Jump to content
UltravioletPhotography

Discussion: UG5 + IR-Cut Filter Stack, Part 1


igoriginal

Recommended Posts

It should also be pointed out that a UG-5 + IR-cut stack is technically a simulation of tetrachromatic vision (UV + blue, green, and red), and not UV + blue-green. Some red does get through in that stack. Whereas, our current scientific understanding is that bees are not true tetrachromatics, because their ability to detect red-wavelength light is highly muted or limited. Instead, I would say that "bee vision" is "color-shifted" (towards the end of higher energy), so that they see predominantly in UV + blue/green (and an increasingly muted perception of color as the wavelengths head further towards the red-end, so that yellows are still visible, oranges are decreasing in saturation, and red is almost non-existent to them). Thus, bees mostly remain trichromatic animals, just like humans ... just color-shifted more towards the violet-blue side.

 

On the other hand, it is "butterfly vision" that is assumed to be true tetrachromatic vision (as well as certain species of birds). However, I am aware that this still remains an area of great contention, because our current understanding of the "color interpretation" of other animals is based upon what wavelengths their eyes can detect, but it's still up to conjecture as to how their brains actually assign (or map) these colors. Meaning, just because we know what wavelengths an animals eyes can detect, doesn't mean we can predict with any certainty how their brains will interpret the actual wavelength in terms of assigning a "color" to any given wavelength. Which is why the better way to state such photography is to add the word "simulated" to the description. (Ex: simulated bee vision, simulated butterfly vision, and so on.)

 

Anyway, the bottom line though is that UG-5 + an IR-cut stack is not "bee vision", because it produces a tetrachromatic result, and not just UV + blue-green. (There is red in there, too, and some red-colored flower parts will show up in the image as red if the white-balance is neutrally-distributed). I used to go along with this inaccuracy for a while, too (excluding the red in the description), until my own exploration of such stacks taught me better (when I started seeing reds showing up in such stacks). Thus, to produce a "simulated bee vision" result with a UG-5 + IR-cut stack, I would de-saturate the red-end results, or even consider bringing any reds close to a grey-scale appearance.

Link to comment
Not sure I follow the line of argumentation here. The camera does *not* record UV + blue in the same channel. Whatever fraction of red going through such filter stacks will typically end up with UV, in the red channel.
Link to comment

Not sure I follow the line of argumentation here. The camera does *not* record UV + blue in the same channel. Whatever fraction of red going through such filter stacks will typically end up with UV, in the red channel.

 

No argument. I am just pointing out that a UG5 + IR-cut stack yields tetrachromatic results.

 

I am not talking about the actual (physical) information in the camera's "channels." I am talking about the resulting images, in themselves.

 

Here is an example, below, of a Coreopsis that I shot comparison photos of. In Visible, it has a red center. In UG5 + IR-cut stack, it still has a red center. (UG-5 + IR-cut stack was white-balanced to a color-neutral target: PTFE).

 

My point is that we cannot call UG-5 + IR-cut stack a "bee vision" stack, because its color yields a tetrachromatic result (regardless of actual channel information). More akin to butterflies and certain birds (or any animal that supposedly sees in a tetrachromatic vision). But not bees (at least, not honeybees.)

 

post-34-0-37712100-1445526190.jpg

 

post-34-0-04265800-1445526198.jpg

Link to comment

Of course, Bjørn, your statement now makes me wonder if when UV + Red are both present within a flower's nectar guide (bull's eye) center, then what will the tetracrhomatic animal see? The UV, or the red? Or maybe a messy and unpredictable overlap of both of them? Quite mind-bogging.

 

(I definitely notice a slight magenta-ish undertone to the red center, in the UG-5 + IR-cut stack photo).

Link to comment

Igor, please think about what you want to write and polish it up before placing it in a post. The constant edits are not acceptable.

I did not understand that Igor was working off a cellie. My apologies!

Link to comment

Sorry about that. I am often limited to use of a small mobile device (cellphone), because I have a hectic work schedule and infrequently have the luxury of sitting myself down in front of my desktop PC, at home. Because of this, I am prone to clipped sentence structures and out-of-bounds editor boxes. This phone screen is not well-compatible to full web browsing, and I often have to type "blind" until I hit "reply" and see my result. Often, necessitating editing because of this.

