and things went downhill from there

While I was updating the firmware in my camera and lenses today, I paused to consider that not that many years ago the idea of fixing issues or adding features to a camera or lens via a software push would never have occurred to me.

I dug out and dusted off my old DX 50mm lens and slapped it onto my camera with the adapter to play with it a bit. My camera does not support the autofocus in this ancient lens, but everything else still works.

On a whim, I set my camera on full manual ISO, shutter and aperture (which I set at 1.8 on both). Then I swapped out the old and new lenses, using each to take a picture of the other. I then took the raw images and gave them identical processing and cropping.

The old lens still works fine - after a fashion. It would probably be okay for distance shooting, though as a crop lens it would likely function more like an 85mm lens than 50mm.

On the other hand, trying to focus on things up close is very fiddly with the old lens. I don't get a split view when focusing, but my camera gives pretty good feedback on that front. Even so, with the lens wide open, the depth of field up close is about 1mm. Fiddly indeed!

I am still mulling over the idea of getting a new lens for my older DSLR - though in fairness, "still mulling" refers to a subject that I only broached just over a week ago.

I decided to go with dakhun's suggestion, and I put together a histogram of how often I use various focal lengths. I parsed through ~10,000 pictures, figuring that would be a decent sample size to establish some trends.

The first thing I did was eliminate 18mm, 50mm, and 200mm. I took out the two extremes because in the heat of the moment, I tend to turn the lens to the extremes (zoom right in, or right out) with the intention of fixing/cropping the picture later if needed. I took out 50mm because I already have a fixed 50mm lens, so the sample had a very strong bias.

When I removed those numbers, noticed some interesting trends in the graph.

- I definitely favour the wider focal lengths over the narrower ranges, with >45% of my shots coming in <48mm, and the other 55% spaced out with increasing rarity up to 170mm, then nothing at all between there and 199mm. I guess if 170mm doesn't cut it then I tend to just zoom in all the way.

- There are definite data spikes in the around 24mm, 70mm and 100mm.

On first glance, it looked like I would probably be better served by a more expensive 24mm lens, but I've worked around data long enough not to fooled by the pretty spikes on the chart. I plotted out a trend curve to see where the majority of my shots were happening, and the peak rose fairly quickly to about the 30mm point and then slid away slowly as it worked toward the narrower ranges.

Based on the numbers, I would probably get decent use from either a 24mm or a 35mm lens, since they both fall in the curve.

My gut feel is that if I got a 24mm lens, it might not see a lot of use. Even though I factored out the extremes from my sample, the fact remains that in the majority of cases, once I zoom out as far as 24mm, I typically just zoom the rest of the way out to 18mm. A lot of the settings in the 20+ range were from me backing off a bit because at the widest setting, the filters encroach on the picture, causing vignetting around the corners.

On the other hand, I fear that if I get a 35mm lens, then my 50mm lens will start gathering dust. In retrospect the 35mm lens is probalby the one that I should have bought instead of the 50mm, but it was twice the price, and money was tight at the time because atara was going to school.

Also, is 35mm enough of a difference from 50mm to be worth the price? I wonder if the camera shop would let me borrow a 35mm lens for a few days to that I could try them both out and see if I think it's different enough to be worth considering...