Tuesday I met Joe and Peter in La Jolla and after dinner we went to Windansea for some photography. I wasn’t too inspired but this angle on the rocks looked interesting, so I decided to abandon my resolution of not producing more “samey” long exposure photography with milky/silky water, and just went for it.
And well, the milk soup bit back. 🙂 After setting everything up, I must’ve touched to focus ring of the lens ever so slightly to screw up focus – it’s not really visible in the web resolution image above, but it makes the image mostly unusable for anything else.
That’s really one thing that drives me nuts about many of Nikon’s AF-S lenses: it’s so easy to screw up the focus with the “instant manual override” feature (you can touch the focus ring any time and adjust focus) because it’s not locked in place by the slotted screw, like on the old AF lenses.
And needless to say, the expensive and heavy 24-70/2.8 has a nice focus ring, but the 24-120/4 (which happens to be my favorite lens) has a flimsy, kind of loose and wobbly thin focus ring that isn’t good for anything. The situations in which I need this “instant manual override” of AF-S lenses is fairly limited. I’d find a “focus lock” switch of sorts far more useful*.
I was using a 6-stop ND filter and it was way past sunset, so when I did the math I ended up at 20 minutes exposure time at ISO100 – too long! I didn’t want to wait that long, so I used ISO200 instead, to get to 10 minutes. Another mistake. The long exposure noise reduction of the D800 didn’t catch all hot pixels at ISO200. All over the image, tiny little hot pixels remained.
I was at f/11, so by using ISO 50 (the D800’s “Lo1” mode), stopping down to f/16 and using a 3-stop filter instead, I would’ve been able to get the same exposure time without the bonus of extra hot pixels. Stupid. I didn’t think.
Last not least, I underestimated once more how quickly civil twilight fades once the sun is gone – my 10 minute exposure (well, 9 1/2 minutes, but the missing 30 seconds don’t really change much, they’re less than 1/4 of a stop) was actually two stops (!) short of being properly exposed in the end, which means too much brightening in post, too much noise, etc. etc. – I could’ve just made the very same photo without any ND filter I guess…
So that’s enough of a sign – I should really leave that long exposure stuff be. Or maybe that old blogpost just invoked bad karma…
*) and distance scales that are more than a crude joke, of course – for example, the 16-35/4 has an indicator at 0.5m, and then infinity – but only at f/22 and 16mm focal length you get hyperfocal at the 0.5m setting! Which means that anything else is guesswork: at 20mm and f/16 (which I’d always prefer over f/22, hello difraction), you have to guess where the 0.85m distance setting is… uhm, somewhere between the 0.5m marker and the infinity symbol… but where?! For a wide angle lens, I find this an incredibly stupid design mistake.