 

But, I can certainly just dicipline myself to be patient enough to wait until I can be in front of a computer, again (although, again, this is infrequent so it would force me to be left out of UVP current events / posts). Apologies.

Link to comment
In other words, this is not an issue of "not thinking" before sending. It's an issue of having to be crippled by my medium.
Link to comment

OK, I understand now!! Thank you for that explanation. I crossed out my original comment about this and offered up apology. "-)

 

It is often helpful to just very briefly mention that you are on a cellie and must edit for spelling/typos. That way we don't get worried about revised content. That said, I wouldn't worry too much about typos and suchlike. I sometimes fix typos for folks when I'm reading. But it's your call how you want to handle it as long as you remind me what's going on.

 

Carry on, no worries !!

Link to comment

but it's still up to conjecture as to how their brains actually assign (or map) these colors.

 

There has been research into the bee (or other insect) brain and its opponent channels. So researchers are understanding better how the insect might assign the colour signal.

 

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

 

Igor, I understand your grapplings with this insect vision whether tri- or tetra-chromatic. It is a fascinating and fun subject! I have thought a lot about it myself and have my own grapplings.

 

Just sticking for a moment with bees...

 

To attempt to produce a model (or simulation) of bee vision, we use a 3-channeled digital camera/lens/filter tool which "sees" in its own way to record human-visible RGB flower colours for translation into GBUv bee vision. To do this we have to add a "4th channel", our UV recording, to the input mix. In short, we are attempting to map 4 variables (RGBUv) onto 3 variables (RGB). In math we'd call this projecting a 4-space onto a 3-space.

 

One typical channel stack is to take the visible RGB flower photo, toss the visible Red channel and replace it with the UV photo's Red channel. This channel stack makes a certain amount of sense because it keeps Green and Blue where they belong and because current Bayer-filtered cameras record a lot of UV in the Red channel. This often makes for a brightly coloured, high contrast result.

RGB(visible flower) + R(UV flower) RRuvGGvisBBvis(model) UvGB(bee)

 

Alternately, there is a stack favored by some where the bee's receptor channels are "right-shifted". (Implied here is that shortwave UV and Blue are "on the left" and longwave Red and IR are "on the right".) So the UV Blue channel is mapped to Blue, visible Blue is shifted to Green and visible Green is shifted to Red. This channel stack sometimes makes for a slightly weaker initial result because the UV photo's Blue channel does not contain as much information. But you can always boost the blue in the editor, so you wind up again with a nice photo. This channel stack makes a certain amount of sense because it preserves the progression of short-to-long wavelengths. Also note that the typical Bayer-filtered UV photograph under a fairly wide UV-pass filter like the BaaderU coughs up mostly blue and some yellow after white balancing. But that is a bit of an artifical construct, so I don't want to carry it too far.

RGB(visible flower) + R(UV flower) RGvisGBvisBBuv(model) GBUv(bee)

 

But what is lost in each of these channel stack mappings? I.E., what could p.o.s.s.i.b.l.y. go wrong? :)

 

In the first model, any visible Red reflections are lost. But, you say, visible Red reflections do not matter because bees do not have a Red receptor. And I say wait just a minute....

True, bees do not have a Red receptor, and their GBUv receptors are not stimulated by a Red-reflecting, UV-absorbing object. But that does not mean a Red-reflecting, UV-absorbing object is invisible to bees! Bees can see the +R-UV object, and they are even able to put together cues to determine that the Red object is Red. [agb: find the reference]

So we have to ask whether this model may misinterpret how a Red object is seen by the bee (depending on whether the Red object is UV-reflecting or UV-absorbing)?

Also note in this model, that any UV recorded in the Green and Red channels (and there is some) is also lost.

To sum up, input information may be lost when mapping 4 variables onto 3. [again, yes/no? comments?]

 

Similarly in the second channel stacking model, information is lost.

 

It could be that these questions/problems arise with channel stacking models because (so far) most of this kind of modeling has been done using broadband UV-Pass filters to record the UV reflections. Thus some UV wavelengths are recorded in the B and G channels. Experimentation with narrower band UV-pass filters may provide a more robust model when stacking channels because the UV record will tend to be monochrome.

 

The various filter stacks give results which cause the same kinds of questions to be asked.

