Jump to content
UltravioletPhotography

UV Induced Infrared Fluorescence 2


Recommended Posts

In another post I mentioned the idea of doing UV-Induced Infrared Fluorescence in full colour. The images below show some initial results and conclusions.

 

I'm getting some rock samples to see if these show anything interesting. And I'll also see if I can do soime Visible-induced IR Fluorescence.

 

In the first set of images, highlighter pens were used to write on microscope slides which were then laid on PTFE. The PTFE is a useful test - it reflects well across the UVA, Vis, and NIR ranges, but it doesn't fluoresce. So if it comes out black it means the image really is the result of fluorecence rather than any pollution from reflected light. The results indicate that both the Nemo + Baader U and flashguns with a ZWB1 are good for UVIVF. However, for UVIIRF the Nemo + Baader is fine, but the flash + ZWB1 was useless.

 

Also intresting here is that the Blue highlighter (tried 2 different brands) does not fluoresce in the visible, but it does fluoresce in the IR. The IR image dos not show much colour, indicating the fluorescence is occurring acoss the whole of the NIR range.

 

Straight reflected visible light image:]

 

post-245-0-02288500-1606474498.jpg

 

UVIVF using Nemo + Baader:

 

post-245-0-78941400-1606474526.jpg

 

UVIVF using Flash + ZWB1:

 

post-245-0-06090500-1606474552.jpg

 

UVIIRF using Nemo + Baader:

 

post-245-0-20213600-1606474581.jpg

 

Now a couple of groups of flowers. It's interesting how similar the UVIVF images are to the visible one (yet the highlighter test indictates that this really is fluorescence) and how different they are to each other. It's also noticeable how the areas of foliage that fluoresce red in UVIVF fluoresce very brightly in UVIIRF: I assume that this is a broad-spectrum fluorescence that starts in the red and continues uninterrupted across the NIR.

 

In this first set of images, the two UVIVF images were WB-ed against the white flower at top-left:

 

Straight Visible image:

 

post-245-0-57694100-1606474805.jpg

 

UVIVF using Nemo + Baader:

 

post-245-0-36147000-1606474848.jpg

 

UVIVF using Flash + ZWB1:

 

post-245-0-76458000-1606474887.jpg

 

UVIIRF using Nemo + Baader:

 

post-245-0-83381900-1606474919.jpg

 

In the following set, the UVIVF images were WB-ed against the grey background at the right hand side.

 

Straight visible image:

 

post-245-0-51312100-1606474997.jpg

 

UVIVF using Nemo + Baader:

 

post-245-0-61666400-1606475026.jpg

 

UVIVF using Flash + ZWB1:

 

post-245-0-95882000-1606475060.jpg

 

UVIIRF using Nemo + Baader:

 

post-245-0-12554200-1606475097.jpg

Link to comment

Bernard, nice images.

 

The highlighters are interesting, and for a reason: you may try to combine them into a white IR fluorescence target for WB purposes. Interesting how the blue one fluoresces in IR only.

Link to comment

If you are going to use flash for any fluorescence, visual or IR, then you really should include S8612 with the U-340, other wise you will be getting a lot of IR with your UVIIRF, and some far-red/IR with your UVIVF shots.

The U-340 will not block far-red or IR.

Furthermore, you should use an IR filter on your lens, to cutoff and isolate what IR range you want to show.

Link to comment

Thanks, Cadmium. The prolem with putting S8612 over the lens is cost - a 5x5cm square (even if available) for each of 4 flashguns would cost more than I'm prepared to spend.

 

I sometimes put an S8612 over the camera for UVIVF - this certainly gives different colours, but I'm not sure it is stopping any reflected light contamination. But I'll do some tests.

 

Re. "you should use an IR filter on your lens, to cutoff and isolate what IR range you want to show" - that's essentially what the tri-colour technique is doing, although it's not yet shown much difference across the NIR range.

 

PS - just ordered a couple of your 20.5 mm U340 filters intended for the Convoy torch: one of these will be combined with my 20 mm 303BP10 filter to provide additional IR blocking to see if I can get a daylight image using the UVFS lens withour IR contamination.

Link to comment
Bernard, Cadmium is right, you need to stop the IR from leaving the flash or it will definitely contaminate the scene for infrared fluorescence. S8612 isn’t the only glass you could use for that.
Link to comment

Bernard, Cadmium is right, you need to stop the IR from leaving the flash or it will definitely contaminate the scene for infrared fluorescence. S8612 isn’t the only glass you could use for that.

 

At present I don't use flash for UVIIRF, only for UVIVF. For UVIIRF I'm using a Nemo + Baader.

