, “white” inner disk).

 

My intention was twofold: 1) to test my on-hand filter options to see which might work for filtering out deep blue fluorescence and blue/violet light bleed from the FL-02 filter on the light source, and, 2) to see if I could use the PTFE as a white balancing target.

 

Unfortunately, the PTFE target showed strong coloration that I could not white balance. My question is: why is this so? Could the PTFE be fluorescing (in the red region?)? Is PTFE appropriate, and if not, what would be better?

 

Aside from my question about the PTFE target, some info I take from this is:

• The flash/filter combo works well for exciting VIS fluorescence (notice all that fluorescing lint stuck to the silicone!)
  
• The B&W023 (#9) filters out all the blue light – but cuts too far for my taste.
  
• The Hoya R72 (#12) shows that there isn’t really any IR forming the image – just barely a hint of red; which I’m not bothered by.
  
• The Hoya Y (#3) and G (#4) appear to do the best job, without overdoing it – I’m still undecided on which I prefer.
  

Until I decide to purchase a dedicated cut-on filter for this, I think I at least have a filter or two on-hand that will serve my purposes well enough.

 

There’s still the question of the PTFE though. Any info on this, and/or suggestions on what may be better, would be great.

 

Thanks,

- Mark

[JCDowdy]

I have seen similar results on Spectralon when stacking filters on a strobe for UVIF so I doubt you have much concern from PTFE fluorescing assuming your gasket material is virgin.

 

What you clearly have is a good deal of overlap, aka leak, between your excitation strobe spectrum and your imaged emission spectrum. Ideally, if all you want to image is pure fluorescence there should not be any overlap and a very high OD barrier between the two as in matched fluorescence microscopy filter sets.

 

Think of it like this, if you stacked your strobe filter onto your lens filter almost no light (<10^-4) should get through. Andrea often mentions the spoon test in discussions of this perennial topic which is another simple way to see barrier leak.

[Cadmium]

Hi Mark, What is the material in the very center, that looks tan/white in the visual image and white in the other images?

[Guest]

I understand, and completely agree, that to achieve true fluorescence-only images the excitation and emission filters must be matched, to avoid any overlap. And I will do exactly that - just as soon as my wallet fills back up again (this is actually a very rare event!). Until then, I'm just looking to see how closely I can approximate UVIVF with the equipment and hardware I already have on-hand. I think what I have is okay, for now.

 

In the past (with other hardware) I've used a stainless steel ball bearing (re: spoon test), and have been able to achieve almost complete separation (of excitation/emission) without using a matched filter set. So, I am a fan of the [metal surface to check filter leak/overlap test]. For this setup though, I'll of course need a better cut-on filter, eventually.

 

As for the material in question, in the center of the targets... That's a tooth/fang protruding through the center hole. In the background, behind the silicone material, are the rest of the bottom teeth of a racoon. A polyurethane coated racoon skull, that is. I think it was a sort of trinket purchased from a local fair a long time ago - which I just found in the attic the other day. And it coincidentally happened to be a good holder for my test subjects :)

[Cadmium]

I didn't have my racoon scull handy, but I tested a few things with the PTFE.

First I tried using both U-340 and UG5 on my MTE with no GG420 on my lens. These make a more red PTFE with the in camera white balance I had.

Then I tried the same shots with the GG420 on the lens, and the PTFE is more gray.

I also had my WhiBal in the shots, which looked more natural.

I don't know how much UV your internal camera filter cuts. And I also don't know what your light filter is passing.

I have never tried using PTFE for UVIVF before. I have usually used marquee in CNX2, or some part of the photo that makes it look natural, without using any WB target.

You might ask Alex what he uses for UVIVF white balance.

[Alaun]

Why do you want the PTFE to look white?

First, with Photoninja, it is no problem to run a WB on the little disc, but …

You do a white balance, when your light source does not have the spectrum of a white light source and when you want the reflected light to look like the reflected light would look, when coming from a light source with a white light spectrum.

With fluorescence, you want to see the color of the fluorescent light. So you need a white balance, that your sensor&camera&developing system presents the correct color. This is probably done closest (in a first attempt) with a WB color setting representing a white light source spectrum?

A further concern: The fluorescent light often is close to a monochromatic light. With a monochromatic light source, a WB would give a black and white picture, but you might have a couple of different fluorescensing substances.

Back to your PTFE disc: as a “standard” it should not show fluorescence at all, so with your setting it should be black, only slightly illuminated by the flurescensing other substances (so any color with respect to the mixing of the fluorescence light coming from the near by stuff.)