For example, when we use certain thicknesses of UG5 + S8612 filter stacks, they pass UV, B, and G, and we are throwing away any Red reflections from the get-go. And so again we have to ask whether Red reflecting, UV-absorbing objects are misrepresented under this filter stack? And again we see that that any reflected UV, while mostly recorded in the camera's Red channel, is also recorded in the B and G channels - the same channels where the reflected B and G which is passed by the filter is recorded. So in the case of the UG5 + S8612 filter stack, I'm not entirely sure how to interpret the results.

However, my particular UG5(1.5mm) + S8612(1.75mm) filter stack seems, so far, to produce remarkably good emulations of bee vision for some flower colours. I am not sure yet whether it works for all.

 

I think it is important to keep in mind the (so-called) theoretical steps of this attempted mapping up-front.

(BTW, I know most of us know all this stuff already, but I try to write with newbs in mind also.)

 

Ignoring saturation for the moment, flowers which are UV-absorbing (-UV) have a natural mapping to blue, cyan and green.

The UG5(1.5mm) + S8612(1.75mm) usually produces this result after a white balance adjustment in-camera or in-editor.

 

Magenta (-UV and +R+B or +r+b ), purple/violet (-UV and +R+b or +r+b ), blue (-UV and +B or +b ) +B, bee-blue B, blue

White (-UV+R+G+B ) +B+G, bee-cyan C, cyan

Yellow (-UV+R+G) +G, bee-green G, green

Orange (-UV+R+g) +g, bee-green g, green

Brown (-UV+r+g) +g, bee-green g, green

Red (-UV and +R or +r) no receptor stimulation, so bee-black dark red or black?

 

Flowers which are UV-reflecting (+UV) present the biggest mapping problem. Here we are dealing with bee colours obtained by stimulation of the UV receptor: +G+UV, +B+UV and pure +UV. Degree of UV-reflection is not considered here.

Now I have to go look at filter stack photos and see how the UV-reflecting mappings work out because I don't remember them all off the top of my head.

 

Magenta, purple/violet, blue +B+UV = bee-UVblue ?

Yellow +G = bee-UVgreen yellow?

White +B+G+UV, all receptors stimulated = bee-white white?

Orange +g+UV, = bee-UVgreen ?

Brown +g+UV = bee-UVgreen ?

Red +R+UV = bee-UV only red ? [but see results below in the next post.]

Link to comment

Here is a tropical hibiscus flower which is scorchingly red and also happens to be mostly UV-reflective.

 

Under the UG5(1.5mm) + S8612(1.75mm) you will see that this reflective Red/UV area becomes a reddish colour, after white balance. That's important to note - the red appears after white balance. Red does not seem to be the predominant colour recorded in the raw file when analyzed with Raw Digger.

 

And so Andrea gets stumped about how to map +R+UV.

 

Equipment[ D600-BB + 105/4.5 UV-Nikkor + Sunlight ]

 

Visible[ f/11 for 1/30" @ ISO-200 :: Baader UVIR-Cut Filter ]

hibiscus_visSun_20152409wf_39928pn01.jpg

 

UV[ f/11 for 15" @ ISO-200 :: BaaderU UV-Pass Filter ]

UV after white balance

hibiscus_uvBaadSun_20152409wf_39950pnEditsOnNef01.jpg

 

UV raw: before white balance

Yes, raw composites look dark. However, Raw Digger shows the file to be properly exposed.

hibiscus_uvBaadSun_20152409wf_39950rawDigger01.jpg

 

UV raw: sample from two bottom petals

hibiscus_uvBaadSun_20152409wf_39950rawDigger.jpg

 

Histogram for the raw sample

hibiscus_uvBaadSun_20152409wf_39950sample.jpg

 

 

Bug[ f/11 for 1.6" @ ISO-800 :: UG5(1.5mm) + S8612(1.75mm) BGUv-Pass Filter ]

Here we want to look at how the Red, UV-reflective areas of the hibiscus are recorded.

