View Full Version : Effect of flash on some bird families.

Arthur Grosset
01-12-2008, 02:00 PM
As noise management improves with the latest cameras I am trying to avoid using flash as much as possible on certain families of birds. My experience is mainly in South America but the point may be relevant elsewhere.

Firstly, there is the effect of flash on birds with some iridescence in their plumage - families such as parrots, hummingbirds, manakins, trogons and jacamars. I find that flash significantly distorts the colours of the birds. Does anyone have any tips about how to minimise this distortion?

Secondly, there is the fact that for many families flash spooks the bird so that the bird jumps, raises its wings or takes off. I find this particularly with some tyrannid flycatchers, hummingbirds, manakins and tanagers. There is a thread on this forum which suggests that excessive flash use on hummingbirds might drive the birds away. I don't believe that flash harms the birds (though I have not seen any proper research on this) but I would rather not spook them if I could avoid it. Does anyone have any suggestions?

Alfred Forns
01-12-2008, 02:49 PM
Hi Arthur I think flash ruins more images than anything else !!!! However I do use flash !!

Just after before and after sunrise (same sunset) the flash will cover the beautiful warm light If used just to fill appropriately then is fine I find it most useful for taming harsh light

The birds I work with do not spook/react to flash I'm sure there are some that will One thing you can do while using the flash is firing two exposures in rapid order The first will have the full flash dialed in and the second weaker Would be like bracketing

Peter Hawrylyshyn
01-12-2008, 05:45 PM
arthur -
As there aren't many iridescent birds here, i can only comment on my experience with HB's in the neo-tropics. I have to totally agree that flash can easily change and possibly distort the bird's colors. I prefer not to use flash on non-action shots. It is very easy to over-saturate colors creating a false bright golden/green sheen. Below are 2 photos ( both taken on a 1D3 at 1/50 f8 ISO 800 with 500mm x 1.4 converter) of the same saphire-vented puffleg (left version had manual flash) at Yanacocha in Ecuador. Although the flash improved the image quality, it changed the colors particularly on the chest of the bird. Furthermore, because the color changes aren't blown highlights, you can't pick up the effect easily in your histogram.

As for the birds spoking with raised wings, i think thats due to the infra-red preflash and not the light pulse itself. Most HB species don't react to it . I think some models allow you to turn this feature off in manual mode.

John Cooper
01-12-2008, 07:03 PM
Hi Arthur, My experience of photographing Australian birds over the past 20 years has not convinced me that flash disturbs a bird. But slight forign noises such as shutter & mirror slap often startles a timid species.
As for effecting the birds colour I have only seen the flash improve the natural tones by removing a cold cast as illustarted in Peter's Humming bird examples.

Ian McHenry
01-13-2008, 02:42 AM
Taking pictures of birds in deep forest or bush as we call it in New Zealand can be difficult due to lack of light but wherever possible I prefer not to use flash.
There is a much stronger likelihood of a redeye type result.
I was fortunate with this shot in a heavily wooded aviary.Flitty passerines and thick bush are not an easy combination.
In this case from earlier shots I was pretty sure that photo without flash would not work.
Ian McHenry

Details Pentax K10D and Tamron 70-300 @ 180mm Macro.
ISO 800 1/180 sec F9.5.
Spot metered. On camera flash.
Red-crowned Parakeet.

Arthur Grosset
01-13-2008, 06:44 AM
Thanks for those useful comments. I will certainly experiment with turning off the infra-red pre-flash when I am next lucky enough to be with hummers but wonder if this will not affect the exposure.

To illustrate my original point I attach a composite (I hope that's allowed) of a female Fork-tailed Woodnymph. Both were taken with a 20D and a 400mm f/4.0 DO lens. The image on the left was at f/4.0, 1/800, ISO 400 (Auto -1/3 stop) handheld with no flash. The image on the right was at f/4.0, 1/250, ISO 400 (Manual) handheld with Better Beamer flash.

The left image had some gaussian blur applied to the background; the right image had the eye cloned a bit to get rid of the flash effect and I reduced the saturation but maybe not enough.

Without going into the artistic merits of the images I much prefer the colours / colors of the left hand image and they are much truer to what the human eye sees (which is not necessarily what other birds might see)

Jim Neiger
01-13-2008, 02:52 PM
I am not a flash expert. I actualy try to avoid using flash. Instead I try to use natural light and post processing to make images. I tend to only use flash when it's used as main light, and then only if it's for an image I really want bad. Part of my reasoning for this approach is that I like to use hand held technique almost all the time. Flash is a pain in the butt when using big lenses hand held. Another reason is that I've found use of the shadow highlight tool in PS can produce similar results to fill flash, making it unnecessary in many cases. Now that I have a 1D3, my low light capabilities have improved dramtaically. The 1D3 is just amazing at high ISOs. This means I can photograph birds in shade without flash using hand held techniques and even get enough shutter speed to do flight images and other action. Of course my goal is to make images in sweet natural light as much as possible.

