UV-Induced Infrared Fluorescence Investigation

[Andrea B.]

ADDED 2 YEARS LATER: The bricks were fluorescing. Duh!!! I can be so stoopid!

 

 

Mark was suggesting an experiment to determine more about the validity of recent UV-Induced Infrared Fluorescence photographs. I had my gear still set up so thought I would run a quick-n-dirty test.

 

In the dark, illuminating with a 365nm Nichia-chip UV-LED, I photographed a Labsphere 99% reflective white standard (made from Spectralon) under 4 different IR-pass filters on the Zeiss UV-Planar 60/4 lens. During the first run, the UV-Led torch was not filtered. For the second run I added a 340nm (1mm) UV-pass filter to the torch.

 

The Spectralon white standard does not fluoresce and reflects approximately 99% of all incident UV, Visible and Infrared light. There is some minor variation across the spectrum, but nothing to worry about.

 

IR-Pass Filters: RG9, 715, 780, 830, all which are 2mm thick.

The RG9 definitely fluoresces when the UV-Led is directly shined on it. I did not see any fluorescence by the 780 or the 830. The 715 I was not sure about. It might have had a very weak fluor? As you will see the photographs seem to indicate that.

 

NOTE: These photos were converted only. Please do not draw any conclusions from the false colours which come from a particular white balance preset on my D610. The 3rd photo in each group is fairly close to what I was actually seeing. But the 1st and 2nd photos in each group are not.

 

White Standard under Unfiltered Black Light with IR-Pass Filters on Lens

• All exposures were 15" long.
  

• When I look at the white standard under the unfiltered UV-Led torch, it appears dark violet. I do not know whether I am seeing high UV (like maybe 390-400 nm) or some actual violet (400-410nm). We have all speculated that our 365nm torches leak a bit of violet.
  

• I do not know for a fact what is going on with the 1st (RG9) and 2nd (715) photos. The RG9 photo may be showing some fluorescence (as noted above)? Maybe that is what is happening with the 715 photo also? Something is happening with these two filters! All comments are welcomed.
  

• The 3rd photo made with the 780 seems to show a bit of reflected ultraviolet or violet which registered in the blue channel. EDIT: The reflected light in the 780 photos does not register only in the blue channel. I would not be surprised by a little bit of violet leak in a 365nm torch, but why is that getting through the 780 filter? Or, if it is high UV, why is that getting through this filter?
  

• The 4th photo made with the 830 is black, as expected. Thank goodness!!!!
  

 

 

 

 

 

White Standard under 340-1 Filtered Black Light with IR-Pass Filters on Lens

• Exposures were increased to 30" due to the filtered torch.
  

• Again the 1st (RG9) and 2nd (715) photos have a big problem. Is it due to fluorescence of the filter glass?
  

• The 3rd photo made with the 780 still shows evidence of some violet leak. The 340 nm filter in a 1mm thickness might not be quite strong enough to block all violet leak. So I will try to repeat this experiment with a thicker 340. EDIT: I am not sure violet leak is being seen.
  

• As before only the 4th photo made with the 830 is black, as expected.
  

 

 

 

 

 

Conclusion

At this point my tentative conclusion would be that for UV-Induced Infrared Fluorescence experiments it might be best to use the 780nm or 830nm filters. BUT, this experiment needs to be refined and repeated to make sure we are all getting the same conclusion across the various shooting platforms. So let us not be too hasty.

 

Another consideration is that some emitted fluorescence (chlorophyll, for example) is in the range 700-750. How would we capture that if we determine our 715 filter is fluorescing and giving false results in UV-induced infrared fluorescence photography?

 

 

Sub-Conclusion

Remove fluorescing lamp shades from the vicinity of such experiments. It is way too easy to get contamination in fluorescence shots, isn't it??

[Cadmium]

What is your conclusion?

 

Oh, and have a Happy Thanksgiving.

[Andrea B.]

Conclusion is there now. :)

 

Do I need to stack a UV-Pass filter over my IR-Pass filters to determine whether they can pass UV?

