Jump to content
UltravioletPhotography

My first attempt at UV-induced visible fluorescence: Helianthus cultivar


igoriginal

Recommended Posts

So this is my first "official" (serious) attempt at UVIVF (uv-induced visible fluorescence) photography.

 

That of a Helianthus ("Sunflower") cultivar from my flower garden.

 

Done at night, in a pitch-black room (with all windows blocked out from ambient street lights and passing cars). Also, a towel pushed under the door frame, too.

 

Ceiling fan turned off, and all AC vents closed, too (to prevent any potential air movement in the room from causing image blur of the flower, during the long-exposure photo).

 

The UV torch which I used was some cheapo generic design. Probably not very deep UV-A emission (I think, maybe around 385 - 390 nanometers bandwidth). But, I plan to get a much better and more powerful UV torch, soon, which emits UV radiation down to 365-370 nm bandwidth.

 

I fitted a 1mm thick Schott UG-1 filter over the torch, using some black electrical tape. (When I get a much better torch, I plan to permanently glue a filter step-ring directly onto the barrel of the UV torch, so that I can easily screw in any filter, easily and neatly, as needed or required.)

 

Tiffen (Wratten) "Haze 2A" (pale yellow) filter was used on the full-spectrum camera (Panasonic Lumix G5) and adapted lens (a 35mm F/3.5 Kyoei design).

 

In my case, I ended up choosing to set my CWB (in-camera "shoot to set" custom-white-balance) on my Panny G5 via the Kelvin-adjustable option, and dial down to the reddest-biased position (2300K). I seem to like the resulting color rendition best, this way, instead of any other white-balancing scheme (unlike setting white-balance to Teflon / PTFE for UV reflective photography, for example).

 

Maybe it's because I failed to implement Schott S8612 into a stack with my UG-1 on my UV torch? In either case, I like this color rendition, best. The purple and burgundy combo are very pleasing to me.

 

(And these colors, by the way, come closest to the colors I saw with my own eyes, when the UG-1 filtered torch was applied to illuminating the flower. So, how far off can I be, if this comes so close to the fluorescing colors which I actually saw?).

 

It wasn't until post-photo, that I have been reading about people including IR-blocking glass on their UV torch. Why is that? I thought these UV LEDs do not emit IR, anyway?

 

Instead, the Kelvin-adjustable white-balance option set to the reddest-end bias (2300 K) seems to work best for UV-induced visible fluorescence photography in my case (but I am sure this is personal preference, while also most closely matching what my own eyes saw. Although I am sure that I did something "wrong" according to others. Haha. And I'd like to know what those errors were, if there are any. Thanks.)

 

(Perhaps with a 365 nanometer-wavelength torch, the blue channel will not be as heavily flooded, and thus may not require such an aggressive Kelvin-adjustment towards the red-end. This may just be something necessary for my 385 nanometer-wavelength torch. I don't know. Like I said, I had failed to include an IR-blocking glass on my torch. But I never imagined that this was necessary?)

 

In any case, moving the Kelvin-adjustable white-balance setting to 2300 Kelvin seemed to force the same colors, in-camera, as what I saw with my own eyes.

 

My results are below:

 

2 photos at the 35m focal length (actually 70mm, due to the Micro-4/3 sensor crop factor of 2x).

 

And then, 2 more photos with some cropping in, to get more detail.

 

NOTE: The pollen granules on the UVIVF image of the flower is really "glowing" and standing out! Neat! Obviously the "mess" left behind by visiting bees foraging for some goods.

 

post-34-0-57247000-1409076843_thumb.png

 

post-34-0-30915500-1409076849_thumb.png

 

post-34-0-79071000-1409076856_thumb.png

 

post-34-0-76937700-1409076863_thumb.png

Link to comment
enricosavazzi

[...]

It wasn't until post-photo, that I have been reading about people including IR-blocking glass on their UV torch. Why is that? I thought these UV LEDs do not emit IR, anyway?

[...]

I don't know why they use an IR-blocking filter. You are correct that UV LEDs produce virtually no NIR. I am only guessing, but UV may in principle excite NIR fluorescence, which might be a reason for using a NIR-cut filter if this NIR fluorescence is strong enough to swamp out the UV-excited fluorescence in the VIS.

 

Edit - I misread, I was thinking about using a NIR-cut filter on the lens. I still don't see a reason to use a NIR-cut filter on a UV LED, except perhaps if the UV from the LED is exciting a NIR fluorescence of materials inside the torch.

Link to comment

Thanks, Enrico.

 

Precisely what I was thinking.

