What is really seen?

[Damon]

I have a visual question that's been bugging me but I am sure someone here will know the answer to it quickly. If I am talking in circles then let me know. I have done it in the past on this site when diving to this stuff. :)

 

For this question we could use any animal that sees into the UV, but for now let's use a bird that can see both Visible and UV. This bird thus looks at a dandelion during normal daylight hours.

What does it see? Are you sure about that?

 

I have both Vis and UV shots of dandelions myself so I just stacked them together and they fused into one. Try it with any of your shots and see what happens. I am not convinced this is what the bird is seeing.

 

Combining images is not the same as what the bird sees because the UV shot I have captured was without visible light entering the lens and the visible shot was captured without UV light entering the lens. My dandelion image was not captured with Vis and UV at the same time like a birds vision would work.

 

Can anyone here with all these filters knocking about, shoot UV and visible at the same time so we can see what the dandelion really looks like? From what I have experienced with taking UV images is that if Visible light is leaking into your setup (let alone full sun) your UV shot is no good-ie. you don't see the UV signature on the dandelion.

 

What happens when you take a picture of a dandelion during the day with a full spectrum camera when using a lens that let's in only Vis and UV? do you see the UV signature or no? I don't know as I don't own one.

 

 

So how then can this bird (or any animal that can see into the UV) see the UV signature of anything with daylight always overpowering it?

 

 

-Damon

[Andy Perrin]

I’m no expert in this but I gather that birds are tetrachrominants and their eyes have four color channels essentially. So the UV is not overlaid on top of the visible signal in the way you are describing.

[Cadmium]

I am no expert either, but here are a few thoughts.

The best filter for mixing UV with Visual would be the UG5/U-330, stacked with S612, but that doesn't reach into red, so it is not really bird range,

but I am using that example because the profile of that transmission is tapered to where the UV transmission is stronger, and the visual is weaker, which balanced out in a way where the UV is not overpowered by the visual.

Draw back is two,

1) it is limited to the UV+Blue+Green range, unless you use a different BG glass which extended the rang into red, but not with the same gradual couture, it would have more of a red peak.

2) The UV part of the mix is not monochrome, thus the color mix is not UV+Blue+Green(+Red), but instead it something more like (UV false colors Green+Yellow+Violet) + (Visual Blue+Green (+Red)).

In order to make a more true mix of UV+Visual, you would need to desaturate the UV then mix it with the Visual range you want to mix with,

and to do that with one filter would require some pretty fancy post processing that would juggle channels around to add/subtract, and and so on.

 

There are plenty of other filters that can be sued for UV+Visible, like stacks using BG3, BG25, even some UG/U stacks, they so the same thing, mix UV with visual, but not the same tapered profile as the UG5/U-330,

regardless, they all show UV patterns not seen with a visual only range filter.

[DaveO]

The simple answer to your first question, is that nobody REALLY KNOWS what birds or bees see. Very skilled researchers have found which wavelengths of light are visible to birds and bees but how they process that information is a "Hard Problem"

 

Here's a reference to Birds and it also contains reference to earlier bee work.

 

Burd M, Stayton CT, Shrestha M, Dyer AG. 2014

Distinctive convergence in Australian floral colours seen through the eyes of Australian birds.

Proc. R. Soc. B 281: 20132862.

http://dx.doi.org/10.../rspb.2013.2862

 

Try this as well:

 

More than colour attraction: behavioural functions of flower patterns

Natalie Hempel de Ibarra, Keri V Langridge and Misha Vorobyev.

 

http://dx.doi.org/10...ois.2015.09.005

 

 

 

Horridge, G. Adrian. What does the honeybee see and how do we know?

 

http://epress.anu.ed...e_citation.html

 

 

Dave

[Damon]

Thanks guys!

I would like to see a flower taken where Full Visible and Full UV hit the sensor. Don't you just end up with a visible image?

Dave, I have read pieces of 2 of those before and they are indeed fascinating. The last link is a 404 page not found deal.

 

The question I wanted to be answered though was how can any UV capable creature see a UV signature in full sunlight? Sunlight seems to overpower the UV markings.

 

My old Gazania UV pics were amazing to me - all that deep black around the middle. If I was live viewing that Gazania UV deep black with my camera and then let some light in - poof the UV goes away. So again, how is anything seeing any of these UV signatures during daylight?