Rats. This is a messy thing. And it's late. I'll think about it tomorrow.

[Cadmium]

Oops, sorry, was the conclusion there before? I was just looking at pictures then.

Sorry, I am missing your point. What are you trying to exactly determine here?

 

"Do I need to stack a UV-Pass filter over my IR-Pass filters to determine whether they can pass UV?"

 

Stack a UV-Pass filter over an IR pass filter?

IR Pass filters don't pass below their cutoff point, they don't pass UV.

Don't go overboard with this idea and start photographing your Nichia through IR pass filters, because you WILL see some 'light' through most of those if viewed directly, even a little with RG1000 2mm,

even if you stack it with S8612 3.5mm, and even with U-340 2mm on the UV torch, this is because these LED's are extremely bright when viewed directly, they will burn through just about any OD, this is why directly viewing LED's as tests is not a very good real world test.

Back to your question, IR pass filters don't pass UV unless you nuke them in direct view with a Nichia LED, which is not reality, it is a drill press.

 

Your point about fluorescing IR filters:

Some do more than others, just like most of the longpass filter glass.

I don't see mine fluorescing when I am shooting photos, only when I point the UV light more directly at the filter.

The result of a strongly fluorescing filter would be fogginess, I think?

Remember, even most lens glass fluoresces some, if bathed directly with UV light.

If you really want to test that to the max, then try a Wratten IR range gel filter, those fluoresce the least of anything other than KV filters, and the highest KV is 550nm.

So if you are concerned about your IR lens filters fluorescing and affecting the results, then get a gel filter of the same nm, and compare results.

 

I still feel like I am missing your point in these tests somehow?

 

Have some fun first. Try some green induced IR fluorescence, use a 715nm or higher filters, and see if you like it.

[Cadmium]

Also, try putting your Cadmium test tile in the photos with your Spectralon. I think that would be a better test for IR fluorescence.

[Andrea B.]

Yep, going to perform more experiments, for sure! And I thank you for the suggestions.

 

I still feel like I am missing your point in these tests somehow?

 

We need to demonstrate that we are seeing UV-induced IR-fluorescence when illuminating appropriate subjects with a UV-Led using an IR-pass filter on the lens. This test, using a NON-fluorescent subject, captured some light which is as yet unexplained. WHAT is it that we are seeing in the RG9 and 715 photos?

 

ADDED 2 YEARS LATER: The bricks fluoresce.

 

[Edit: I restored some accidentally deleted phrases.]

[Andrea B.]

Remember, even most lens glass fluoresces some, if bathed directly with UV light.

 

A dedicated UV lens will not fluoresce. The UV-Planar, UV-Nikkor and UAT, any of which I use in most tests, do not fluoresce.

[Cadmium]

So this is to see if the PTFE shows any light? It looks pretty much the same in UVIVF. Dull gray/purple.

Try some green light and the tile.

 

I was not implying that your lens is fluorescing.

The light you see on the PTFE is not coming from the filter fluorescing.

Try some gel filters.

 

I think you would be better starting with green/visually induced IR florescence, higher nm, 780nm/830nmm/...

Then go back to UV induced.

[Andrea B.]

The point is that if we claim we are recording UV-induced Infrared Fluorescence then we need to rule out contamination for our given UV-Led illumination and given lens filtration.

 

I've just shown that when using my 365nm UV-Led illumination and my IR-pass filters (RG9, 715, 780) on the lens, then there is some kind of light being recorded that does not come from fluorescence of the subject, a Spectralon disk, because a Spectralon disk does not fluoresce. That light will also be present when shooting fluorescent subjects under the same illumination and with the same filtration. So the question is: Is this light a contaminant? And, where is it coming from?

 

What am I missing here?

Shooting in the dark, if some light is passing through an IR-pass filter, then it must be IR. But I am only illuminating with UV. If the IR-pass filters block UV and Visible light, then what is this mystery light I've recorded? It could be that high red or low IR around 700nm (say, 695-705nm). But why would my Nichia 354nm UV-Led emit any red light? Is there some red light coming from other parts of the torch? Even when I filter the torch I'm getting this red light.