Link to comment
But what is stopping your UVIVF from having IR contamination then? I am honestly very confused about your multiple setups at this point. I’m no longer sure I understand which filters and which light sources were used for what.
Link to comment

But what is stopping your UVIVF from having IR contamination then? I am honestly very confused about your multiple setups at this point. I’m no longer sure I understand which filters and which light sources were used for what.

 

The filter/lighting combinations that work are those against the highlighter test images in the first post above, i.e.

 

Nemo + Baader for UVIVF

Flash + ZWB1 for UVIVF

Nemo + Baader for UVIIRF

 

In fact, in tests I have just done, adding the Baader to the Nemo doesn't make any difference, so I'll probably just use the naked Nemo with it's in-built filter.

 

For UVIVF using Flash + ZWB1, I'm using a non-modified camera, so the IR sensitivity is already low. However, I can see a very small amount of leakage, which can be addressed by putting an S8612 on the camera. But the leakage is small, and as the S8612 seriously dampens red fluorescence, so I'm in two minds as to whether to use it all the time. Here are the comparison shots, both at the same exposure:

 

No S8612 on the camera. In the full-resolution image you can see a faint magenta on the PTFE background:

 

post-245-0-33598500-1606604677.jpg

 

With S8612 on the camera:

 

post-245-0-19831100-1606604691.jpg

Link to comment

On the camera for UVIVF you need a Baader UV/IR cut maybe? Also for that matter, what is stopping UV from getting into the camera? Even unmodified cameras will get some UV, I think?

 

To be clear I think the bulk of it is definitely visible light fluorescence but I am sure if the specimen is dim enough you will get contamination if you don’t fix it ahead of time. It’s the fluorescence version of “you can always force some OOB contamination through any filter” that we know so well.

Link to comment

S8612, not on the camera, on the flash.

I am going to say U-340 instead of ZWB1 (for anyone who is confused by that), and by the way, the Nemo has ZWB2 (I think) which is more like U-360/UG1 than it is like U-340/UG11.

U-340 is best because it cuts below 400nm.

For the Nemo or Convoy S2+ you only need the U-340 filter, because the LED's are fairly narrow band, and don't have much visual or IR emission above the visual violet/blue residue.

The U-340 will block anything from escaping the torch above 400nm (see example below).

However, U-340 will not block everything coming from the torch above 400nm with a flash.

Flashes emit a huge amount of visual and IR, and all of that needs to be blocked, that is why you stack the S8612 with the U-340, then you limit the flash to UV only under 400nm.

S8612 has no equivalent.

Some other BG glass can block Far-Red/IR down to the same cut off, but only when it is much thicker, BG40, for example, would require 4.5mm thick to equal the 2mm of S8612.

Keep in mind the usual glass sheet comes in 4mm thick, anything thicker would be specialized, or would need to be built up from more than one sheet thickness by gluing it.

BG40 is second best to S8612. See graph.

Then when you do IR fluorescence, either induced by UV or induced by 'visual' (usually best to use green visual light for that), you need to use an IR longpass filter on your camera lens,

otherwise you will be recording a mix of visual and IR fluorescence, and IR fluorescence is very weak and needs much longer exposure time than does visual fluorescence.

 

So here is the basic set up for visual fluorescence:

UV torch filtered with U-340 2mm.

Barrier filter on the camera lens to block any reflected UV from entering the camera. Wratten 2E works well for this (or other filters that cut at about 415 or above with NO footprint close to 400nm or below).

Or if you want to use a flash,

Full spectrum converted flash filtered with U-340 2mm + S8612 2mm. Again, barrier filter on the camera lens...

 

For IR fluorescence:

So here is the basic set up for visual fluorescence:

UV torch filtered with U-340 2mm.

IR longpass filter on the camera lens to block visual and define IR to any point you want (keep pin mind that a 720nm filter, for example, shows a mix of Visual and IR fluorescence).

Or if you want to use a flash,

Full spectrum converted flash filter it with U-340 2mm + S8612 2mm. Again, IR longpass filter on the camera lens...

 

One more thought, if you are doing UVIIRF, then you may want to put a Zeiss T* filter in front of your IR longpass filter, because even IR longpass filters can fluoresce from UV light.

 

Get those ducks lined up, and you got it.

Anything else? maybe I am forgetting something.

Oops, I think I forgot something, do you know what it is?

With UVIVF (visual fluorescence) you need to block any fluorescence above the visual range from entering the lens,

so you need to use some kind of visual range filter, stacked with your 2E (or other UV barrier filter), like BG38, BG40, Baader UV/IR-Cut, etc.,

whatever filter you would normally use for shooting visual photos with your full spectrum camera.

 

You don't need to worry about that with a stock camera or with IR fluorescence (unless you want to get fancy and do some kind of IRF bandpass shot.

Of course, you can't do IRF with a stock camera.

 

I probably still forgot something?