 

-Damon

[DaveO]

Sorry about the missing link, try this:

 

https://press.anu.edu.au/publications/what-does-honeybee-see-and-how-do-we-know

[Andrea B.]

Damon, I've made posters which will help answer your questions. The're around here somewhere. So let me go dig them out.

 

Posted below in Post #9.

[Cadmium]

Thanks guys!

I would like to see a flower taken where Full Visible and Full UV hit the sensor. Don't you just end up with a visible image?

Dave, I have read pieces of 2 of those before and they are indeed fascinating. The last link is a 404 page not found deal.

 

The question I wanted to be answered though was how can any UV capable creature see a UV signature in full sunlight? Sunlight seems to overpower the UV markings.

 

My old Gazania UV pics were amazing to me - all that deep black around the middle. If I was live viewing that Gazania UV deep black with my camera and then let some light in - poof the UV goes away. So again, how is anything seeing any of these UV signatures during daylight?

 

-Damon

 

You can take a photo that contains full UVA + full (human) Visual range, Schott S8612, BG40, and BG38 all do that,

however the Visual range overpowers the UVA and you will see no hint in most situations of the UVA in the photos, you will just see a visual photo.

So you would somehow need to partly suppress the visual range of the UVA+RGB range.

However, even if you had a filter that transmits a higher % UVA range combined with a lower % RGB range, you still have the camera's Bayer filters mixing the UVA false colors with the RGB colors.

The UG5/U-330 + S8612 stack does a pretty good accidental job of the dual transmission % filter idea, depending on the type of suppression in the stack, albeit not exactly normalized for the visual range.

 

The real problem remains the UVA false colors, you can't separate the UVA range from the Visual range given the Bayer filters.

Thus your colors are mixed inaccurately: (UV false colors Green+Yellow+Violet) + (Visual Blue+Green (+Red)).

I would have to think about a Bayer-less camera... hmm.

However, there again, the birds and bees probably don't see things monochomatically,

and it's hard for me to imagine the math on that scenario, but the monochrome version might be an inaccurate mix also.

And, like Dave says....Nobody knows exactly how or what they see (except an illusive team of Birds and Bees that meet together secretly every 137 years ;) ).

 

Below is a comparison between Visual Human, Simulated Bird, and Simulated Bee (left to right).

Left: The Human Visual shows no UV pattern, just our Visual Blue, Green, and Red.

Center: Bird shows the UV pattern, with our Visual Blue, Green, and Red.

Right: Bee shows the UV pattern, with our Visual Blue, and Green, but no Red.

 

[Andrea B.]

Damon, I found one of my posters. It's over-oversimplified, but makes the point.

 

On the top left is a yellow dandelion the way we humans see it in the sun. The colors/wavelengths the dandelion reflects are shown on the top right.

(We can quibble elsewhere about whether UV is considered a "color" or not. Depends on who is looking -- human or bee!)

 

The dandelion's center absorbs Blue and absorbs UV. Alternately, we can say the center reflects Red and reflects Green. (Oversimplified! The center actually reflects yellow, but we model color using R, G and B.)

• To us humans the dandelion center thus appears Yellow.
  
• To the bee the center appears to be Green because the bee has no visual receptors for detection of reflected Red. (Oversimplified! I am granting the bee the capability to detect greenish-yellows.)
  
• To the camera which has been converted and rigged to see only reflected UV, the dandelion center is dark because it is reflecting no UV.
  

The dandelion's petals absorb only Blue, i.e., reflect Red, Green and UV.

• Again, to us humans the dandelion petals are Yellow because we have no visual receptors for detection of reflected UV.
  
• To the bee, however, the petals appear to be a color. What color? Well a "bee color" which for which we have no name, but which is sometimes called "UV-green" in the bee vision literature.
  
• To the camera which has been converted and rigged to see only reflected UV, the dandelion petals are bright false colour because the camera is recording reflected UV from those petals.
  

 

That's all there is to it - human vision, bee vision and converted camera vision. We sometimes get too casual about the fact that converted camera vision is a single channel thing and that our UV photos do not depict the entirety of what the bee sees.