 

Bee vision simulation after white balance

hibiscus_ug5s8612Sun_20152409wf_39963pn99Crop.jpg

 

Bee vision simulation raw: before white balance

hibiscus_ug5s8612Sun_20152409wf_39963rawDigger01.jpg

 

Bee vision simulation raw: sample from two bottom petals

hibiscus_ug5s8612Sun_20152409wf_39963rawDigger.jpg

 

Histogram for the raw sample

All three channels have data. If we typically see reflected UV recorded in the red channel (before white balance), then where is all this blue coming from? The flower itself seems to have very little reflected visible blue. Many red flowers do have a touch of blue or a slight orange cast. But I can't see enough actual blue in this Hibiscus (either in my colour-profiled photograph above or when viewing the flower in person) to account for all this recorded blue under the UG5(1.5mm) + S8612(1.75mm) unless perhaps, as mentioned above, the UG5 is so wide (broadband) that we are capturing a lot of UV in the blue channel. A spectral analysis of a hibiscus petal would help determine what is going on here.

After looking at this, I decided to make a very much smaller sample in a more uniform area of the petal.

hibiscus_ug5s8612Sun_20152409wf_39963sample.jpg

 

Bee simulation raw: very tiny sample

In Raw Digger I blew the photo up 500% and selected as uniform a sample as I could from one of the slightly brighter areas on the petal. Here is a screen shot of that.

hibiscus_ug5s8612Sun_20152409wf_39963tinySample.jpg

 

Histogram for the very tiny raw sample

Even more blue than before !

So -- what do we do about modeling a Red, UV-reflective colour in bee vision? Do we map it to the redd(-ish) colour we get after white balance? Or should we map it to this raw combination of r+B, where there is somewhat more B than R ?? I do not know.

hibiscus_ug5s8612Sun_20152409wf_39963tinySampleHisto.jpg

 

Here is an edited version of the raw composite.

Some illumination and contrast was added. Then a bit of detail enhancement and some saturation. (Not sure if I took it far enough.)

Is this at all representative of bee vision? I do not know.

hibiscus_ug5s8612Sun_20152409wf_39963rawDigger01rawDiggerillumContrSat.jpg

Link to comment

Now the opposite, a predominantly Red flower which is very UV-absorbing (+R-UV) except on the tips which are very UV reflective showing some +R+UV and +R+g+UV.

There are accents of orange to yellow-orange on the tips of the petals and on the disk flowers.

 

Zinnia haageana

4 August 2015, Charlotte Rhoades Butterfly Garden, Southwest Harbor, Maine, USA

 

Equipment[ D600-BB + 105/4.5 UV-Nikkor ]

 

Visible[ f/11 for 1/200" @ ISO-200 :: Baader UVIR-Cut Filter :: Sunlight ]

zinniaHaageana1_visSun_20150804charlotteRhoadesSwhME_38048pn01.jpg

 

 

UV[ f/11 for 1/20" @ ISO-800 :: BaaderU UV-Pass Filter :: Nikon SB14-Mod ]

After white balance

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38068pn01.jpg

 

UV raw composite before white balance

Raw Digger shows that this photo has some areas of underexposure.

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38066rawDigger.jpg

 

UV raw composite sample

An area on the UV-absorbing portion of the ray was chosen for the sample.

Screen Shot 2015-10-22 at 4.45.27 PM.jpg

 

Histogram for raw sample

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38066sampleHisto.jpg

 

 

Bug[ f/11 for 1/1.3" @ ISO-200 :: UG5(1.5mm) + S8612(1.75mm) BGUv-Pass Filter Stack :: Sunlight ]

Bee vision simulation after white balance

zinniaHaageana1_ug5s8612Sun_20150804charlotteRhoadesSwhME_38056pn01.jpg

 

Bee vision simulation raw: before white balance

zinniaHaageana1_ug5s8612Sun_20150804charlotteRhoadesSwhME_38056rawDigger01.jpg

 

Bee vision simulation raw: sample from petal

Screen Shot 2015-10-22 at 4.46.49 PM.jpg

 

Histogram for raw sample

zinniaHaageana1_ug5s8612Sun_20150804charlotteRhoadesSwhME_38056sampleHisto.jpg

 

Bee vision simulation before white balance with contrast, illumination and saturation pushes.

zinniaHaageana1_ug5s8612Sun_20150804charlotteRhoadesSwhME_38056rawDigger01pn01.jpg

Link to comment

Earlier on I tentatively mapped UV-absorbing Red (+R-UV) to either dark red or black in the bee vision model.

If the preceding work is valid, then perhaps the mapping should be to dark magenta?

 

And UV-reflective Red (+R+UV) was mapped to red in the bee vision model.

Again if the preceding work is valid, then perhaps the mapping should be to light magenta?