 

WB with UV- or IR-reflective pictures is a different story.

[Andrea B.]

Teflon has good resistance to UV degradation. It could eventually become contaminated by pollution in the atmosphere (or other contaminants) and then possibly exhibit some kind of fluorescence. But I seriously doubt that your PTFE is fluorescing in any way at all under that UV ring flash.

 

What you are seeing is the reflection of the light shined upon the Teflon that has not been filtered out by the illumination filter nor by the lens filter. Although I'm not sure about that coated raccoon tooth. There might be a small bit of fluor there. Human teeth fluor, so prolly also for raccoon teeth.

 

The colour of that reflected light in the photos presented above depends on (among other factors) the Bayer filter on your Canon 5D, the internal UV/IR blocker in the 5D, the UV ring flash filter and the filter used on the lens. It also depends upon the 5D white balance setting in use during the shoot and whether your conversion software preserves that WB or not.

 

In photos [2, 5, 6, 7, 8, 11], I'm thinking that you probably have recorded some violet or violet/blue leakage and maybe some high red also.

 

In photos [3, 5, 9. 10], it is probably some high red leakage between 690-710 nm.

 

The FL-02 lamp filter can pass both violet and red. Not very much as per its transmission chart, but then it doesn't take much for it to show up in a photo.

 

Internal UV/IR blocking filters usually can pass violet, violet/blue and high red. They're usually just some kind of BG glass.

 

You did not mention the exposure data for these photos?

f/x for y" @ ISO-zzz

Knowing that would help determine just how reasonable it is that we are seeing violet or red leakage being recorded. ;)

 

 

********

 

For Visible fluorescence photos, we typically suggest using a Daylight or Auto White Balance setting because there is no way to make a colour profile when the subject is illuminated with UV light for production of Visible fluorescence effects. Colour corrections must be made "by hand". I usually take some test shots and then adjust colour balances on the test shot while trying to actually look at the fluorescence. This is tricky, to say the least.

 

Note that if we wanted to be able to "properly" white balance in UVIVF work, we would need to find some subject which happily fluoresces R+G+B equally to give "white" under UV stimulation. This is not going to happen.

 

I think the OG530 did the best job of supressing the blue/cyan fluorescence.

 

END of COMMENT.

 

 

 

**********

The following are just my notes on those filters. I'm not familiar with most of them, so could not fully interpret your results until I learned them. Everybody please skip this part. :D

 

[2] no filter

[3] Hoya Y(K2): Longpass filter which cuts in at 460 nm. (Blocks UV, violet & some blue.)

[4] Hoya G(X0): Longpass filter which cuts in at 525 nm. (Blocks UV, violet, blue, cyan & some green.)

[5] B+W 415 + 489.

[6] B+W 415 (GG400): Longpass filter which cuts in around 380 nm.

[7] B+W 489 (KG3): Transmits UV, Visible and near IR. [Question: What is this filter supposed to be for??]

[8] Lexan: Polycarbonate plastic which transmits from just below 400nm.

[9] B+W 023 (OG530): Longpass filter which cuts in at 530 nm. (Blocks UV, violet, blue, cyan & some green.)

[10] Hoya 85: Longpass filter which cuts in just below 400 nm with a sloped attenuation of violet & blue.

[11] Tiffen Hot Mirror: Transmits between approx 360-740 nm. Does not block all near UV nor near IR. (Thus rather useless it seems?)

 

 

Results

 

[2, 6, 8, 11] No Filter, B+W 415, Lexan and Tiffen Hot Mirror record similarly. Blue/cyan lint fluorescence is present. Teflon reflects violet/magenta.

 

[5, 7] B+W 415 + 489 and B+W 489 record similarly. Blue/cyan lint fluorescence is present. Teflon reflects violet/magenta.

 

[3,4,10] Hoya Y(K2), Hoya G(X0) and Hoya 85 record similarly. Some blue/cyan lint fluorescence is present; much is supressed. Teflon reflects red tones.

 

[9] B+W 023(OG530). Blue/cyan lint fluorescence is supressed. Teflon reflects red tones.

[DaveO]

Here are shots of my standard darkroom set up to show the effects on both a Spectralon target and a Colour Checker

 

Visible Light: Nikon D750 Full Spectrum Modification, Nikon Rayfact PF10545 MF-UV 105 mm f/4.5 lens, Metz 15 MS-1 flash, 1/200 s @ f/16 ISO 200, Baader UV/IR Cut Filter.