 

And as for exciting residual NIR emission from an object, as a result of UV-induced fluorescence ...is it possible that the UV LED bandwidth of the torch can determine the resulting intensity of NIR emission from the fluorescence?

 

After all, we know that fluorescence is the result of lower-energy wavelengths being emitted from an object due to being bombarded by higher energy, after a remaining portion of that energy is also lost through molecular friction / mechanical acceleration (also known as "heat").

 

Thus, perhaps a higher-energy (365 nm) torch would yield a higher percentage of NIR as a component of total fluorescence, compared to a lower-energy (385 nm) torch? (Or maybe I have it backwards?)

 

Is it a valid consideration to assume that variable energy inputs will also yield proportionately variable spectral output fragments and associated intensities of each spectral range? (All other parameters remaining equal, that is).

 

Just a thought.

Link to comment

So, perhaps I am asking the fundamentally wrong question, here.

 

Instead, I should ask: Do I NEED to stack an IR-blocking glass with my Haze-2A filter already on my lens?

 

Is this a mandatory step, or is it an *optional* step which is only required to protect from IR contamination under certain conditions?

 

And if so, do my results in this case indicate any significant IR contamination? If that's the case, then why does the resulting image out-of-camera match what my own eyes saw?

 

(Unless, for some freaky reason, I can see NIR? Haha.)

Link to comment

Besides that, wouldn't NIR contamination cause red-channel information to rise, in proportion to the blue-channel?

 

Thus, if that were the case, then wouldn't pulling up a histogram window in live-view (especially through an EVF, where no stray light could escape into the dark environment, unlike an LCD screen) be of any use to discern potential NIR contamination as a component of the total spectral response?

Link to comment

A broadband ultraviolet excitation source can induce none, some or all of UV, Visible or IR fluorescence,

depending, of course, on the underlying fluorescence signature of the subject being photographed.

Therefore, as I've mentioned in a few other threads, if you want to record Visible fluorescence,

then you must block both UV and IR from entering your lens.

 

While it is likely that the fluorescence you induced in the Sunflower is Visible,

you cannot claim it as Visible unless you add IR blocking to your lens in addition to the UV blocking.

It is more accurate to label the photographs as UV-Induced Fluorescence. Which is a perfectly fine label.

 

In absence of an IR blocking (or Visible blocking) on a lens you would have no way to determine

whether any red fluorescence in a subject was from an IR emission or from a Visible emission

unless you knew ahead of time how that subject fluoresced.

 

IIRC, fluorescence signatures are very specific ("spiky")? Typically, narrowband excitation filters and recording filters are used to induce and record these specific fluorescent wavelengths. In our home studies using broadband filters we are usually just trying for a good - and interesting - photograph. You have succeeded in that.

 

I don't know either why there would need to be IR blocking on these torches?

Link to comment
Why would you use full-spectrum modified camera to try to record UV-induced visible fluorescence? Why not use unmodified camera for this task?
Link to comment

Alex, you describe my usual setup for fluorescence. If a 'normal' camera is at hand that is what I use. I still add the UV/IR blocking and Kodak 2E filter to the camera lens, to be on the safe side. There may be some violet/deep blue leakage from the excitation source and this is best avoided to gain clarity.

 

The deep red fluorescence by chlorophyll (665 nm if memory serves) is in the visible range and should come out on the photos as well, if you see it with your eyes. Bracts and stems can display this beautiful fluorescence. I rarely if ever observe it directly in the flowers themselves.

Link to comment
Yes, Bjørn, I also use regular unmodified camera for fluorescence. Hence my question to Igor. But I do not use yellow filter that often. Not when the fluorescence is mainly in blue part of the spectrum.
Link to comment
I make photos with and without the 2E (about the only filter I have left from the film days, by the way) and select the one with the higher clarity. Often there is no clear difference, sometimes the one with the straw-coloured filter appears better. However, as you already have indicated, blue-tinted fluorescence will of course be attenuated by a yellow filter even a pale one such as the 2E. But for me this apparently is not occurring frequently.
Link to comment

Why would you use full-spectrum modified camera to try to record UV-induced visible fluorescence?

Why not use unmodified camera for this task?

 

Fair question, so I'll throw in my answer also.

 

1) If I am already set up with a broadband cam to do a particular series of UV/IR work, then I will just continue to use that broadband camera/tripod/etc for the fluorescence work.

 

2) Ordinary un-converted cameras can also record some amount of UV and IR - especially older ones like the D70 or similar. So even on those we should use appropriate excitation and recording filtration to be sure we are inducing and recording what we say we are.

