Colour Calibration

[Pete Eeles]

Hi all - a question for all of the "proper" photographers out there

Over the autumn / winter, I'm going to be visiting various museums and want to get photos of specimens that are "faithful" in terms of their colour (in particular, the Mountain Ringlet specimens from Ireland). Now - when Mark Colvin and I visited the Oxford University Museum of Natural History, there is a calibrated setup and the museum also kindly took the photos for us - what more could you ask for?!

However, other museums don't have such facilities, and I'll need to take photos myself. My question is - what is the best way of ensuring that the colours of specimens I take are "true" when I make any adjustments when post-processing? I was thinking of taking a photo of a colour chart / grey-scale chart and using that as my reference. Hope that makes sense!

Thanks,

- Pete

[dilettante]

Does your camera have a custom white balance setting? If so, then you can calibrate that (usually using an 18% grey card) and use it for subsequent shots of the butterflies. Recalibrate as necessary if the lighting is not consistent.

If you don't have a custom white balance, then your options are limited to post-processing, and that will depend on what software you're using, but the starting point would be to take a photo of a colour-neutral subject (such as the 18% grey card) under the lighting. Ideally you'd include a sample of that card in every photo, but depending on the software you may be able to automatically apply the same correction to multiple photos, so only need one photo with the grey card (assuming the lighting is consistent, which it probably won't be!).

Also it's probably not possible to completely defined 'faithful' colours. Artificial lights probably have incomplete spectra compared to pure white light. Your idea of faithful may be as it appears on a sunny day but that will be different to diffuse daylight to evening light etc. And that's before we take into account perceptual aspects and how the brain interprets colours under different lights. And if there's irridescence involved...

[JKT]

The spectra mentioned by dilettante can indeed be a problem. For that reason I tend to use my own light, which comprises of two flashes and a diffuser. That usually gets me pretty good starting point. Using a color chart in the beginning, end and occasionally in the middle will help. The iridescence is indeed another ballgame, which I haven't found a way to master.

The macrolep setup is attached...

[Nickkor]

Hi Pete;

Similar to Dilettante's suggestion you could try a 'Digital Grey Kard DGK-1 White Balance Card / Gray Card for Digital Photography' (Amazon) or similar, these are a set of 3 credit card sized cards - one each of black, white and mid grey. This technique is applied post shoot using your RAW software and does not require the white balance to be pre-set in camera; cost is reasonable at about £12.

Shooting in RAW format take a shot of the 3 cards together under the ambient lighting conditions to be used for your actual images. After loading the images into your RAW software use the image of the 3 cards to set the black, white and mid grey point and save the resulting 'white balance'. The resulting white balance can then be copied to your remaining RAW images and should remove any colour cast caused by the lighting. I have a Nikon camera and have used the above process with Capture NX2 software, I tested the cards taking a picture of the slightly cream coloured table cloth in our dining room, the default image from the camera rendered the cloth as white, applying the white balance obtained from the 3 card image restored the correct slight cream colour.

Regards,

Nick.

[MikeOxon]

This is a huge subject that starts, as others have noted, by using a defined light source. A continuous spectrum lamp together with a grey card as a reference is a good start.

The camera interprets the colours of the subject by photographing through colour filters - usually three types arranged in a matrix over the photocell array that forms the image sensor. The precise formulation of these filters is one of the 'secrets', jealously guarded by camera manufacturers, and one reason why it is a good idea to stick with the mainstream 'professional' brands. The process of extracting the colour information from the sensor is described in an article on my website at: http://home.btconnect.com/mike.flemming/usingraw.htm

After that, all the colour information is stored as triplets of numbers within the image data. For consistent reproduction of colours, it is essential that these numbers are interpreted in the same way, at all stages of the image processing from camera through to VDU or printer. This is the purpose of 'colour profiles'. I also have an article about this on my website at http://home.btconnect.com/mike.flemming ... olour1.htm

After having said all this, remember that 'colour' is a quality constructed by the brain and not a property of objects as such. The brain adapts to different surroundings, so that we see differently in artificial light than in daylight, for example. A camera does not compensate in this way and has to be set according to the ambient light. The brain's perception is context sensitive. I have posted the following image before but you can check with a colour editor that the two marked squares on this cube are identical colour and brightness - only the context is different!

based on image by R.Beau Lotto
more illusions at http://www.echalk.co.uk/amusements/opticalillusions/colourperception/colourperception.html

The bottom line is that it is not possible to transfer a coloured object from one context to another (e.g. a photo) and expect the brain to interpret the result in the same way. The camera may not lie but the brain frequently does!!!

Mike

[ScottD]

I was thinking of taking a photo of a colour chart / grey-scale chart and using that as my reference. Hope that makes sense!

GretagMacbeth Colorchecker perhaps?

[Willrow]

Further to the sound advice already provided Peter, consider investing in a GrafiLite a portable Natural Lighting System that comes complete with it's own Neutral Viewing Mat (18%) for Accurate Colour Assessment.You can find out more about this product from Colour Confidence who are market leaders in Colour Calibration products.

Over many years I have taken countless numbers of record photographs of Moths both at home and on field excursions (Moth Traps) using this type of 'simulated daylight' with an 18% Grey Card background, the results are as near as you can probably get to realistic and providing you don't do something silly in post processing etc afterwards, should suffice. In fact if you get your subject right in the first place, no post processing will be required, why alter something that was supposed to be realistic in the first place. While on this subject, museum specimens are rarely if ever good examples, often faded, dried up, chemicalized etc, you might very well reproduce the specimens 'faithfully' - yet whether they represent the original voucher/token specimen is impossible, death removes all vitality and this affects colours in particular, even allowing for the most modern preserving techniques this problem persists.

You should reach a point where you are happy with the results achieved, like most things in life...it takes practice before getting it right I'm sure you will find your way around the problem, but I would advise doing some homework and then practice your technique prior to your museum visit. There are many Colour Calibration products on the market and the subject is full of conflicting advice that can often cost the unaware, don't let it become too complicated...I'll do that for you

Regards,

Bill

[Pete Eeles]

Thx for all of the excellent replies all - I'll definitely be getting some practice in when I get home and putting this advice to the test!

Cheers,

- Pete