 

Alternately, perhaps the UG5(1.5mm) + S8612(1.75mm) stack is simply passing too wide a selection of UV? Or possibly passing too much blue? We know this combo passes more blue than green by looking at the charts. But that has to be considered in light of (no pun intended) the amounts of blue and green light in sunlight. There is more green than blue, so perhaps it is ok that the filter stack passes more blue.

 

To complete this little bee vision experiment with red flowers, I need to provide some channel stack composities for us to look at also.

Link to comment

This Tropical Hibiscus is a UV-reflective Red flower.

Tropical Hibiscus Channel Stack #1: Ruv --> R, Gvis --> G, Bvis --> B

hibiscus_uvBaadSun_20152409wf_39950_RuvGvisBvis.jpg

 

Tropical Hibiscus "Right-shift" Channel Stack: Gvis -->R, Bvis --> G, Buv --> B

hibiscus_uvBaadSun_20152409wf_39950_rightShift.jpg

 

This Zinnia is (mostly) a UV-absorbing Red flower (except for the ray tips).

Zinnia haageana Channel Stack #1: R=Ruv, G=Gvis, B=Bvis

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38070_RuvGvisBvis.jpg

 

Zinnia haageana "Shift" Channel Stack #1: R=Gvis, G=Bvis, B=Buv

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38070 _rightShift.jpg

Link to comment
A few humans (<10-9 of general population) are said to be RYGB tetrachromats. If a suitable display device could be found, these fortunate individuals could actually experience tetrachromatic simulations of animal vision. The rest of us can only try to imagine....
Link to comment

All human colours are RGB combinations regardless of whether they are seen with 3 or with 4 cone types in the eye. So human tetrachromats don't see anything outside the RGB possibilities. They can, however, distinguish subtle differences between colours better than the rest of us.

 

As the daughter of an RG-colour-blind father, I supposedly have 4 cones. I do not know that for a fact. I'm just going by the accepted genetics. However I am truly not aware of having any super powers when it comes to distinguishing colours!! I can say that might I notice colours more than my husband does (he is not colour-blind), but I think that is just the typical female-male thing. I do sometimes have a terrible time when clothes shopping trying to find the "right" colour top to match a skirt - especially if purples are involved. I've never yet met any two purples which "match".** But then women are very picky about stuff like matching colours. So I really don't think there is anything real going on with those 4 cones. I wonder how they test for being a true tetrachromat?

 

Added:

** "I've never yet met any two purples which match" was said somewhere, sometime, by somebody else. But I loved the phrase and picked it up. :)

Link to comment

Igor, Yes and no, but wrong.

It is a recipe. Like baking a cake, you can't simply talk about two ingredients. You need to be specific.

This means specific glass brands, types, and thicknesses.

Indeed if you stack a specific thickness of UG5 with a specific thickness of S8612 (for example) you will not have any red or IR transmission.

On the other hand, if you use the wrong thicknesses, or the wrong brands of glass, then you will have a strange recipe that may not block red and IR.

The 'wrong' stack may be pretty, and there is nothing wrong with being wrong, but it is still wrong when it comes to UV+Blue/Green with no Red/IR.

Many glass brands claim to be 'equivalent' to other glass brand types, when in fact some of these are not equivalent, even at any thickness.

You need to be specific about a formula when you are talking about a formula.

 

Schott UG5 is very similar to Hoya U-330, and these can be interchanged in the recipe for UV+Blue+Green. These are the only two I have tested, but they work for no red when you use the correct stack thicknesses.

I always prefer Schott glass because Schott data is much more precise, with a data resolution of 1nm, and no missing sections of data.

Take a look at any Hoya data sheet and you will see that data resolution is only 10nm, and there are large sections that are missing, like the entire visual range for U-340 (for example).

 

Boon did a lot of work on getting the UG5 stack to work precisely, with no Red/IR, and I would like to give him credit for that.

Link to comment

Cadmium Steve - thank you thank you thank you for this reminder about being specific when presenting work with filter stacks!! I'm going to go back and add the thicknesses to my filter stack mentioned above.

 

I just ordered some new S8612 glass, a 1.75mm and a 2.00mm, the latter a new thickness for me. On a side note, I finally got some oxidation on my original S8612 filter. This is not at all unexpected - especially because I have used the filter extensively in wildly varying conditions. BG glass is very prone to this oxidation as you have always taken pains to point out. I tried a couple of things to remove it, but so far no.