 

Bee vision cannot be simulated properly even when using a camera with specialized UV+Blue+Green filters because there is no way to depict in RGB triplets any color involving reflected UV. In the poster above I had to use a light green and dark green to indicate reflected green+UV and reflected green respectively.

 

Here is another dandelion chart which was used by the University of Nebraska Extension Center a couple of years ago at a conference about pollinators. I had to brighten the center of the dandelion because the UV+B+G filter makes it too dark and that center is not dark to the bee. I also pushed the petal false colour towards yellow to stress the fact that the dandelion is bi-coloured to a bee. (But remember yellow petals are not what the beezes seezes. They are seeing UV+green petals which we have no proper way to represent.)

 

[Andrea B.]

I know several of you are a bit tired of these posters as I seem to haul them out at regular intervals. But they do make the point about how subjects are seen by humans vs. bees vs. converted cameras. Colors made from (UV, B, G) triplets are not easy to represent by (B, G, R) triplets. Yeah, duh! Lol.

 

BTW, in your own work, you are certainly free to represent a (UV, B, G) palette any way you like - as long as it is consistent. :D We've had many lovely representions of flowers in bee vision here on UVP.

[Andrea B.]

SHORT VERSION: Bees do not see UV signatures. Only a converted camera + UV-pass filtered lens can "see" a UV signature. Chart tells you why.

 

 

I should just post my SHORT VERSIONS of everything. Save a few electrons.

Hmm.......

[Cadmium]

Hey stranger... nice to see you. :)

 

And here is a UV+Blue+Green using (desaturated UV) = Blue + (Visual Blue) = Green + (Visual Green) = Red, RGB = GBU, basically transposed color, like that of EIR.

(using two shots however, and this is a 3 color Bee stack, not a 4 color Bird stack, and if this is how bees see, then I want what they are having ;))

[Andrea B.]

So pretty!

 

Looks like Valentine Rudbeckias. "-)

[Damon]

Thanks Cadmium for your thoughts and work with this. I had a feeling you would grind away at this and come up with some good stuff. I like your direct line of thinking.

 

Andrea, thanks to you of course as well. I read and re-read you info. That's a good deal of interesting information in a succinct format. It looks as if it only answers my query only partly though.

 

If you don't mind I am going to continue thinking about this. I will try my best not to ask questions that you have or anyone else have already answered here or elsewhere on the site if at all possible.

 

 

[Damon]

OK Andrea, the way I have been thinking I will take this bit by bit. :D

I liked the long version.

 

The dandelion's center absorbs Blue and absorbs UV. Alternately, we can say the center reflects Red and reflects Green. (Oversimplified! The center actually reflects yellow, but we model color using R, G and B.)

So the center in reality really reflects yellow because Red + green = yellow

• To us humans the dandelion center thus appears Yellow. Red + green = yellow
  
• To the bee the center appears to be Green because the bee has no visual receptors for detection of reflected Red. (Oversimplified! I am granting the bee the capability to detect greenish-yellows.)
  

Since bee can’t see red (but the red doesn't go away, it’s just not red) then it can’t see yellow either because there now there is no red in the red + green = yellow deal. It can only see whatever false color of the red + green turns out to be. So overall the bee sees the combination of absorbed UV, UV false red + reflected green + UV false green

[Andrea B.]

I know those posters are very oversimplified. I'm never sure whether I should use them or not.

 

To us humans the dandelion center thus appears Yellow. Red + green = yellow

 

So the center in reality really reflects yellow because Red + green = yellow

 

Yes, but our dandelion on the poster is just a model, maybe a rather bad one? Yellow wavelengths are reflected by real dandelions, and those yellow wavelengths are not R+G. Reflected yellow = absorbed blue.(**)

 

Some Bee Vision theorists grant some yellow detection to the bee using the long right-hand shoulder of its green receptor. (Had I a bit of time, I'd try to find that reference. Sorry!)

 

If we were to deny the Bee any yellow vision at all via its green receptor (as some do),

then to the Bee the dandelion would have a dark center (call it bee black or bee dark grey) and petals which reflected UV. And, to repeat, reflected UV is detectable to a Bee. So the Bee sees the dandelion in a similar manner to the light green & dark green example on the poster.