 

Another thought, you will see no color from IRF unless you are using a lower longpass filter or no no longpass (IR longpass) filter at all (AKA 'false color', as they say).

 

Here are some mixed examples. NOTE: "LUM" a word I have sometimes use for UVIIRF (credit Ninjin).

Also note the interesting difference between using UV to excite the 850nm IRF and using visual (green) light to excite the 850nm IRF. Yep! :wink: I rather like the visual green induced version.

post-87-0-78259500-1606613429.jpg

 

S8612 vs BG40 thickness equivalence:

post-87-0-20466500-1606614853.jpg

 

U-340 2mm visual violet leak block. Filter lest through invisible UV to excite fluorescence, but no visual violet/blue escapes the torch filter.

post-87-0-41395800-1606615063.jpg

 

IRF Cadmium

post-87-0-67454200-1606615379.jpg

Link to comment

Steve, well done, you really clarified things.

 


I would say that a 720 nm longpass filter passes IR-only by definition, even if you can see something. The human eye has no sharp cut-offs and you need to put a limit somewhere. Pointing my 532 nm green laser at the wall and putting my Hoya R72 in front of it, I saw a very faint red crescent moon shape. It was the 808 nm "pump", which is not collimated as well as the 532 nm green, thus the crescent moon. It was so faint I had to use my full-spectrum camera to find it, but once found, I saw it. That doesn't mean 808 nm is red, it is definitely IR. But this is a whole new topic, and I won't discuss it here.

 


Seeing Andy's work at subtracting contamination in UVIVF images, I think the same could be done in UVIIRF. The contamination image can be obtained by putting a longpass filter (better deep red than R72) on the flash. I don't say this could solve the problem definitely, but may be a temporary solution.

Link to comment

LED's are weird, those kind of tests are misleading, LED's can have very intense light that cuts though high OD. Not a good test. Not real life.

Pointing a laser directly at a camera/filter is not what we are coming close to here.

 

I think you are better off using a higher IR filter.

 

I can see red with even an 830nm filter if I use an incandescent bulb... and even very weak transmission can be enough to compete with very weak IR fluorescence.

So to define exactly what your camera is recording, use an IR longpass that is high enough to cut off that you don't want your camera recording.

Link to comment
Ionic longpass filters usually have very high ODs. They are very good at blocking shorter wavelengths. If you try to see a powerful blue LED through an Hoya R72, you will see red. That's because, like UV LEDs, blue LEDs fluoresce red and emit a tiny bit of red light. I don't think an LED can punch through something like OD 10. Looking at my LEDs of various wavelengths with my camera + R72 + in-camera white balance, I saw only faint false-yellows, indicating short IR wavelengths between ~700 and ~800 nm. And those are bright LEDs, they heat up things.
Link to comment

I want to keep it simple, so random people who are reading all this can understand.

My post sounds terribly complex to me, hard enough for someone new to understand.

 

I may be wrong, but from what I have read Bernard say in this topic so far, it doesn't sound like he is doing UVIIRF exactly correctly.

 

I will say this, UVIVF is much easier to do than UVIIRF, if for no other reason than IR fluorescence requires much L O N G E R exposure times.

If the exposure time of UVIIRF is not long then you are probably not doing it right, you are probably capturing something other than IRF.

Link to comment

Oh, and by the way, the fact that I can see some faint red through 830nm 2mm doesn't mean I can see 830nm,

it means I can see the low transmission of 720nm (if you look at the 2mm version graph).

Use 3mm and that becomes 780nm. Still all that isn't part of what I was trying to outline above.

Link to comment

Cadmium summarized this well. For UV inducing fluorescence, you want a clean UV only light source. Then a narrow specific filter on your camera to support your claim.

But we are not all scientists here trying to prove that 385nm excitation of flower A leads to 850nm emmission on pollination day Q.

 

If it looks artistically good to you and you have a vision for that subject, than by all means hit it with UV and 5% IR to get a blue background in your subject with a Red pop fluorescence.

 

Sometimes we get too technical here and the other 200 people whom have created accounts here but never post anything, should know artistic freedom is good. Explore your color palette. Some subject have interesting colours mixed in with various leaks. Just don't claim anything, and your fine. If you want to make a claim to a specific subject doing something, than you need to be throw with all your controls for each subject, every single time you image it.

Link to comment

David, I mostly agree. Sometimes we tend to be too meticulous, too worried about leaks, colors, range, sharpness, white balance and so on. I like Bernard's images and the red kick given by IR leaking from the flash. Artistically, they are nice. There aren't really any rules, as long as someone states that a certain image is not "pure", or the processing used.

 

For example, I can't say than an image is UV-only if I know my filter has OD 3 for example and leaks a bit, I have to be transparent and say that, so that other people looking for pure UV can compare their images to mine and know what to expect. But as long as you are clear about what you did, you are fine, and a bit of freedom and experimenting with colors doesn't hurt sometimes.