Walt Anderson
01-14-2008, 08:57 AM
The color changes you are noticing are due to differing white balances between the flash and the ambient light. All flashes are made to match noon day light which is 5500d Kelvin. Your ambient light in the forest is cooler {shaded} and has a green color cast. If you want to match this light you need to add a gelatin filter to the flash head. I would suggest a very light green to start with. Check with a pro camera shop for Rosco filters and ask for a sample book. Hope this helps.

Walt Anderson
Visual Echoes Inc.

Arthur Grosset
01-14-2008, 10:10 AM
If you want to match this light you need to add a gelatin filter to the flash head. I would suggest a very light green to start with.

Walt Anderson
Visual Echoes Inc.

Thanks for that suggestion, Walt. I will certainly experiment with this the next time I am photographing in the rain forest. I presume that the filter should be added to the head rather than to the fresnel lens on the Better Beamer.

I have tried in the past to eliminate the effect by reducing the temperature on ARC. It does reduce the effect but not entirely. Would adding a filter in Photoshop give the same effect?

Walt Anderson
01-14-2008, 12:16 PM
The filter would work attached to either location. You need less material when attached to the head. You are trying to match the light from the flash to the ambient light. Using ACR you could do two conversions, one for the ambient light and one to correct the flash exposure then bring them into PS as layers and mask them.

Walt Anderson
Visual Echoes Inc.

Julian Robinson
01-16-2008, 09:29 AM
You are correct about the colour changes - the distortions may be related to structural colouring of plumage and the directionality of the flash, although this does not explain some observed effects. I am yet to find or devise a satisfactory explanation that covers all effects. It is certainly much more than just a matter of colour temperature and it only occurs on some birds, and to differing extents. Some have marked hue shifts and obvious changes in brightness as well as unnatural iridescence. Photographers have postulated that parrot fluorescence might be responsible but such fluorescence is usually limited to small patches while the observed changes cover large plumage areas and most affected birds have no fluorescence at all.

I have an example of obvious hue-shift here where you can see side by side that flash causes colours that simply do not exist in a normally illuminated Crimson Rosella (the greens), and greatly reduces other colours (the purples)...


To forestall possible suggestions, this has nothing to do with the green plumage seen on immature C Rosellas!

It is interesting as to what actually causes the colour changes, or at least observations of when and what. In the following photo you can see that the male Superb Parrot appears practically fluorescent/iridescent under flash - much brighter, but no hue shifts - while the females are almost unaffected...


I haven't found a way of reducing the colour changes produced by flash other than by reducing flash intensity and relying more on daylight as you did. Cross polarisation has been suggested, but would be a pain to implement on the run with moving birds, and would reduce available light significantly.

As it happens this is an area of particular interest to me, so I once conducted some experiments on the extent to which flash spooks birds. I own the 20D (well I did until 10 days ago) which has about the loudest shutter noise of any camera on earth, so it was initially impossible to distinguish between the effect of flash and that of the shutter. In the end I tried firing the flash alone sometimes (each time sacrificing what would undoubtedly have been a fantastic photograph :)) to see what happened. My observations are:

- flash alone will cause some birds to flinch but I have never seen one fly away just because of flash. Other birds are unphased by it. All birds that I tried it on become used to flash after a couple of shots and show no reaction at all after the first few. But a few times birds have flown off after a barrage of flash shots, which may have been because of the flash. (These flash tests had to be tried on a newly arrived or newly seen bird because of this habituation).

- the shutter noise of the 20D on its own will cause some birds to flinch and a significant number to fly off. Again they became habituated generally, although I think it caused ongoing wariness. The 20D is a loud camera.

- the combination of flash and shutter noise was synergistic, in that some birds would flinch or jump badly and were much more likely to fly off. This is reasonable since a two-sense (sight and sound) shock is 10 times (or something) more shocking than a similar level one-sense shock.

I did some research mostly on the web and came to the conclusion from the limited number of papers I found that flash does not damage birds at all, but can alarm some, and does cause loss of night vision in nocturnal birds (but less than spotlighting does!).

There is a fair bit of ill-informed discussion on the matter too. One paper (which I can't find now) cited a test to prove that flash distressed birds. They put a remotely fired flash on a beach near a flock of some kind of shore bird, and when the flash was fired the birds all flew off. Duh! They would equally have flown off if you trod on just one twig at 30 metres, but neither action would have "distressed" them, it's just what birds do when something unusual happens in a hurry.

