Testing to distraction

You may want to do your own testing of cameras, lenses, or ‘digital film formulas’. The key to testing is the basic scientific one: identify and control the variables. When I checked the Kodak P850 against the Nikon P610 I made a side-by-side tripod mount so the cameras’ views would be as alike as possible:

twinned

Unless you are comparing two cameras, you won’t need this. If you’re just comparing lenses for example, simply mounting the camera on a tripod so it doesn’t shift between lens changes works. Another key to successful testing is to fix white balance, ISO, and either shutter speed or aperture so that the only difference between shots is one aspect of exposure, not all of them.

It’s a bit peculiar that they sell us ‘professional’ cameras that automatically pick white balance, ISO, shutter speed, and aperture (not to mention focus) all according to some magic artificial intelligence program that purports to be smarter than the photographer – and almost never is. If we wanted that type of simplicity we’d just get a cheap point-and-shoot to begin with. Okay, let’s be kind and say they do this so that people can graduate from point-and-shoot to making their own choices. Of course you can always do it the other way if you’re feeling lazy one day and just want to grab some quick shots without having to work out the best settings. Besides, that way you can blame any errors on the camera.

One of the most important things for doing testing is a consistent subject. Professional evaluations should take place in the studio with controlled lighting and charts: lens resolution, gray scale, and colour rendition. That’s pretty boring, but the opposite method which I’ve seen used too often by photographers is to just shoot away at stuff and decide later if the picture is good. The first method may be dull, but the second is utterly lacking in true objectivity.

If you’ve followed this blog you have been subjected to the mind-numbing dullness of ‘The Sheds’; my easily-accessible consistent subject matter that has green tones and red tones and lots of lines. It works very well, to a point. But supposing we want to get more objective results? Or supposing you don’t happen to have a convenient fixed subject to work with? What then?

Enter the cards. It’s all in the cards! You can of course buy cards specifically made for photographic purposes. Or you can visit the hardware and dollar store and save yourself a small fortune. It does not matter that the results aren’t some professionally-graded standard; the purpose is comparative, and you compare one shot of the homemade card to another to see the difference in whatever aspect you changed between them.

Behold the gray scale card made of foam board from the dollar store, paint chips from the hardware, and a bit of glue to put them together:

IMG_1243IMG_1244

The first image is a monochrome shot of the card, the second in colour. Notice the colours? You won’t see them in real life, and that’s the important lesson: what you see, what the camera sees, what the computer ‘sees’, and what anyone looking at their screen sees are not going to be identical. Here’s what happens when we use the “Enhance Colour” function:

1244CE

Understand that this is the computer interpreting shades as colour and then increasing what it perceives to be there.

Now here is the colour version desaturated to black & white using GIMP ‘Lightness’, ‘Luminosity’, and ‘Average’ conversion methods:

1244LI1244LU1244AV

You can see subtle differences in some of the shades between them and certainly in comparison to the monochrome shot.

The background for this, by the way, is a much larger black-and-white card constructed of foam board and poster board. The reverse of that is this:

IMG_1246

This can be used for comparing colours, and understanding that they don’t always show up the way they should. For example the four colours there (clockwise from upper left) are red, green, blue, and yellow. Really yellow. On my screen it looks decidedly beige. Probably yours too, as computers use red, green, and blue to build colours from even though yellow is the true primary, not green. Complicated isn’t it?

This would be “correct” on my screen:

1246YC

Now imagine what would happen if you made a colour card out of paint chip samples. Hmm. I may just do that.

As you can see, testing isn’t much fun. Certainly not as much as shooting pictures and getting the results you want, but understanding what’s behind the dull stuff makes it easier to do the fun stuff. You don’t really need to use cards like this (or a resolution chart, which is basically sets of progressively closer fine lines) but it gives you the principals for making changes for actual pictures.

IMG_1245

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s