Jump to content
UltravioletPhotography

Thinking about Tones in Simulated Bee Vision


Andrea B.

Recommended Posts

This spring I started experimenting with a Steve-recommended UG5 + S8612 filter stack which passes a lot of UV, some blue and some green. Before that I had not attempted much simulated bee vision (BV) beyond a couple of RGB stacking experiments. But I got into it, as they say, and found it quite interesting to convert & edit the fotos made with the stack.. I have now tried to include a BV frame for each flower I shoot along with my standard UV-signature fotos made with the BaaderU, AndreaU, CopperU and a small collection of Steve-Stacks.

 

Observations:

 

1) UV-Dark Areas

The UG5 + S8612 stack reproduces our reasonable BV colour assumptions for UV-absorbing areas after white balancing in the converter. (I say assumptions because we cannot actually ever know how another creature sees.)

  • Flower colour --> BV colour
  • UV-abs + yellow = -UV+R+G = green or yellow-green
  • UV-abs + white = -UV+R+G+B = cyan
  • UV-abs + blue/violet = -UV+B/-UV+R+B = blue
  • UV-abs + magenta = -UV+R+B = blue, or cyan depending on saturation
  • UV-abs + red/orange = -UV+R/-UV+R+G = ??

However, UV-absorbing areas are not always reproduced with darker tones under the UG5 + S8612 stack. My solution for this problem involves a Photoshop Luminosity layer of the UV-signature over the BV frame. Examples are below.

 

[Note: Before various critics appear, kindly note that I have no idea whether anyone else wants to claim a "first" on stacking a UV-signature as a Luminosity layer over a visible or BV frames. I've been stacking UV-sigs as Lum layers since 2007. I do not particularly care whether I was the first or the ninety-first person to do so. It is so unimportant.]

 

 

2) UV-Bright Areas

I'm not so sure about what colours "should" appear with the UV-reflective areas? I have only a few actual examples so far. Note also that UV-reflective yellow areas on flowers appear much yellower under the UG5+S8612 than do UV-absorbing yellow areas which appear green or yellow-green. And UV-reflective white stays white under the UG5+S8612 after white balancing.

 

Given a combination of reflective colours which includes UV, we are stuck with imagining a colour impossible to humans however natural it is to the Bee or other insect, so why not just go with what appears in the foto? Using a Luminosity layer stack for my BV fotos ensures that the UV-reflective areas stay bright and the UV-absorbing areas stay dark.

  • Flower colour --> BV colour
  • UV-reflect + yellow = +UV+R+G = yellow or yellow-green
  • UV-reflect + white = +UV+R+G+B = white
  • UV-reflect + blue/violet = +UV+B/-UV+R+B = blue
  • UV-reflect + magenta = +UV+R+B = ??
  • UV-reflect + red/orange +UV+R/+UV+R+G = ??

Wild Radish (Raphanus raphanistrum)

For this flower, the UV-dark central bullseye and veins are also obvious under the UG5 + S8612 stack. Note that the UV-reflective yellow portions of the petals appear yellow, not green as do the UV-absorbing yellow portions in other flowers.

From left to right: Visible, UV, BV. Expand your browser to 1200 pixels width to see side-by-side.

600_8304pn_400px.jpg600_8309pf_400px.jpg600_8319pn_400px.jpg

 

 

Oxeye Daisy (Leucanthemum vulgare)

But for this flower, the UV-dark rays are very bright under the UG5 + S8612 stack.

From left to right: Visible, UV, BV. Expand your browser to 1200 pixels width to see side-by-side.

leucanthemumVulgare_visFlash_20150725shoreCottageME_36386pn_400px.jpgleucanthemumVulgare_uvBaadSB14_20150725shoreCottageME_36428_400px.jpgleucanthemumVulgare_ug5s8612Sun_20150725shoreCottageME_36443pf_400px.jpg

 

 

 

Oxeye Daisy: UV frame as Luminosity layer over BV frame.

My solution for restoration of the "proper" tones in the Oxeye Daisy BV foto is to stack the UV frame as a Luminosity layer over the BV frame. Now we have UV-dark areas properly represented and still have the cyan and green of BV - now become a dark cyan and a dark green. (Of course, you can also put the BV frame as a Colour layer over the UV frame to get the same result.)

leucanthemumVulgare_ug5s8612Flash_20150725shoreCottageME_36428-69pnLuminos.jpg

 

 

 

Wild Radish: UV frame as Luminosity layer over BV frame.

For the wild radish, the Luminosity stack is not really needed, but here is what it looks like anyway.

 

This stack also illustrates a general problem with stacking - the frames do not always match up. Some painful retouching of petal edges was required for this one. And sharpness suffers because of the mismatched caused by the flowers' motion between frames due to some breezy conditions. It is best to shoot indoors if careful frame matching is required.