 

Regarding that second reason, I will also observe that I'm a bit of a "purist" about my inducing/recording filtration because that was the way I was taught by Shane and Vivek. :D I use two filters - one on the torch and one on the lens no matter what camera is in play.

 

Everyone please do note that we are certainly not espousing a particular set-up for fluorescence work here on UVP "as a website", OK? My only request as the Head Housekeeper and Second Bear (Bjørn being Chief Factotum and First Bear) is that all fluorescence work is presented with accurate titling and description including mention of filters used or non-filtered items used.

 

************

 

odds and ends:

 

White-balance for fluorescence work - I have never quite figured out what to do about that.

 

Try an experiment with IR-pass filters to see if you can induce some NIR or IR fluorescence. You can see my attempts here.

http://www.ultraviol...d-fluorescence/

Link to comment

I understand your reasoning, Andrea. But what about the difference in transmission of the modern OEM ICF (who uses D70 these days?) and the BG glass used to replicate the effect in full-spectrum cameras. They are definitely not the same. What you say about modern unconverted cameras recording UV and IR is true, but in this particular case it rarely plays any significan role. ICF employed in modern cameras cuts UV very well, and even some 400-405nm range. And same with IR. The transmission in the remaining part of the visible spectrum by ICF is also not identical to any regular BG glass. Adding yellow or pale filter will have no affect in this part of the spectrum. I think it is much easier to just change the camera, than to fiddle with WB in post-processing.

 

Besides, Igor did not use BG glass over the camera sensor in his setup, which means he recorded more red and less blue than the unmodified camera would.

Link to comment

A broadband ultraviolet excitation source can induce none, some or all of UV, Visible or IR fluorescence,

depending, of course, on the underlying fluorescence signature of the subject being photographed.

Therefore, as I've mentioned in a few other threads, if you want to record Visible fluorescence,

then you must block both UV and IR from entering your lens.

 

That makes perfect sense, Andrea. I appreciate the clarification.

 

I suppose I was just "lucky", in this case, to not record any (significant) IR contamination (since my resulting image out-of-camera, after white-balance adjustments, matches what my own eyes saw).

 

Regardless, I have my answer now. I just simply failed to include an IR-blocking filter along with my Uv-blocking Haze-2A filter. Next time, I'll know better. :D

Link to comment

Why would you use full-spectrum modified camera to try to record UV-induced visible fluorescence? Why not use unmodified camera for this task?

 

I've thought about that, too. However, I already had my full spectrum-modified Panny G5 on hand (as Andrea also explained as part of her reasoning), and was already quite adapted to the system and where all of the buttons, controls, custom functions, and saved settings were. (In other words: convenience.)

 

After all, fiddling in the dark with a camera is cumbersome as it is. Thus, having "button memory" helps.

 

Although I can certainly try some other cameras out. This was never an issue for me. It was more about system familiarity.

 

(Another reason behind me going this route: I also had the curiosity to see how the image would look ... in real-time through the EVF ... by putting different filter combinations on both, the lens and the UV-emitting source. In other words, I wanted my first experience with UV-induced fluorescence to give me insight by playing around with different set-ups and filter stacks. That kind of flexibility is not possible with an unmodified camera.)

 

Otherwise, I agree. Outside of experimentation and gaining insight (through trial and error), an un-converted camera would probably be less of a hassle, once the person (myself) understands what this is all about. But, I definitely didn't want to take the "easy" or "no-brainer" way, the very first time, because I wanted to also understand why things appear the way they do, based on various filter uses and settings. And there is no better teacher than hands-on experience.

 

However, as Andrea and Bjørn stated, I'd still make sure to implement filters on the lens of an unmodified camera, anyway ... because there is no guarantee that an unmodified camera would exhibit "proper" (well-balanced) VIS colors, without the "better-safe-than-sorry" approach of always using IR / UV cut filtration.

Link to comment

Yes, Bjørn, I also use regular unmodified camera for fluorescence. Hence my question to Igor. But I do not use yellow filter that often. Not when the fluorescence is mainly in blue part of the spectrum.

 

Actually, I was getting too much blue, based on my initial settings. So much so, that the resulting image (in-camera AWB) was nearly blue-monotone, and looked nothing like what my eyes were seeing.

 

Putting the pale-yellow (2A) filter on the lens balanced the colors out, a bit, and made the in-camera image more closely match what my eyes were seeing ... which included some yellow-greens and burgundy-reds in the VIS fluorescence, too.