 

Ultraviolet Light: Nikon D750 Full Spectrum Modification, Nikon Rayfact PF10545 MF-UV 105 mm f/4.5 lens, Nissin Di866 Mark II flash, 1/200s @ f/16 ISO 200, Baader UV-Pass Filter.

 

Ultraviolet Induced Visible Fluorescence: Nikon D750 Full Spectrum Modification, Nikon Rayfact PF10545 MF-UV 105 mm f/4.5 lens with Baader UV/IR Cut Filter, Nichia NCSU033A UV-LED with Baader UV-Pass Filter, 10.0 s @ f/16 ISO 1250.

 

Dave

[Guest]

@Alaun: You have a good point I hadn't considered. Given the nature of UVIVF, the PTFE should not look white - rather, anything which does not fluoresce should have no color/illumination at all (ideally). I think I was just trying to apply standard practice (white balancing), though it is not appropriate in this application. My mistake - now I know better.

 

@Andrea: the PTFE I used here are virgin and new, so they shouldn't have any fluorescing contaminants. And for the racoon teeth - while bone should fluoresce anyway, these teeth/skull are coated with polyurethane to help protect/preserve it - and polyurethane itself also fluoresces (a kind of pale, greenish blue).

 

@DaveO: It looks like that Baader UV/IR cut filter is bleeding through just a bit of deep blue.

 

Now, I re-ran my test, using a filter combination that seems to work surprisingly well (at least, for non-research grade hardware). I used the same flash, flash filter, and camera, with a few added target materials as well. Not only did I get decent results, but I think I can now explain what I saw.

 

I know this isn't by any means perfect, but while the excitation/emission filters I'm using are certainly not matched, I think this will serve well to get some colorful UVIVF images relatively free from excitation bleed/contamination. Actually, with the FL-02 glass its not strictly UV (more like, UV+deep blue/violet), but I can live with that for now.

 

As for cutting the blue fluorescence from the ever present lint - well, I do have a couple new cut-on filters ordered which may help with that. If not, I'm just going to go with Andrea's advice... get a can of compressed air and try to blow the stuff away!

 

- Mark

[Cadmium]

---> I think the O530 did the best job of supressing the blue/cyan fluorescence.

 

---> [9] B+W 023 (O530): Longpass filter which cuts in at 530 nm. (Blocks UV, violet, blue, cyan & some green.)

 

---> [9] B+W 023(O530). Blue/cyan lint fluorescence is supressed. Teflon reflects red tones.

 

Andrea, each instance of "O530" should be Schott OG530

[Cadmium]

Here are shots of my standard darkroom set up to show the effects on both a Spectralon target and a Colour Checker

 

Ultraviolet Induced Visible Fluorescence: Nikon D750 Full Spectrum Modification, Nikon Rayfact PF10545 MF-UV 105 mm f/4.5 lens with Baader UV/IR Cut Filter, Nichia NCSU033A UV-LED with Baader UV-Pass Filter, 10.0 s @ f/16 ISO 1250.

 

Dave

 

Dave, Yes, my UVIVF test the other night of PTFE looked the same color as your test above shows, a kind of blue/gray.

Thanks for the comparison.

[JCDowdy]

Looks like things cleared up nicely once you created an adequate excitation/emission barrier. Well done!

 

Curious about the silicone though, it seems to bright to just be illuminated by adherent fluorescent dust. Problem with dust is you can't simply filter out its UVIF without losing UVIF you want to see. It is hard to imagine how much dust is everywhere until you start doing this.

 

The non fluorescent Viton and EPDM may also be useful as specimen supports. I have collected and tested various black plastic items from household waste and broken toys. Some have a filmy hazy fluorescence that wont clean off but others look totally black under UVIF.

[Guest]

@JCDowdy: I can't seem to find information anywhere online as to whether or not silicone fluoresces (under UV/blue irradiation). I don't think it should, and from what I saw in the images I took above my guess is that the brightness of silicone target is due to three things:

 

1) Silicone is particularly tacky, even compared to the Viton and EPDM targets, so dust/lint tends to stick to it signficantly more (and tends to stay put once stuck). For this reason, there is more dust/lint on the silicone target's surface, providing more localized fluorescent/secondary illumination on the surface of that target.

 

2) This silicone target is translucent, which effectively diffuses this fluorescent/secondary illumination from the dust/lint on its surface (allowing for the light to cover a larger area, vs simply remaining tiny hotspots of light).

 

3) The silicone target in the image is backed by a polished stainless steel disk; which will help reflect the fluorescent/secondary light back out into the silicone, thereby diffusing even more of that light (versus that light being lost/absorbed to the background).