 

600_8319pn01Luminos.jpg

 

 

Small Warning: From my reading I have noted that there is not a consensus about how Bees (or other insects) interpret tones. The Luminosity stack illustrated above which preserves UV tones in the BV frame, may be more for human benefit than a it is a good simulation of Bee Vision. I rather like the idea of Luminosity stacks for BV, but I remain open to further refinements of our photographic attempts at BV.

Link to comment

I went out to shoot various colour flowers this afternoon in order to finish the colour correlations I started listing above.

  • Flower colour --> BV colour

  • UV-abs + yellow = -UV+R+G = green or yellow-green
  • UV-abs + white = -UV+R+G+B = cyan
  • UV-abs + blue/violet = -UV+B/-UV+R+B = blue
  • UV-abs + magenta = -UV+R+B = blue, or cyan depending on saturation
  • UV-abs + red/orange = -UV+R/-UV+R+G = ??

  • UV-reflect + yellow = +UV+R+G = yellow or yellow-green
  • UV-reflect + white = +UV+R+G+B = white
  • UV-reflect + blue/violet = +UV+B/-UV+R+B = blue
  • UV-reflect + magenta = +UV+R+B = ??
  • UV-reflect + red/orange = +UV+R/+UV+R+G = ??

Example: UV-absorbing + visible Magenta with low saturation

Here is a foto of a small Echinacea forest. During high summer the flowers wear crowns of bees. I sampled the Visible magenta rays and found that they are not highly saturated in this particular cultivar. One such sample read RGB = (237,165, 201). Thus it was no surprise that under the UG5 + S8612 stack, the rays were cyan coloured. The Echinacea rays are UV-absorbing.

 

Equipment: D600-broadband + 105/4.5 UV-Nikkor

 

Visible Light [f/11 for 1/500" @ ISO-400 in Sunlight with Baader UVIR-Block Filter]

echinaceaPurpurea_visSun_20150804charlotteRhoadesSwhME_.jpg

 

 

Bee Vision Simulation [f/11 for 1/13" @ ISO-400 in Sunlight with UG5 + S8612 Stack]

echinaceaPurpurea_ug5s8612Sun_20150804charlotteRhoadesSwhME_.jpg

 

 

Ultraviolet Light [f/8 for 1/30" @ ISO-400 with SB-14 UV-modified Flash and Baader UV-Pass Filter]

echinaceaPurpurea_uvBaadSB14_20150804charlotteRhoadesSwhME_.jpg

 

 

Extracts, unresized. There is a particularly cute, small, orange and yellow banded Bumblebee in this area.

echinaceaPurpurea_visSun_20150804charlotteRhoadesSwhME_01.jpg

 

In UV, the bee fur is a bit reflective.

echinaceaPurpurea_uvBaadSB14_20150804charlotteRhoadesSwhME_01.jpg

Link to comment

Next, here is a Mexican Zinnia (Zinnia haageana) which is red with a yellow-orange squiggle on the tips. The rays are UV-absorbing except for the tips which are UV-reflective. However, under the simulated Bee Vision UG5+S8612 stack, I got a surprise. The UV-absorbing red rays are a dark red, the yellow squiggle is green and the UV-reflective red ray tip becomes a sort of pinkish-orange.

 

I've made shots of my Labsphere white standards under the UG5+S8612 stack, measured the white balance and created a Preset in Photo Ninja for converting fotos made with this stack. So I did everything "by the book" and still got an unexpected dark red and an orange in the simulated Bee Vision. I thought we would only get greens, yellow-greens, blues and cyans ??

 

Does anyone have any suggestions about this anomaly?

 

Example: UV-absorbing + visible Red (rays excluding tips) and UV-reflecting + visible Red (tips)

 

Equipment: D600-broadband + 105/4.5 UV-Nikkor

 

Visible Light [f/11 for 1/200" @ ISO-200 in Sunlight with Baader UVIR-Block Filter]

Samples taken on the red rays are like RGB = (163,0,8) in the darker areas and (239,0,1) in the lighter areas.

The yellowish squiggle on the ray tip has samples like RGB = (255,175,0).

zinniaHaageana1_visSun_20150804charlotteRhoadesSwhME_38048pn01.jpg

 

Bee Vision Simulation [f/11 for 1/1.3" @ ISO-200 in Sunlight with UG5 + S8612 Stack]

zinniaHaageana1_ug5s8612Sun_20150804charlotteRhoadesSwhME_38056pn01.jpg

 

Ultraviolet Light [f/11 for 1/20" @ ISO-800 with SB-14 UV-modified Flash and Baader UV-Pass Filter]

zinniaHaageana1_uvBaadSun_20150804charlotteRhoadesSwhME_38068pn01.jpg

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...