 

In fact (as explained in my comments above), even with the pale-yellow (2A) filter on the lens ... I even had to make further white-balance adjustments in my camera, and move the Kelvin-temperature to 2300 K, in order to finally get the colors in-camera to match what my eyes were seeing! :D

 

So, if anything, the image was way too blue-dominant, in-camera. Much more so than even what my eyes saw. Without the pale-yellow (2A) filter on the lens, I couldn't get it to match the actual fluorescence as observed by my eyes.

 

(It's possible though, that all of this necessity for extreme adjustment was due to the fact that I had no IR-block filter stacked with my UV-block filter. But I don't know, because I failed to use one on the lens. Next time, I will try the stack. What confuses me, is how any potential NIR contamination could make the image too blue-dominant? Wouldn't some NIR contamination shift color dominance toward the red-end? Hmmm.)

 

In either case, my results worked out for me (luckily), after some tedious and tiresome settings "acrobatics."

Link to comment

Ok. It just hit me. I think I finally discovered the true source of why my VIS images were too blue-dominant (before some extreme white-balancing adjustments).

 

I remember that the background which I used (behind the flower) was fluorescing, too. In a very vivid blue color! And since flower petals are often translucent and permit some light to also pass through and "backlight" the petals, I believe I have now found the source of my overly blue-dominant initial results (when in auto-white-balance).

 

Apparently, the towel which I used for the background (specifically, the dye or paint in the towel) behind my flower was adding its own fluorescence glow to the resulting image (including passing strong blue colors through the petals).

 

What a strange mix-up. :D

 

Mystery solved, perhaps?

Link to comment

Could be, Igor. :) Fluorescing backgrounds might indeed gum up the works. Many fabrics and papers have optical brighteners which fluoresce.

 

Hope you noted what Alex wrote above. Most likely we don't have NIR fluor. We only need to be aware that it might be possible.

And see the following.

 

*****

 

Alex - all good questions for sure!! And thank you for asking those and pointing out some of the finer details. But I have no answers. I just have not experimented with Fluorescence shooting all that much or thought about it all that much. I should do this I know. It always boils down to the time factor.

 

who uses D70 these days?

Quite a lot of hobbyists or beginners still pick up a D70 for pennies just to give UV a whirl and see what it is like.

 

What you say about modern unconverted cameras recording UV and IR is true, but in this particular case it rarely plays any significant role.

 

Like I said, I was taught to use two filters in fluorescence work as would be done in any serious lab work, and I've stuck to that. I think for the generic Fluorescence photography we do here, this might be called using both a Belt and a pair of Suspenders. But that way I always know my UV-Induced Visible Fluorescence Pants are being held up securely. :D

 

I still think that if you omit a lens filter when using an unconverted camera, it is only fair to label it UV-Induced Fluorescence. There is nothing perjorative in that label - I'm just striving for accuracy - especially because we have a lot of readers who really would not know the finer datails of how an unconverted camera may or may not record UV or IR.

 

OTOH, most people omit the excitation filter on their UV torches, it seems. But I'm not sure whether to lean on that or not. Would we call it Mostly-UV-Induced Fluorescence? :D :) It begins to get tedious. While we strive to be scientific here in our presentations and labels, we are not expecting laboratory-quality experiments.

 

I think it is much easier to just change the camera, than to fiddle with WB in post-processing.

 

I've made some camera profiles so that I no longer have to perform a white-balance step when editing in any converter/editor. So I will try those on future UVIVF photographs. That way I can go back to some of my older fluor fotos and see what they should have looked like. Or maybe I can redo some of them.

 

*****************

 

After all, fiddling in the dark with a camera is cumbersome as it is. Thus, having "button memory" helps.

 

This is where your Photographic Assistant becomes useful. Have someone stand by the light switch to toggle it at your command. What, you don't have a P.A.?? Me neither. :D

 

I drag a small lamp into the closet for use while I get everything set up for the fluor foto and then, after chasing the cats out of the closet, shutting the door and readjusting the flower they have tried to play with, I refocus, turn off the lamp and fluor away.

Link to comment

Admin Note: While I was either reading or writing this thread, I apparently leaned on or otherwise engaged my Track Pad and temporarily un-approved a couple of posts above. To a member it would have appeared as though two posts had suddenly disappeared!! It is fixed now, but please accept my apology for this unintentional mistake.

Sorry 'bout that.

Link to comment

I still think that if you omit a lens filter when using an unconverted camera, it is only fair to label it UV-Induced Fluorescence. There is nothing perjorative in that label - I'm just striving for accuracy.