 

4) Any blue light bleed at all from the imperfect match between my excitation/emission filters will also diffuse effectively in the silicone material, adding to the overall effect.

 

For now that 1/2/3/4 combo is my guess - at least until I can find info otherwise on whether silicone fluoresces. I'll keep looking.

[rfcurry]

Mark,

It appears to me that the silicone is fluorescing. Do you know the exact chemical structure of the silicone? The term silicone is used for any polymers with repeating units of siloxane to which materials are added to provide the properties, color, texture, etc., desired. One of the additives may fluoresce.

[DaveO]

Just to complete the set, here's the shot without the Baader U over the UV-LED

 

and the corresponding SS spoon shots, first without the Baader U

 

then the Baader U over the UV-LED

[Cadmium]

Dave, Yes, very nice tests. Thank you.

[Andrea B.]

Mark, maybe wash the lint off that Silicon ring, pat dry with a microfiber towel and reshoot? The lint cannot light up that ring by as much as is seen here!

 

Something to note about "leakage". If a UV-pass filter leaks around 400nm, it is transmitting violet, not blue.

 

So, next please note that spectral violet is not easily recorded with many cameras. Our eyes are stimulated in both our Red and in our Blue channels by spectral violet [380-450 nm, usually]. The Bayer pigments are supposed to achieve a similar overlap for spectral violet. But some cameras' internal filters block too much around 400nm and low violet is not recorded at all. Other cameras record spectral violet as blue because of their Bayer pigments or because of the way their internal UV/IR blocker passes the violet. Of course, any camera not fully blocking IR will also see some contamination of spectral violet. And finally, colour profiling is often needed to properly bring out spectral violet in a photograph.

 

[i've lost my reference link to this violet problem.]

 

Another point to remember: traditional scientific or industrial fluorescence filtration is very narrow-band on both the illumination and lens. Our photographic experiments with UV-induced fluor are very wide-band. How much does this affect our photos? Well, we do not know because we don't test with narrow band filters. No reason why we should do that because our purposes are photographic not scientific. And the expense would be huge.

 

[if I have repeated something for the nth time, please excuse me! I try to keep certain facts in the foreground when fluorescence discussions arise because I'm never sure who knows what. :D]

[Cadmium]

Something to note about "leakage". If a UV-pass filter leaks around 400nm, it is transmitting violet, not blue.

 

That would depend on the UV bandpass filter. If it is U-340 or UG11 then it cuts off at or below 400nm, but if it is something more like U-325c (or U-330, UG5...) then you have not only violet but also blue and green, depending on the light source output. Even U-360 or UG1 may transmit a little above 400nm mixing with the actual fluorescence.

The narrowest light source I have is the MTE Nichia 365nm, but other lights (like a compact fluorescent black light) haves wider band widths.

In short, it is best to limit the light to under 400nm, and limit what the camera sees to 420nmm and above, and if the light also has some red/IR light, then limit the visual band also.

My typical lens stack is BG38/BG40+GG420. You can also use Wratten 2E gel filter instead of GG420.

[Andrea B.]

I should have been more specific. :D

I was referring to filters like BaaderU and AndreaU and U-360-Stacked. I think it is possible that we are getting bits of violet from those. But the demarcation between ultraviolet and violet is not fixed. I don't really worry about a little violet "leakage" (i.e. small amounts of violet transmission).

[Cadmium]

It depends on how accurate you want the visually fluorescing colors to be.

I would not use any of the filters you mentioned to filter an MTE, including a Baader U, it is much more precise, convenient, and less expensive to filter the light source using a U-340 2mm thick filter on the MTE, and this cuts the UV to below 400nm (see graph). Also, U-340 can be easily obtained in other sizes, shapes, thicknesses, to attach to larger light fixtures and flashes (for example).

 

I think the 400nm/420nm is the ideal cut off and separation, however, it may not matter so much where the cut-off point is,

as long as there is a separation between UV/Visual threshold, so that none of the light from the light source is part of the reflected visual light entering the lens/sensor.

The only light entering the lens should be visual range fluoresce.

Some visually fluorescing light may illuminate and reflect visually off other surfaces, but no reflected excitation light should enter the lens/sensor.

No part of the light source should be present visually to the camera lens/sensor.

 

I have seen some people use '400nm' LED light to excite fluorescence. I have not experimented with this, but I have seen examples of it.

That situation might require moving the separation up a little higher using filters with different cut-off points, but there still need to be a separation.

 

[Andrea B.]

Well, as mentioned, none of this is easily accomplished with broadband filters. Should anyone ever be interested, there are narrowband filters made especially for fluorescence work.