 

ADDED 2 YEARS LATER: The bricks fluoresced.

[Cadmium]

What does your Spectralon disk look like when you put it in a UVIVF scene?

When I photograph PTFE in UVIVF, it looks much like your Spectralon UVIIRF tests. Kind of a dull dark dirty gray/purple (white balance variable).

Is my PTFE fluorescing or not? It isn't reflecting the UV, because I have it filtered.

Is your Spectralon fluorescing or not?

 

You're not missing anything.

I think.... the light you are seeing is fluorescing from the materials in the scene.

Try filtering your LED with U-340 2mm + S8612 3.5mm, just to rule out the most you can. Use RG780+ to rule out lower visual, and that should exclude everything.

If the only 'lighting' is 365nm, and no visual, no IR, and you are shooting 780nm...850nmm+, then what are you seeing?

 

Try a scene with all three, Spectralon, Cadmium tile, and Color Checker. That would be an interesting comparison.

Try it in both UVIVF, and UVIIRF to compare both of those.

 

Somewhat unrelated, but then again, somewhat related. Found this while trying to find my pic of PTFE in UVIVF, but didn't find that yet.

[Cadmium]

I didn't find the UVIVF PTFE test I was looking for, but this will do fine, and even shows black rubber sheet also.

The background under the gourds is PTFE sheet. Note how it looks dull gray/purple.

The black area at the left is black rubber sheet.

Is the PTFE fluorescing? Yes, it is. Everything you see in that photo is fluorescence, not reflected.

Many things, in fact most things fluoresce a small amount in UVIVF. It is even hard to find things that are completely black in UVIVF.

The black rubber is pretty black, and that Krylon 1602 ultra flat black spray paint that Shane has mentioned before is black.

http://www.filmtools...pray-paint.html

 

[Cadmium]

Also curious what the background material is that you have your target sitting on. It is also showing light.

[Andrea B.]

Spectralon can transmit if thin enough. Spectralon reflects at about 98-99%. Spectralon does not fluoresce. It cannot absorb photons and then emit them. Similarly, PTFE.

 

 

The background is bricks.

 

 

If the only 'lighting' is 365nm, and no visual, no IR, and you are shooting 780nm...850nmm+, then what are you seeing?

That is exactly the question I have been trying to ask.

 

ADDED 2 YEARS LATER: The bricks were fluorescing.

[Cadmium]

You have shot your Spectralon in visible range fluorescence and it looks black?

[Cadmium]

"Spectralon does not fluoresce. It cannot absorb photons and then emit them. Similarly, PTFE."

So then PTFE looks black in UVIVF?

 

http://www.ultravioletphotography.com/content/index.php/topic/1929-ptfe-in-uvivf/page__hl__%2Bptfe+%2Bfluorescence__fromsearch__1

[Cadmium]

If you put U-340 2mm on your Nichia LED torch, you will only have 365nm light. If you want to further convince your self, then stack S8612 3.5mm on the LED with the U-340 2mm.

If you don't have U-340 2mm, you could use UG11 2mm.

If you do all that, you will for sure only be illuminating your target with UV.

 

I guarantee you, once you plop that test tile down in front of your UVIIRF shot... you will go "wow". :)

[Andrea B.]

If you read to the end of the link you provided, you will see that Yes, PTFE is black in UVIVF when any overlap between excitation wavelength and emission wavelength is removed. That is, you don’t want your illumination to leak into your lens. Mark eventually accomplished that and showed the black PTFE.

 

In my experiment above I can’t figure out what overlap I’m seeing between a UV torch and an IR-pass filter. That is a 300 nm gap. How can there be overlap? Yet something is wrong with those photos. The Spectralon is not black.

[Cadmium]

I understand. I was more looking at DaveO's examples, which look closest to what I view PTFE as in UVIVF, which look similar to yours above.

 

How does your black Spectralon look in a similar test as your above?