 

But if one uses a pale 2E filter (cutting below 410 nm) or yellow filter (cutting below 460-470nm) we are effectively removing part of the visible spectrum, which some books state to start at 390nm, don't we? So may be we should call it UV-Induced Part-Visible fluorescence? I understand your desite for accurate terminology, but the problem with many therms are that they are not accurate enough.

Link to comment

True dat, as they say. :D

 

The acronyms could become horrendous.

 

We will just have to call it "Fluor" as I did in the beginning before everyone tried to teach me what to call it. :D :) :)

It was a lot easier to write.

 

But as long as the excitation equipment and recording equipment is described, everyone can decide what they want to call the light they are seeing. Meanwhile we try to indicate some general terminology for those who are new to the lovely process of photographing Fluor.

Link to comment

One of the reasons I started to characterize the ICF of DSLRs was for the application of those cameras to UV Induced Visible Fluorescence photography.

 

It is a tricky situation because older models often leak both UV and IR and newer models often have too aggressive ICFs which start at 420nm and end around 680nm clipping the visible spectrum at both ends. That project was actually started before Baader introduced the nice UVIR cut filter, however even that has undergone revision with regards to wavelength cut-off points.

 

First I believe it is important to eliminate any IR from the light source when possible, this allows subsequent modification of any Red/IR fluorescence captured to be defined by the lens filter of your choice e.g. "visible bandpass" or longpass cut-off filters (my favorite).

 

Second, many UV sources also leak blue light. This is difficult to handle because blue fluorescence from the subject may also be involved and it is very tricky to cut off the blue leakage at the source without attenuating the UV. Often a compromise must be made e.g. is it acceptable to clip a little blue around 400-420nm. The 2E only passes 9% at 420nm however it was one of my most used filters in fluorescence photography because blue fluorescence emission is often broadband and therefore clipping a little blue off the "UV" end didn't really effect the image colour appearance.

 

A modified full spectrum camera does provide the flexibilty to tune the cut-off wavelengths through additional filtration e.g. 680nm or 700nm etc. In fact I liked the D70 for that reason - a "semi-broadband" unmodified camera.

 

Daylight balanced slide film always seemed to work best for Visible fluorescence but often required scanning and colour correction and so I found Daylight balance at capture followed by subsequent modification in the RAW converter to work quite well.

 

I have observed the fluorescence (dark room, dark adapted vision) of thousands of gem materials over the years and the most interesting point was that if the observation was made by two people, it was common for them not to agree on the exact colour, especially weak fluorescence. So you can post process for the "perfect" match but it may still not be "perfect" for the other person.

 

Chlorophyll fluoresces in the red and NIR and so a decision needs to be made - how much IR to include or exclude. For "visual" representation of the fluorescence, the IR contribution must be removed from the image pre-capture (especially on a full spectrum camera) AND that leads to the issue of where to define that cut-off point. Perhaps a longpass filter with a similar spectral cut-off slope to that of the human eye, which is what many ICF filters attempt to replicate. Bringing us back to the unmodified camera.

 

In addition, reducing the spectral range in this way also reduces CA effects.

 

This really makes it clear why it is important to clearly describe the equipment used, light source (specific), excitation and emission filters*, model camera (including unmodified vs modified).

 

* The exact model of the filter can be important e.g. in the case of the Baader UVIR cut filter, production changes effected the wavelength range.

Link to comment

I have observed the fluorescence (dark room, dark adapted vision) of thousands of gem materials over the years and the most interesting point was that if the observation was made by two people, it was common for them not to agree on the exact colour, especially weak fluorescence.

 

Very interesting. This be must be due to the fact that dark adapted vision is scotopic (rods only) and therefore less acute for distinguishing colours? And also perhaps each individual has slightly differing scotopic vision?

So if we can somehow set our cameras to record fluorescence colours accurately, then the photograph would be useful for seeing the "true" colour of the fluorescence in absence of accurate measuring with a spectrometer ??

 

 

This really makes it clear why it is important to clearly describe the equipment used, light source (specific), excitation and emission filters*, model camera (including unmodified vs modified).

 

This is what is called the Takeaway. :D Or the Summary or the Key Point.

Link to comment
enricosavazzi

Besides that, wouldn't NIR contamination cause red-channel information to rise, in proportion to the blue-channel?

[...]

Depends on the WB and on the NIR wavelength. Regardless of WB, NIR close to the VIS range is usually recorded as pink/red, while deeper NIR (900-1,100 nm) records at approximately the same level on all color channels (in practice, all Bayer filters are equally transparent to these wavelengths). So at these wavelengths, no, you do not get a prevalence of red.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...