All that said, I find that flash rarely has any effect on birds at all, mostly their reaction is to turn towards the camera and wonder what the **** that was, in an apparently calm and bird-rational manner.

Lastly I agree with you and Alfred, flash can ruin a lot of bird shots. Despite this I use fill-flash a fair bit simply because it saves many otherwise impossible situations. Without the benefit of a 1D, shadows and highlights for me is just a means of showing off spectacular shadow noise to its greatest advantage. Although, like flash, I do use it a fair bit, in moderation and often in conjunction with noise reduction.

PS I notice that this forum is extraordinarily sensitive to language, it not even allowing the word h*ll (where the * was an 'e')! That seems pretty silly?

Arthur Grosset
01-16-2008, 10:03 AM

Thanks for that very detailed analysis. We seem to be pretty much in agreement about the effect and you are much further down the line when it comes to the cause.

I hadn't thought about shutter noise as a contributory factor but I will bear it in mind. And I'm racking my brain to try to remember who it was who wrote a book about hummingbirds with a detailed analysis of the physics of iridescence. He operated some time ago and used to book an extra seat on planes for his camera equipment. Does anyone recall who this was?

Oh, and I think that **** is fairly easy to interpret!

Julian Robinson
01-16-2008, 10:41 AM
Arthur - I'd be very interested to read the hummingbird iridescence physics if/when you find it!


Peter Hawrylyshyn
01-16-2008, 09:56 PM
Julian - A very informative post.

There are several postings on the web and in publications which explain HB iridescence. It is the result of "Interference" of light waves strengthening or weakening when reflected off a thin film. Iridescent portions of HB feathers are unique in that they contain layers on tiny air-filled cells. The distribution and color of iridescent feathers varies by species. The surfaces of these "air bubbles" selectively reflect some colors stronger than others dependent on the angle of incidence and wavelength. The HB adjusts its feather tips relative to the angle of the sun to create the iridescence. Changes by as small as a few degrees can turn the iridescence on or off.

This method for producing iridescent color is different from the effect of pigmentation found in most bird feathers.

Upon further thinking, a simple explanation might be that when the flash lights up the HB - it will be at a different angle to ambient or natural sun light which will therefore produce a different iridescent pattern. Also the greater the intensity of light in the flash - the greater the iridescent effect. This can be seen in Arthur's two pic's of fork-tailed woodnymph.

Best reference i found was : Iridescent Color of Hummingbird Feathers by CH Greenewalt et al published in the Proceedings of the American Philosophical Society (Vol 104 #3 pg 249-255 in June 1960) .

There's another quick post at: : http://www.mnmicroscopy.org/ProjectMicro/Explorations/Nov1998.html

I'm still unclear whether such "air-filled" feathers are unique to HB's or would also explain the iridescence in parrots??

Put into practice - i find when using flash with HB's , pushing the histogram to the right will exacerbate the flash effect on iridescence. So i try to keep it centred close to the middle and adjust brightness as needed in PS.

Hope this helps

Julian Robinson
01-16-2008, 11:52 PM
Thanks Peter, also informative and I think has finally led me to understand what is going on with birds that change colours under flash! I understand about iridescence and structural colouration, but there has been some confusion in my mind (and that of others I think) as to what the relationship between the two is. Many birds use structural colouration to add to the repertoire of available colours in a non-iridescent way i.e. a consistent colour not dependent on viewing or illumination angle. Blues and greens for example are not normally available as pigmentation, so are produced by structural means.

But iridescence is also structural. The difference being that iridescence has a directional component i.e. what you see is entirely or partly dependent on your viewing angle and the angle of the illumination.

So a Bronzewing is iridescent, but a blue bird is not, but both rely on structural colouration.

My dilemma until today was why and how this could be so.

I could not find a full-text source of your reference article other than by subscription, but fortuitously it did point me to another paper that for me solves the problem!! (I'm excited because this has occupied a small space in the back of my mind for years). See this lovely paper... http://www.matematicalia.net/index.php?option=com_content&task=view&id=59&Itemid=64

It's in Spanish so I'm relying on a Babelfish translation, this paper concludes that non-iridescent structural colouration comes from QUASI-ordered nano-structures (as distinct from iridescence that comes from regularly repeating, ordered, structures). They had to do some fancy Fourier analysis to show this, and only indirectly, but it is the answer we needed. They conclude...