[Andrea B.]

Could it be our UV torches??? I think it must be the torches emitting light from some electrical component other than the Nichia Led chip or reflecting something reddish from some part of the torch interior. The torches must be emitting/reflecting some high red or low IR which naturally passes thru the 715 and 730 filters, but begins to have difficulty getting through the 780 or 830 filters.

 

Nothing else can explain the light & shadow effect seen above in the UV-Induced Infrared photos.

 

BUT.......why doesn't the 340 filtration prevent this?

 

This is sure a stumper.

 

I'm happy you and E. have observed the same results, and I look forward to your posts later.

[ulf]

When I measured the emission from a "365nm" "3W UV-LED of unknown manufacture but not Nichia, to my surprise it had quite a lot of light in the visual range and into short IR too.

If the semiconductor type Nichia is using is the same you'll get a similar emission pattern from your torch.

 

The resulting emission graph below is directly from the spectrometer and not calibrated for intensity.

The grating in my spectrometer give a sensitivity decreasing a lot in the IR range compared to UV.

Thus the emission of the LED in IR is higher than the graph show.

 

 

 

The U-340 has an IR leakage that partly overlap the LED's emission.

http://www.uqgoptics...tting_UV340.gif

[Cadmium]

http://www.ultraviol...&attach_id=3374

 

[Andrea B.]

Very interesting result, Ulf. Thank you for posting it. It is somewhat surprising to learn that some UV torches can emit such a large amount of visible light !! I do not think that my particular UV-torch does this.

 

---

 

Here are the results of directly shining my Nichia 365nm UV-Led torch through different filters and stacks in total darkness. I aimed it at a 99% Spectralon white rectangle. Note that most of us can see some UV light between 380-400nm, or at least between 390-400nm. It appears as a bright whitish-violet colour.

 

Schott 340nm (4mm) UV-Pass -- see UV light (as described above)

Schott 340nm (1mm) UV-Pass -- see UV light (")

BaaderU UV-Pass -- see UV light (")

I don't quite know how to explain it, but I can see the UV light being reflected off the Spectralon (back into my eyes I suppose, ouch) but the disc is black (black-ish) because it is non-fluorescent.

 

Baader UV/IR-Cut -- see some reflected violet light, not at all bright

 

Schott 340nm (4mm) + Baader UV/IR-Cut -- no light seen, so 340 cuts violet.

Schott 340nm (1mm) + Baader UV/IR-Cut -- no light seen

BaaderU + Baader UV/IR-Cut -- no light seen, so BaaderU cuts violet.

 

Schott 715nm (2mm) Red+IR-Pass -- no light seen

ADDED: This is important --> Placing this 715 filter over my torch and looking directly at the torch thru the filter I can see some very dull red. So my tentative conclusion is that the 715 must be passing enough UV for it to become apparent in the long-exposure photographs made above. How can I further test this?

Schott 780nm (2mm) IR-Pass -- no light seen.

Schott RG (2mm) IR-Bandpass -- no light seen

 

So, having just made these little tests, I am more baffled than ever about what I managed to record in the photographs above. ADDED: Crossed out that remark as I may have gotten to the bottom of this little temporarily-baffling mystery. We shall see.

 

 

EDIT: Added results for thinner 340 filter.

EDIT: Added new observation about the 715nm (2mm) IR-Pass Filter

[Andrea B.]

NEXT UP: Try various solutions. Perhaps stack a longpass filter over the 715 IR-pass? Perhaps use a thicker 715 IR-pass?

 

In ordinary sunlight photography, a small UV leak through a 715 filter would not contaminate the IR because there is so very little UV in sunlight relative to the 50-55% IR content in sunlight. But in the dark, when there is no ambient IR at all and only some IR-fluorescence, the amount of UV from the UV-Led torch may be (relatively) overwhelming and can potentially contaminate the UV-induced Infrared photograph.

[Cadmium]

> Here are the results of directly shining my Nichia 365nm UV-Led torch through different filters and stacks in total darkness.

