Custom Camera Calibration

The other day I had an email fall into my inbox from leading UK online retailer…whose name escapes me but is very short… that made my blood pressure spike.  It was basically offering me 20% off the cost of something that will revolutionise my photography – ColorChecker Passport Camera Calibration Profiling software.

I got annoyed for two reasons:

1. Who the “f***” do they think they’re talking to sending ME this – I’ve forgotten more about this colour management malarkey than they’ll ever know….do some customer research you idle bastards and save yourselves a mauling!
2. Much more importantly – tens of thousands of you guys ‘n gals will get the same email and some will believe the crap and buy it – and you will get yourselves into the biggest world of hurt imaginable!

Don’t misunderstand me, a ColorChecker Passport makes for a very sound purchase indeed and I would not like life very much if I didn’t own one.  What made me seethe is the way it’s being marketed, and to whom.

Profile all your cameras for accurate colour reproduction…..blah,blah,blah……..

If you do NOT fully understand the implications of custom camera calibration you’ll be in so much trouble when it comes to processing you’ll feel like giving up the art of photography.

The problems lie in a few areas:

First, a camera profile is a SENSOR/ASIC OUTPUT profile – think about that a minute.

Two things influence sensor/asic output – ISO and lens colour shift – yep. that’s right, no lens is colour-neutral, and all lenses produce colour shifts either by tint or spectral absorption. And higher ISO settings usually produce a cooler, bluer image.

Let’s take a look at ISO and its influence on custom camera calibration profiling – I’m using a far better bit of software for doing the job – “IN MY OPINION” – the Adobe DNG Profile Editor – free to all MAC download and Windows download – but you do need the ColorChecker Passport itself!

I prefer the Adobe product because I find the ColorChecker software produced camera calibration profiles there were, well, pretty vile in terms of increased contrast especially; not my cup of tea at all.

Now this is NOT a demo of software – a video tutorial of camera profiling will be on my next photography training video coming sometime soon-ish, doubtless with a somewhat verbose narrative explaining why you should or should not do it!

Above, we have 5 images shot on a D4 with a 24-70 f2.8 at 70mm under a consistent overcast daylight at 1stop increments of ISO between 200 and 3200.

Below, we can see the resultant profile and distribution of known colour reference points on the colour wheel.

Next, we see the result of the image shot at 3200 ISO:

Now let’s super-impose one over t’other – if ISO doesn’t matter to a camera calibration profile then we should see NO DIFFERENCE………….

……..well would you bloody believe it!  Embark on custom camera calibration  profiling your camera and then apply that profile to an image shot with the same lens under the same lighting conditions but at a different ISO, and your colours will not be right.

So now my assertions about ISO have been vindicated, let’s take a look at skinning the cat another way, by keeping ISO the same but switching lenses.

Below is the result of a 500mm f4 at 1000 ISO:

And below we have the 24-70mm f2.8 @ 70mm and 1000 ISO:

Let’s overlay those two and see if there’s any difference:

Whoops….it’s all turned to crap!

Just take a moment to look at the info here.  There is movement in the orange/red/red magentas, but even bigger movements in the yellows/greens and the blues and blue/magentas.

Because these comparisons are done simply in Photoshop layers with the top layer at 50% opacity you can even see there’s an overall difference in the Hue and Saturation slider values for the two profiles – the 500mm profile is 2 and -10 respectively and the 24-70mm is actually 1 and -9.

The basic upshot of this information is that the two lenses apply a different colour cast to your image AND that cast is not always uniformly applied to all areas of the colour spectrum.

And if you really want to “screw the pooch” then here’s the above comparison side by side with with  the 500f4 1000iso against the 24-70mm f2.8 200iso view:

A totally different spectral distribution of colour reference points again.

And I’m not even going to bother showing you that the same camera/lens/ISO combo will give different results under different lighting conditions – you should by now be able to envisage that little nugget yourselves.

So, Custom Camera Calibration – if you do it right then you’ll be profiling every body/lens combo you have, at every conceivable ISO value and lighting condition – it’s one of those things that if you don’t do it all then you’d be best off not doing at all in most cases.

I can think of a few instances where I would do it as a matter of course, such as scientific work, photo-microscopy, and artwork photography/copystand work etc, but these would be well outside the remit the more normal photographic practices.

As I said earlier, the Passport device itself is worth far more than it’s weight in gold – set up and light your shot and include the Passport device in a prominent place. Take a second shot without it and use shot 1 to custom white balance shot 2 – a dead easy process that makes the device invaluable for portrait and studio work etc.

But I hope by now you can begin to see the futility of trying to use a custom camera calibration profile on a “one size fits all” basis – it just won’t work correctly; and yet for the most part this is how it’s marketed – especially by third party retailers.

Become a patron from as little as $1 per month, and help me produce more free content. Patrons gain access to a variety of FREE rewards, discounts and bonuses.