Jump to content
UltravioletPhotography

Convallaria majalis [Lily of the Valley]


Nico

Recommended Posts

Chalwatzis, N. 2013. Convallaria majalis L. (Asparagaceae) Lily of the Valley. Flowers photographed in visible and ultraviolet light, including simulated bee colours. https://www.ultravioletphotography.com/content/index.php/topic/308-convallaria-majalis-lily-of-the-valley/

 

Convallaria majalis

DE: Maiglöckchen

EN: Lily of the Valley

 

Flowers were photographed near Bensheim, Germany, 24 May, 2013.

 

All images were taken with a broadband-modified Panasonic Lumix G1 and the EL-Nikkor 80mm/f5.6 at f8.

 

Visible light image with IR-neutralisation-filter ND (Optic Makario), ISO 100, 1/40s.

post-14-0-52284700-1369602078.jpg

image reference: NCH_P1080761130524

 

UV-image, Baader U-filter 2”, ISO 400, 8s, sunlight

post-14-0-85750000-1369602080.jpg

image reference: NCH_P1080762130524

 

UV-image, Baader U-filter 2”, ISO 400, 8s, sunlight

post-14-0-90236500-1369602082.jpg

image reference: NCH_P1080771130524

 

“simulated bee-colours”: composite image of the UV image (-> blue) (greescale conversion) and the blue (-> green) and green (-> red) channel of the visible light image.

The goal of this is to mimic the spectrum that bees and other insect-pollinators can see.

post-14-0-72571800-1369602075.jpg

image reference:NCH_P1080762_RGB_UV_bw_II130526

 

 

While in the visible light spectrum these flowers appear evenly white, the UV-reflexion is stronger in the inside compared to the outside. The latter is almost UV-dark. However, the UV-channel has a lower intensity than the blue and green channels, which results in a bee-yellow.

 

[published: 26 May, 2013]

Link to comment

I like to elaborate this statement "the UV-channel has a lower intensity than the blue and green channels, which results in a bee-yellow."

 

From a purely photographic perspective it might hold water although implicit assumptions on a flat spectral response of the sensor is employed.

 

However, seen in a biological context, we have to know whether the visiting pollinator responds similar to each channel or spectral range. This would entail weights associated with each colour stimulus are equal. Were they not equal, the outcome in terms of false-colour interpretation may differ significantly. That "bee-yellow" might turn into a "bee-blue" instead..

Link to comment

I agree that there are implicit assumptions. However, I’m certainly not assuming an even spectral response of the sensor. This would be true if both images were taken with the same aperture, shutter speed and sensitivity:

What I did is (aperture constantly at f8):

 

visible light: 1/40s at ISO 100

UV spectrum (Baader): 8s (!) at ISO 400

 

So we have roughly 13 f-stops difference, which means 213 times difference. Now, even if we account for a more narrow spectrum (roughly 80 nm (UV) compared to 300 nm (400-700)) and the difference in sensor sensitivity (that I cannot quantify) and some light absorbed by the filter, I thinks it’s unlikely that UV is underrepresented in my composite image. Maybe it’s even the other way round. In order to answer that I thinks we would require much more data about the physical and biological response of the insect eye that I do not have.

I’m aware of these limitations but I still find it sensible to look at the entire spectrum, even if it is not easily quantifiable. I also find your induced fluorescence very interesting. Do you have any hints how this could contribute to flower-pollinator communication. For our eyes it becomes visible only in the dark, since the intensity is below daylight …

Link to comment

There are literature references stating bees are many times more sensitive to UV than other spectral bands (off my head, I have to search for the actual references), so we may use a selective weighting for the UV band in these composite false-colour emulations. Meaning we probably should boost the UV channel quite a lot compared to the others, which was my point. Another point to keep in mind that the UV image upon which the compoite is based should be acquired separately as an in-camera b/w capture as this represents most closely, in terms of luminance, what the camera actually has recorded in the UV band.

 

It also is a fact our cameras roll off their response into the UV band. Ideally a flat sensor (plus a UV-optimised lens with +- 70% UV-transmission through the lens) should be down about 6 stops in UV (2-3% of daylight) instead of the usual 8 to 11 or so found. An increase of 13 stops sounds a little pessimistic, though, and may point to the lens attenuating UV itself. We are never able to get the true response curve from the manufacturer of the cameras (I have asked, more than once, Nikon representatives, but always get vague answers). However, compare a UV model against a normal version of an astronomical camera and you get an impression of how our cameras suffer from a reduction of their UV response:

 

http://www.ccd.com/images/alta_f42uv_sens.gif

 

Source: www.ccd.com/images/alta_f42uv_sens.gif

Link to comment

Thanks, Bjørn for the information.

I think there are many aspects to consider: E.g. the distribution of the different wavelengths in sunlight, which reaches the surface of the earth.

 

I have found this for example:

post-14-0-96752500-1369682256.jpg

source: http://www.advanceda...12/10/aafeature

 

In any case, the 13 stops are a rather extreme example with the pretty low reflecting Convallaria flower. Other flowers that are UV-bright require shorter shutter speeds, sometime below 1s. Geranium sanguineum required just 0,6 s, so we have 3-4 stops difference less.

What I can estimate with my gear (as with any other UV equipment) is the relation of the UV reflection of a flower compared to the background reflection. As you have also described, some flowers are significantly brighter (than the background) and in others like Convallaria, are not. So if flowers are found to be (relatively) UV-dark, it is save to propose that the UV-reflection doesn’t contribute to an overall brighter appearance (obviously).

Anyhow, I would be interested what difference you find with a dedicated UV-lens, if you compare the required shutter speed under the same situations with the filter that transmits the visible spectrum on the one hand, and with the Baader-U 2” filter on the other.

… I have the feeling that my desire for a “real” UV lens might be growing. On the other hand a good UV-flash is probably the investment that should come first.

Link to comment

I have measured similar spectra of incident sunlight. This is also essentially the similar output spectrum one gets from an uncoated Xenon flash tube.

 

One might say that a pollinator with enhanced UV sensitivity "corrects" for the drop in incident irradiance taking place around and below 400 nm. In that interpretation the hypothetical pollinator responding to UV, B, and G bands will "see" a white, ie. flat energy spectrum.

 

I did run a comprehensive test of UV vs visible light exposures using two different Nikon cameras and two UV-Nikkor lenses (to get concurrent exposures) over at nikongear.com a few years ago. See http://www.fotozones.com...-disc-shootout/ for additional details. It is documented there that the UV exposure has a very broad "acceptable" range, thus, describing the offset against visible light is not possible except in a broader sense. Opening 8-11 stops re visible light should bring the UV exposure into a useful range.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...