 

However....., Bee Vision is so very complex(*). As photographers we naturally want to try to illustrate how Bees see using colors and light/dark contrasts. But that does not even begin to model Bee Vision.

 

Dig into some of the Bee Vision papers. They are all quite fascinating. I've worked through Horridge (...not that I could pass an exam on my reading at this point!!...) and a couple of others.

 

******

 

All that said, I love making photos with the various UV+B+G filters as an approximation of how Bees see the world.

 

 

*****

(*) Bees incorporate edge changes/reflections when detecting the achromatic colors, black/white/grey. They don't just "compare" lights/darks.

 

(**) Perhaps I should try to make a poster using a Blue/Yellow axis. Reflected yellow = absorbed blue + absorbed UV. This would be better because it avoids the oversimplification above.

 

****

 

You can always ask all the questions you want to ask. I don't mind. But I want to be clear that I'm not a Bee Vision authority!! Any corrections are always welcomed.

[Andrea B.]

Horridge: Anti-intuitive Bee Vision

page 160.

 

 

 

 

The neuron properties are more suitable to detect line-labelled modulation than a tonic, maintained photon flux. None of that seems to be adapted to the discrimination of colour as a palette of persistent sensations. Moreover, within the optic lobe there is no sign of the colour triangle, or of neurons that could discriminate between colours.

 

Horridge is saying here that bee brains make use of changes in the signal intensity which reaches their eyes. The modulations of a green signal tell the bee something different that do changes in the blue signals, for example.

 

Part of that signal modulation happens because bees fly and winds blow. "-)

 

Horridge remarks somewhere that many older experiments with bee vision were misinterpreted because it was not realized that the bee was detecting changes in signal intensities rather than making use of a trichromatic color palette.

[Andrea B.]

This might work better as a model (although still oversimplified). :D

The "flower" here is not labeled but is intended to be something like a dandelion or any yellow flower having a UV absorbing center.

I've often used a mottled CMY palette to model reflected UV. Fun!

 

[Damon]

Let me sink my head on this a bit.

I thought you said you were going to make this easier to understand haha. Just kidding! I appreciate such effort on your part. It's very nice.

It would be great to put a nano-camera behind a set of bee eyes and look at some stuff. Of course there is no bee brain to decipher it but I still want to do it. There was that group that looked at the mammal eyeballs remember? http://rspb.royalsoc.../20132995#sec-4

 

As an aside--My wife and I have documented every single flowering plant in Southern NJ via visible light pictures and descriptions etc and were intending to make a field guide--aka you see it blooming you can ID it.

The top 3 colors as far as abundance are from high to low:

1.White

2.Yellow

3.Pink

These must be the "colors" that are available to the widest range of insects/pollinators I am guessing.

 

Thanks again

[DaveO]

Damon, here's an attempt to reconstruct what it might be like to be a bee with a compound eye:

 

Dyer, Adrian G.; Howard, Scarlett R.; and Garcia, Jair E., Through the Eyes of a Bee: Seeing the

World as a Whole, Animal Studies Journal, 5(1), 2016, 97-109.

Available at: http://ro.uow.edu.au/asj/vol5/iss1/7

[Damon]

Cool! Thanks Dave!

[rfcurry]

Damon,

 

The people who would have the best opportunity to perceive what birds do is those humans that are aphakic - lacking the lens or the yellow pigment of the lens - of the eye. Most aphakics register UV as a pale blue Claude Monet's use of more blue on his water lilies after his cataract surgery is believed to be because he saw more blue in the UV. Human children begin life without the yellow pigment in the lens that blocks the UV in adults, so they have the pathway for interpretation of UV.

 

Of course, at dawn or dusk, when the UV contributes a higher percentage of our sunlight, the amount of UV in our visual palette will increase. Both the blue cones and the rods of the eyes, in some animals tested. are more sensitive to UV than to blue light. Thus, some crepuscular feeders, e.g., white-tailed deer, are capable of UV vision (and their white tails, or flags, are easy for other deer to follow when escaping from predators).

 

Sorry, I blather on..

[Andrea B.]

I've never been able to figure out why high UV would be seen as pale blue, when spectral violet is definitely not blue? I would think that UV would be some variation on spectral violet?? Guess I'm just curious...