> I aimed it at a 99% Spectralon white rectangle. Note that most of us can see some UV light between 380-400nm, or at least between 390-400nm.

> It appears as a bright whitish-violet colour.

 

Let's break this down.

 

> Schott 340nm (4mm) UV-Pass -- see UV light (as described above)

> Schott 340nm (1mm) UV-Pass -- see UV light (")

> BaaderU UV-Pass -- see UV light (")

> I don't quite know how to explain it, but I can see the UV light being reflected off the Spectralon (back into my eyes I suppose, ouch) but the disc is black (black-ish) because it is non-fluorescent.

 

If what you are seeing is reflected UV light... then would this reflected UV light be blocked to your eyes if you looked at it through a GG435 filter?

 

I am not going to argue one way or the other if your target fluoresces or not. However, everything else will fluoresce, if even in some small amount, when the Nichia 365nm LED is filtered with these first three examples. These three are all working normal.

They transmit UV, they block visible, and things fluoresce.

 

> Baader UV/IR-Cut -- see some reflected violet light, not at all bright

This is also normal, you are seeing the violet and blue from the Nichia, because the Baader UV/IR-Cut transmits a little below 400nm, so you are seeing a tiny bit of light that is a little below 400nm and light that is blue above 400nm.

 

> Schott 340nm (4mm) + Baader UV/IR-Cut -- no light seen, so 340 cuts violet.

> Schott 340nm (1mm) + Baader UV/IR-Cut -- no light seen

> BaaderU + Baader UV/IR-Cut -- no light seen, so BaaderU cuts violet.

 

These stacks are all working normally, they block visible and they block UV when stacked together. Nothing fluoresces to the eye, and everything is black.

 

 

I am confused by these below, are all of these with the filters on the torch and you are looking directly into the torch? With just your eyes (goggles), or using your full spectrum camera in live view?

Even with eye protection, I would not recommend this test, only do this sort of thing with a Nichia with your camera in live view please.

These are all normal, as I said before:

"Don't go overboard with this idea and start photographing your Nichia through IR pass filters, because you WILL see some 'light' through most of those if viewed directly, even a little with RG1000 2mm,

even if you stack it with S8612 3.5mm, and even with U-340 2mm on the UV torch, this is because these LED's are extremely bright when viewed directly, they will burn through just about any OD, this is why directly viewing LED's as tests is not a very good real world test."

 

 

> Schott 715nm (2mm) Red+IR-Pass -- no light seen

> ADDED: This is important --> Placing this 715 filter over my torch and looking directly at the torch thru the filter I can see some very dull red. So my tentative conclusion is that the 715 must be passing enough UV for it to become apparent in the long-exposure photographs > made above. How can I further test this?

> Schott 780nm (2mm) IR-Pass -- no light seen.

> Schott RG (2mm) IR-Bandpass -- no light seen

 

So at this point, all of these tests are normal, other than you don't think the target should fluoresce at all, therefore you are saying that the light you see on the target is reflected visual light.

[Cadmium]

So you are thinking that the 365nm UV only light is illuminating your target, and being reflected back through your 715nmm IR filter?

This I doubt, simply because when viewed directly the Nichia LED is very intense and can be seen through almost anything.

For example, if you use your same stacks that you mentioned above, that block both UV and visible, and shine your MTE though them,

then everything is black nothing fluoresces, but if you view that directly with your full spectrum live view (not just goggles please!) you will see bright light.

 

Here is another test I just now did:

Nichia 303 and Convoy S2+ (not simultaneously), both filtered with U-340 2mm. Directly viewed with full spectrum live view.

First with RG715 2mm stacked on torches, then with 4mm, swapped 2mm then 4mm, then 2mm, then 4mm... by hand in live view.

Nichia can be seen through both 2mm and 4mm, and amazingly there is no difference in brightness.

 

Additionally, I performed the same test using RG1000 instead of the RG715.

First using 2mm then 3.75mm, swapping those also shows no difference in brightness between the two. However RG1000 is dimmer then RG715.