 

In this case however, I think Bernard is looking for pure UVIIRF, and so the IR leak is not good. Then, defining where red ends and IR starts is a whole different mess, who says the boundary is at 700 nm, who says it is at 680 nm, 750 nm, I think I have even seen 800 nm. It is the same in UV. I can see 340 nm, it looks blue (and, BTW, I have seen other sources which say UV at around 350 nm looks blue, people have experimented down to 300 nm, and so this confirms that what I saw is actual 340 nm UV), but I don't consider it visible light. For me, visible is 400-700 nm, and by definition I consider light outside either UV or IR.

 

And about sleeping members, as UVP Admin calls them here, first post, bottom line, please post something (I don't mean to be harsh). Even if your images are not pretty, even if you are a complete beginner, you can learn, and most of all you can inspire others. I have been inspired by other members here, and I am always learning. Sometimes people here try things no one has yet tried, you may be one of them.

Link to comment
I like italic. I don't know why it is called that, it isn't "Italian", we call it "corsivo", because in hand writing the pen "corre" ("runs") on the paper. Mysteries.
Link to comment

Oops, I think I forgot something, do you know what it is?

With UVIVF (visual fluorescence) you need to block any fluorescence above the visual range from entering the lens,

so you need to use some kind of visual range filter, stacked with your 2E (or other UV barrier filter), like BG38, BG40, Baader UV/IR-Cut, etc.,

whatever filter you would normally use for shooting visual photos with your full spectrum camera.

( I will edit above and insert that).

 

You don't need to worry about that with a stock camera or with IR fluorescence (unless you want to get fancy and do some kind of IRF bandpass shot.

Of course, you can't do IRF with a stock camera.

Link to comment

Steve, I was wondering about your comment on the UVIVF;

 

"So here is the basic set up for visual fluorescence:

UV torch filtered with U-340 2mm.

Barrier filter on the camera lens to block any reflected UV from entering the camera. Wratten 2E works well for this (or other filters that cut at about 415 or above with NO footprint close to 400nm or below).

Or if you want to use a flash,

Full spectrum converted flash filtered with U-340 2mm + S8612 2mm. Again, barrier filter on the camera lens..."

 

As I see it, for the full spectrum converted flash you'd only need the S8612 on the flash if you were doing the imaging with a camera converted to multispectral (UV-Vis-IR) imaging by removal of the internal filters. I would have though that with a standard camera, with the internal filters in place, then the IR from the flash should be blocked, or is that not the case - would the S8612 still be needed on the flash with a standard camera?

 

Fully agree with you about the need for a blocking filter on the lens, and ideally a low fluorescence one. Even if the internal filters in the camera (again in an un-modified one) are good at blocking the UV, then some lenses can themselves fluoresce under UV. UV should be blocked before it gets to the lens for fluorescence work.

Link to comment

All of the blocking filters I mentioned on the lens are low fluorescence. Yes.

 

You need both U-340 and S8612 on the flash.

Your flash should only be emitting UV, no visual, and no IR.

Right?

 

You are saying you only need S8612 on the flash if you are using a full spectrum camera because it somehow replaces the internal filter that was removed?

If anything, you could use the flash with U-340 only, and no S8612, but that will have more far red from the U-340 700nm+/- bump.

Pylon experimented with that idea years back:

https://www.ultravioletphotography.com/content/index.php/topic/1463-365nm-uv-flash/page__view__findpost__p__9952

 

Is that what you mean?

Perhaps I don't understand what you mean?

Link to comment

If you only put S8612 on the full spectrum flash, then no mater if you use full spectrum or stock camera, you will get massive amounts of visual light, you absolutely don't want that when you are doing fluorescence.

You ONLY want UV to excite the target, no visual light, no IR. You ONLY want UV coming out of the flash, stock or full spectrum camera.

Right?!

I may not understand what you are saying.

Link to comment

OK, maybe you mean that we don't need to put the BG38, BG40, UV/IR-Cut filter on the camera lens with the 2E(or other) to block IR?

That would be a good point, if that is what you meant?

 

Do you need the S8612 on the flash if you use a stock camera?

I think so, because the IR from a full spectrum flash is so strong...

You can always test that idea.

Link to comment

I think with a full spectrum camera, when doing UVIVF, you want to filter the lens how you would for regular visual shots, and ADD the 2E type UV blocking filter.

You also want IR blocked from exiting the flash. You don't want Far-Red/IR from the flash mixing with red fluorescence, that would be true of both stock and full spectrum. Right?

So again, you only want UV exiting the flash. You only want to excite, not to illuminate the visual spectrum.

Right??

Am I missing something here?

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...