iridescent colours

[nitrus]

I was photographing humming birds in Cost Rica where there was a bird the violet sabrewing with a very intense purple colour but when the photo appeared in the screen this colour had turned bright blue. we were using quantum flashes. Does anyone know why this happens?

[jiro]

Did you check the White Balance setting on your camera? Try to set it to flash and see if that works.

[nitrus]

it was set on flash and moreover we tried with different white balances and all change the hue from the actual purple to blue

[Geoff F]

Yes, purple is one of the tricky colours. Every year I try to photograph iris flowers, particularly the early varieties, and end up ditching nearly everything.

The only suggestion I can make is to set a Custom White Balance beforehand and try to avoid bright sunshine.

I have made some improvements through editing my photos but, mostly, I never really understood what I did.

[Jim B.]

I looked at some images of a Violet Sabrewing.Beautiful bird!Many of the images showed more blue than purple depending on the angle of the light.The iridescent feathers might be part of the problem?

[Snarkbyte]

I agree with Jim.

The flash might well have changed the incidence angle of the major light source, especially if the flash was close to (or mounted on) the camera, and no reflector was used. Since the iridescent color depends on angle of incidence, a small light source like a flash can narrow the color band that you see from a particular perspective. I looked at some other photos on this bird on the web, and I've seen several where the blue iridescent color is visible on some parts of the bird, apparently from the same effect.

See this for example. Note how the bright blue areas surround or are near areas of specular reflection. Diffuse reflection from a broader range of angles produce the dominant purple color.

[nitrus]

I did some more investigation of this problem and found that the images come up in the finder (mac) as violet /purple but as soon as processing in a raw convertor (Bridge) starts they convert to intense blue. This happens in Adobe Bridge and in Capture One. I don't think it is a problem with the flash - although we used 4 flashes all off camera of course and triggered with cords, or the camera - when seen initially the colour is acceptable. By the way thanks for taking an interest in this.

[Frankie]

Monteverde? I had the same thing happen, but I wasn't using a flash. I have the same issue with shooting the Blue Morpho butterfly, so I'm convinced it's the iridescence that causes the problems, especially since it all depends on how the light hits it. The Blue Morpho's inner wings are actually brown until the light hits it the right way and it appears blue.
I don't know the answer to photographing iridescence, but would be interesting to learn better ways to capture the brilliant colors.

[nickjohnson]

Hi Angela,
I'm not a Mac user so can only answer in general terms. The clue here is that you report different colour rendition depending upon the software app your using. E.G the viewer is OK but the others are messing up. Might be that the default colour space for the RAW converters – and the way that colour space is being displayed – is causing a colour shift. Have you / can you change the colour space setting ? Even very good (and expensive) monitors have trouble showing all the colours in quite small colour spaces – E.G. Adobe RGB. So any colour that is outside the display capability is mapped or translated onto a colour that can be displayed. Most of the time this happens completely unnoticed by us – it just works. However, very occasionally the mapping gives rise to strange results. HTH

Regards,

Nick.

[Steaphany]

Many Birds feather coloration is actually determined by the wave phenomena of light. Parrots, with their brilliant coloring actually have a brown pigmentation, ever see a Parrot while their feathers were wet ? When a Parrot's feather's dry, their coloring reappears.

When dealing with iridescence it becomes even more tricky as the effects from angles of incoming and reflected light become more pronounced. This is why iridescence usually has a color gradient associated with the character of the color. Another aspect is that the reflected light from an iridescence subject needs to be captured as if it was a light source, not a passive reflecting surface.

Another aspect that you need to consider is how your camera works with polarized light. This is the reason just about all polarizing filters available have a quarter wave retarder layer on the camera side, it converts the linearly polarized light to circularly polarized light. Even with the distance being close to moderate, add a CPL filter and see how that works out, remember that you'll have to adjust the filter to maximize the iridescent reflection, not attenuate it.

[techmind]

All the above points are relevant, but there's a further issue.

Iridescent colours can potentially be almost spectrally pure (ie a single visible wavelength of light) and as such could lay significantly outside the colour-gamut of your camera or working space. Morphing the colours into your gamut can lead to colour distortion ... but that's the easy bit.

For all the effect that goes into maintaining colour fidelity once the signal enters the processing chain from the CCD ... the biggest weakness in the imaging chain is usually the physical colour filters on the CCD (or CMOS) sensor itself. The RGB pixels on the sensor do not have the same spectral response as the sensors in the eye (you can get unweildy research cameras eg a 'video photometer' for £40k which do have accurate filters, but this is not practical for normal photography!), and therefore the colours will be interpreted subtly differently from the eye - and no amount of mathematical jiggery-pokery can fully recover that as in condensing a full spectral distribution down to three channels information is irretrievably lost. In practice, under normal conditions the colours are pretty good, but near-monochromatic light can reveals problems. People who try to take photographs of "rainbow" light spectra from prisms or other scientific equipment always come across this issue.

The eye colour responses are characterised by XYZ colour-matching functions (loosely equivalent to RGB)... but the X function has an additional sensitivity in the violet (this explains why violet (shorter wavelength than blue) can be 'faked' using a computer RGB display screen by adding a little bit of red into a predominantly blue RGB colour mix). If the camera RGB filter doesn't have a some blue 'leakage' in the red channel (they don't normally) then it can only see deep violets as blue.

If you try taking a photo of an energised UV 'blacklight' tube you are likely to find that comes out blue in the photo while it's violet to the naked eye.

See also: http://techmind.org/colour/