We have made studies of coloration in the skin and pens of numerous species of birds [ 11-17 ], in the skin of mammals (in face weaves and of escroto of primates and two species of marsupiales) [18 (http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=es_en&trurl=http%3a%2f%2fwww.matematicalia.net%2findex.p hp%3foption%3dcom_content%26task%3dview%26id%3d59% 26Itemid%3d64#ref18)], and in the wings of libélulas and 19 (http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=es_en&trurl=http%3a%2f%2fwww.matematicalia.net%2findex.p hp%3foption%3dcom_content%26task%3dview%26id%3d59% 26Itemid%3d64#ref19)butterflies[, 20 (http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=es_en&trurl=http%3a%2f%2fwww.matematicalia.net%2findex.p hp%3foption%3dcom_content%26task%3dview%26id%3d59% 26Itemid%3d64#ref20)], with similar results.

In summary, we can say that the use of analysis of Fourier allows to establish that cuasi-ordered structures can produce noniridiscentes colors by the constructive interference of the coherent dispersion of light waves.

[where pens=plumage I assume].

So now, at least to me(!) it all becomes clear...

-- Many birds (the "dull" ones) use only pigmented colouring - they are generally B&W, brown, yellowish or reddish.

-- Many birds use structural colouring (based on quasi-ordered nanostructures) to add greens, blues and purples to their plumage. These colours are non-directional.

-- A few birds use structural means based on regularly ordered nanostructures to exhibit iridescent colours. These colours will change with angle of view and illumination (so bronzewings etc).

-- AND ... some birds must have both kinds of structural colouring present at the same time, so under uniform (normal) illumination appear a certain set of colours but under directional light (eg from camera flash) a whole different set of colours becomes evident at certain angles, competing with the non-directional colours. Iridescence on top of normal structural colouring.

So Arthur's Hummingbirds and my Superb Parrots must have iridescent colours that are roughly but not quite the same as their non-directional normal colours, so they brighten up and somewhat change what you see when you use a flash. My Crimson Rosellas must have iridescent colours that are completely different from the normal coloration. Also (actually maybe this is more accurate) the Rosellas' iridescence occurs on parts of the plumage that are otherwise very dark so when the iridescence appears it represents all of the colour visible - a new colour independent of the previous dark feathers. In the other birds mentioned, the iridescence appears on plumage that is already bright and coloured so it will only modify the net effect rather than introduce an entirely new colour.

This is great, not only the answer to my long-sought mystery, but also a really nicely worded (as far as you can tell from a machine translation) paper that directly addresses the issue. It it worth reading the last few paragraphs (the ones that follow what I quoted above) of that paper, they have a fine romantic view of it all and compare this phenomenon for the eyes as music is to the ears!

Thanks so much for pointing me to this.

Arthur Grosset
01-17-2008, 04:58 AM
Best reference i found was : Iridescent Color of Hummingbird Feathers by CH Greenewalt et al published in the Proceedings of the American Philosophical Society (Vol 104 #3 pg 249-255 in June 1960) .

Peter, you won't believe that at about 5 am this morning I woke up and said to myself "Greenewalt" who I suddenly remembered was the author of this book I had read many years ago in Brazil. I have now discovered it on Amazon and have ordered a copy. I suspect that the reference you mention is included in the book which was also published in 1960 though there is now a "new" edition in 1991. The title is "Hummingbirds" by Crawford H. Greenewalt. ISBN 978-0486264318. If it adds anything more I will let you know.

I am also extremely grateful to you and Julian for interpreting all this information and putting it into language that I can understand. I think I now realise what is going on with flash on these birds. It sounds as though there is no way to replicate what the human eye sees in an image using flash but, with an understanding of what is going on, we might be able to minimise the flash effect. Or maybe we just wait until we get sensors that function at ISO 6400 without any noise!

Barry Goggin
01-31-2008, 12:22 AM
I know I am coming late to the game but this links up with something I was interested in recently. As noted before, structural coloration allows feathers to show colors e.g. blues that pigments don't show. What was interesting to me was whether it is possible to have a lack of color in an amelanotic feather that has structural coloration e.g. in a leucistic bird.
It turns out that there was a paper in 2006 that addressed this in an amelanotic Stellar's Jay feather. Their conclusion was "The washed-out color of^ the amelanotic jay feather was thus most probably caused by^ the loss of the basal melanin layer, suggesting that melanin^ functions to absorb incoherently scattered white light from^ the feather barb thereby increasing the purity of the color^ produced by the spongy layer." This means that the bird's amelanotic feathers would have a washed out coloration but would not be a pure white. I give the reference below which you can view online. There is a photo of the bird and you can see faint blue color.
So given the importance of the structural coloration, it is probably not surprising that flash would affect the colors exhibited by such birds.