 

******

 

I must note that the rather difficult and very interesting modeling of bee vision does not necessarily preclude using a UvBG filter stack which passes some red light. We must represent red objects in some RGB way to make a human visible model of bee vision. And there will never be such a model which accurately shows us how the bee sees because bee vision involves UV and because we are projecting 4 variables onto 3 when attempting such modeling. And these simple models entirely exclude other variables such as saturation or brightness.

 

As I figured out tentatively above, one of the bigger problems in this kind of modeling is not whether red does or doesn't leak thru our filter stacks, rather it is whether or not we should be using such a broadband filter as the UG5 or the Hoya 330. I think not. But I am not yet ready to shell out $$$ to purchase narrowband UV-pass filters. :)

 

Red is not the only problem in modeling bee vision. Green leaves are a huge problem in bee vision models. Under various UG5 + S86112 stacks, green leaves show up as green. But bees see most green leaves, apparently, as some kind of bee-grey. Short of hand-painting the colours in Photoshop, there is not yet a very accurate way of representing green leaves in either filter-stack or channel-stack bee vision models.

Green leaves = +uv+b+g-R or +uv+b+G-R

 

Reference:

UV as a Component of Flower Reflections and the Colour Perception of Hymenoptera, Chittka et al., 1993, VIsion Research 34:1489-1508

This paper analyzes 573 species of flowers.

 

My Bug stack is UG5 (1.5mm) + S8612 (1.75mm).

 

Link to comment

I'm sitting here. Staring at a bunch of Red Gerberas given to me for my birthday.

All I can think about is how the blasted bees see these blasted red flowers!

Sigh.

Link to comment

Happy birthday Andrea

Here is a tip for your birthday, wash that sticky S8612 in some household (say 3%) hydrogen peroxide, & enjoy the re-newed glass ;-)

Then for lasting pleasure, keep a little satchel of silica gel in the filter case.

Cheers

Col

Link to comment

My, oh my. I get stuck distracted with yet more work for the past 24 hours, and then I come back and have much to catch up on. Wew. :)

 

@ Andrea,

 

First but not least - HAPPY BIRTHDAY!!!

 

Thank you very much for that very informative, highly-detailed (and beautiful) collection of examples concerning some of your ongoing thought-experiments. I must be honest and say that it is going to take me a number of days to get back you, because I have some deadline-related paid projects to complete before Monday morning (or my butt is toast). But I definitely look forward to taking the time to read over the above content with greater scrutiny, once that is behind me. VERY NICE work.

 

For now, I want to respond to at least one thing: I never stated that red flowers are "invisible" to bees. I simply stated that they probably see something else there, rather than human red (or, the bee version of red might even be a grayish color, according to their color-sense. You never know.)

 

Some other scientific writers have even proposed that bees make up for their lack of the color-sense of red-associated wavelengths by color-mapping it in green! Meaning, if a bee were to come across a red rose, it would see the plant as one contiguous green object - petals and leaves alike. Clearly, the rose wouldn't be invisible to them. It would just default to their sense of green, when there is no color information to interpret. I'd say that this is as much a possibility as any other consideration. Who knows?)

 

@ Steve,

 

Ahhh. Yes, that makes sense, since I remember some of the transmission-curve data which you've sent my way over the years.

 

However, if that be the case, then I am not entirely convinced that it is "proper" (in your words) to use a thick-enough UV + VIS stack to cut off all red, if one's desire is to actually simulate animal tetrachromacy (as opposed to bee trichromacy). This is because some other insects can see red-associated wavelengths, so why would we use an entirely red-blocking stack marked "bug vision", when such a stack should probably be exclusively referred to as "bee vision?"

 

(If some other bugs - not bees - can see red-associated wavelengths, then why use a UV + VIS stack which blocks red? That seems counterintuitive contradictive.)

 

In other words, what I am proposing is to very deliberately use a UV + VIS stack which permits some red to pass through, when the goal is to simulate animal tetrachromacy ..... whereas, a much thicker (red-blocking) stack (UV + blue/green) can be reserved for simulating bee trichromacy.

 

Alas, my assertion is that both types of stacks (a UV+VIS stack which excludes red, and a UV+VIS stack which includes red) each have their use, if contextually-applied to the simulated animal-vision in question.

