Film vs Digital: Testing the Nikon D90

It's getting to that time of year when the sun passes lower across the sky and I get a flood of low, winter light coming in through the front window. Good for still life images, so today I set about giving the Nikon D90 a decent test. Full-on, 'test freak' approaches are simply not in my nature but I do want to get the maximum out what I've got and understand what the limitations are. Once I've found the limitations and I can work creatively with what I do have. My yardstick is what looks pleasing to the human eye is fine by me, so no investigations in to the the extremes of chromatic aberration here.


I have taken a number of still life images in the past using a Mamiya RB67 and a Rolleiflex TLR with Ilford Pan F, HP5, etc. I'm also fairly knowledgeable about film based process and 'actions' - expose for shadows/develop for highlights, film speed adjustments, etc.

I am not that conversant with digital practice and as a result I can't say that I've been impressed with digital imaging in comparison to the results I get from film but I've only had the camera for a few months. I do acknowledge that the attention I put in to the medium format film based work in terms of the process is greater than shooting with a 35mm equivalent digital camera. I use the D90 in a sloppy way, automatically expecting something more from digital capabilities. For street and family pics it's pretty good. However, when things get a little more serious, cracks begin to appear. Exposure is usually the first stumbling block. Sensors really don't seem to work with light in the same way as film and too aften the results look nasty, with overblown highlights, etc.

Appreciating that the problems could be with me/lack or knowledge I decided to run a still life test, aiming towards getting best creative result with each run of images and answer questions like: do I still expose for shadows or is digital better exposing for highlights? Hardly life and death stuff, but important! Forgive me as a stumble through my own personal discoveries and errors.

So, with a humble pear, a homemade black background, a Nikon D90 w. Nikon 50mm 1.8 AF lens, and a tripod I set about the task. After binning a few taken at the last known settings - incandescent light (doh) - I took a test shot on my usual Aperture Priority (AP) setting at suggested speed for f5.6 @ 400 iso:

Yuck. Just imagine it in colour. Very unpleasasnt. Overblown highlights on the side of the pear and the black backround is not black. In real life it would be hard to say this was even a strong highlight. The pear was light green and I could make out all the details. Not a specular spot in sight. No Photoshopping that I know of (curves, levels) could mend this. The highlight burn looked ugly and obvious. I feel that this simply would not have happened with film which is much more forgiving in that it compresses the tonal scale better.

Deciding that AP was not the best way forward I went for a manual makeover - spot metering, Manual setting (M), iso 400, f5.6. On this part of the test I found out that the small button on top of the Nikon D90 for adjusting the exposure up and down (which I often use) doesn't work in manual settings. More images for the wastebasket.  However, I was getting closer with a better exposure on the right/light side of the pear and the background blacked out.

Better. However, it is underexposed by at least one stop at c. 1/320th second and I was seeking more defintition on the line of the right hand (darker) side of the pear. I took the exposure reading from the lighter side to see if I could overcome the tendency of digital sensors to overcook highlights and/or extend the tonal range too far. I thought this might help place the lighter side in the right tonal scale area. I got a better result dodging shadow areas at 5% than the opposite action of burning in non-existent highlight detail.

Then a flash of inspiration (forgive the pun) and I dug the Koday Grey Card out of the drawer. I changed the aperture setting to f16 as the DOF at this distance is very shallow and I wanted back to front sharpness in the pear. I made a reading from the grey card.  1/2 a second. I took a few shots at slightly different settings but the immediate impression was positive. The background was dark and the pear well defined. A good balance, but still the light area of the pear was getting driven to highlight city with little hope of a way back home.

A cup of tea and some calm reflection on what I know from film based work and I bracketed 2 stops either side of the suggested reading. 2 stops over should at least put the light area on to a more accurate zone and I could bring the shadows up a bit of necessary. The result:

And a bit of dodging the left hand side at 5% for shadows. Closer to what I was aiming for. I guess with a better working knowledge of photoshop I could make a better attempt but it seems you have to do a lot of post-processing with digital files, althought to be fair it could take a few attempts on the darkroom to get a decent black and white print.

Verdict: While it's unfair to compare an mid level dslr to a rolleiflex TLR or Mamiya RB67 medium format camera, the Nikon D90 is very capable when applying the same care to the processes as I do with medium  format film. I also need to learn more about post-processing for digital camera photos. I still think digital sensors have a problem with compressing tonal ranges into something manageable. For still life I will stick with film and a good medium format camera. Looking down into a flip up viewinder is also far better for composing than the D90's viewinder or Live View.