 

Anyway, it's nice to hear from you! (Long time no chat!) I actually lost your email, because I had to switch web-hosting services, and in the process, all of my old emails (as well as address book) went through the digital chopping block. So, I am going to PM you for your email.

 

@ Col,

 

Peroxide, huh?

 

I'm a white vinegar kind of guy. And this has worked beautifully for me, as well. (Especially for removal of lens fungus, when it is found, where no other cleaner would have such as good of an effect without damaging the glass. Alcohol doesn't cut it, when it comes to lens mold/fungus. Nor does your typical "lens-cleaning solution." But no fungus I have ever met could stand up to vinegar and win. :-) But, peroxide, huh? Interesting.)

Link to comment

Col, thanks for the hint. I will try that.

 

I'm up too late tonight wondering what I got wrong in the earlier discussion??

Perhaps it is just better to let the photos speak for themselves.

 

(oh argh - bad phrasing there, photos speaking! Goodnight, all. See you Monday.)

Link to comment

Igor, I don't mean to offend you, however I find your initial statement erroneous and misleading, especially for those who have less experience with UG5 stacks than we do,

and I believe this should be corrected in order to not confuse anyone who might be lurking in the balcony, and that is my main point in participating in this discussion.

I have skimmed here for numbers and specificity, but I still don't see any.

There is no reference point.

This topic needs a foundation. You need to define the stack you are using.

It is incorrect to say that "UG5 + IR-cut" is UV+Blue+Green+Red.

"UG5 + IR-cut" is what you stack it to be, the recipe/formula is the essence of the transmission profile.

You have overgeneralized one of glass the types ("IR-cut"), and ignored the thickness of both glass types.

I am asking.

One should define the exact glass, and thicknesses being used in a stack in order to know what is being transmitted to the camera by the filter stack.

If the glass being used can be calculated and plotted, then a more solid foundation can be referenced, if not then the discussion is nebulous and full of confusion.

 

The UG5 stacks I find most useful for UV+Blue/Green are these:

UG5 (or U-330) 2mm + S8612 2mm. This stack transmits from from 310nm to 530nm, with out of band transmission suppressed below 1E-03.

UG5 (or U-330) 1.5mm + S8612 2mm. This stack transmits from from 310nm to 565nm, with out of band transmission suppressed below 1E-03. Most people seem to prefer this stack for how it looks.

 

Using other types and thicknesses of Red/IR suppression glass, such as S8612, BG39, BG40, BG38... will change the Red and/or IR that transmits above the 1E-03 line.

Some quite beautiful flower photos can be shot with UG5, BG3, BG25... when stacked with even BG38 2mm (and the others), but this is not intended to be bee vision or butterfly vision,

mostly in my mind because most of these extended stack profiles lack a smooth curve, and have a Red/IR 'bump'. Nonetheless artistic.

Link to comment

I am in complete agree with you (over describing stack thicknesses), Steve. I was not offended in the slightest, of you pointing out where I was lacking. Quite the opposite - I am thankful for you pointing that out.

 

In fact, I am the first person to [usually] stress stack thickness in many of my own photos (I even make it a point to embed that info directly into the photo, through text boxes superimposed over the image). So this is not even my contention. I was simply stating (generally speaking), that a red-absent stack should not be labeled a "bug vision" stack, but rather a "bee vision" stack, because not all insects are "red-wavelength blind." That was my chief suggestion further along in the conversation. Yes, the thickness information should be included (that goes without saying, and I admit to this oversight with my posted photos above, since I did it all in a hurry), but that was not what I was talking about at this point along the discussion (concerning my last response to you, and my additional thoughts).

 

Otherwise, that you pointed out my folly in my initial post, I concur. No disagreement. (Here is the info you requested, which I failed to mention: 1.5mm-thick UG-5 and 2mm-thick BG39).

 

Thus, I absolutely agree with you, Steve (over the need to be pedantic with one's use of optical filters, when discussing photo results). I've actually learned much from you, because of my association with you over the years. So, I am grateful for you pointing out my oversight within my initial post. Thank you!

Link to comment

Igor, Thank you.

Your stack using BG39 works fine, however S8612 would be preferred because it

1) transmits UV stronger (the main advantage),

2) cuts off visual at the same place (565nm) and,

3) suppresses Red/IR ever so slightly more below 1E-03 (than the BG39 stack).

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...