I’m starting to take stock of the last couple of months’ work that I’ve done on this blog. In October and November, two things hit me at about the same time. They were interrelated and complementary, so I pursued both.
The first was the challenge of making sharper pictures for the firehouse series. That led me to developing a computer-mediated sharpness measuring technique, and to experiments with various shutter-release schemes on the D800E and D4. I developed a way to measure camera motion’s effect on the captured image using an oscilloscope. I also started to explore the sharpest f-stops for several macro lenses.
Shortly after starting the sharpness-enhancement project, I received the Sony alpha 7R (which I’ve been calling the a7R, with the lower-case a substituting for Sony’s lower case alpha). The first thing I did was look at shutter shock. I did some of my usual noise and dynamic range testing, and found indications of still-not-understood ISO-setting in-camera processing of the raw files. I found the camera to be pretty close to “ISO_less”. One of the attractions of the camera was its short flange distance, and thus the ability to use many different third-party lenses. In experimenting with these lenses, I found corner color shifts and corner smearing (the former easily fixed, the latter not so) with some of the shorter and more symmetric Leica M-mount lenses, but no such problems with any Nikon F-mount lenses, even very wide ones. M-mount lenses of 90mm or greater focal length appear to be unaffiliated with either problem. I discovered that lens adapters were generally shorter than they are supposed to be. When I tested the Sony/Zeiss 55mm f/1.8 FE Sonnar, I found it to be remarkable, and, considering its price, astounding.
During my testing, I documented shutter-induced vibration that affected the sharpness of the image over a wide range of focal lengths and shutter speeds. I was by no means alone; Lloyd Chambers, Ferrell McCullough and many others also discovered and documented the problem. I spent a lot of time trying to pin it down. I developed a program to simulate the effects of camera motion and other factors on images. I kept the oscilloscope approach from the Nikon D800 sharpness work, but dropped the computer-analyzed targets in favor of ISO 12233. Using that target and a Helicon variable neutral density filter, I determined that shutter vibration affects image quality with a normally-mounted 135mm lens at shutter speeds of over 1/1000 second, which was a real surprise. The Sony’s shutter can affect the quality of images made with lenses as short as 55mm. I debunked the theory that the effects that we’re seeing are the results of sensor displacements of smaller than one sensel.
I also found that shutter shock causes blur in images lit by electronic flash with durations between 1/600 second and 1/1350 second. Ferrell McCullough suggested trailing curtain synch as a workaround. I tried it, and it worked.
While I was working on defining the shutter slap problem, Joe Holmes, Huntington Witherill, Mike Schultz, Mike Collette, and others were working on trying to solve – or at least mitigate – it. One of the obvious things for me to do now is to test some of their proposed solutions. I am considering whether that’s a good use of my time. One of my reservations is that such testing could expand to occupy all my waking hours. The number of lenses, tripods, heads, and mounting arrangements is seemingly unbounded. Both of my testing regimens are labor intensive. When using the ISO 12233 target, images have to be made over a broad range of shutter speeds at 1/3-stop intervals, assembled into Photoshop layers, and analyzed visually. I’ve recently improved the precision and sensitivity of the oscilloscope work – I hope to report on my progress here soon – but, as often happens in such cases, more sensitive measuring instruments react more strongly to things that you’re not trying to measure as well as their intended target. The improved oscilloscope procedure shows the effects of distant footfalls, electric fans, and, for all I know, micro-earthquakes. This means that a fair number of test runs have to be discarded.
And then there’s a mammoth disincentive to my embarking on time-killing testing centered on making the a7R images sharper: I’ve decided that it’s not the right camera for the firehouse series, at least as long as I stay with available light. It has no image quality advantages over the D800E. The Sony does let me use Leica M-mount lenses, but I have F-mount ones in the focal lengths I plan to use that should produce similar results (One of the really nice things about having the a7R around is that it lets me compare M and F-mount lenses on the same body) and focus closer to boot. The smaller size and lighter weight of the Sony is no advantage in the firehouse series, especially since I’ve decided that I have to construct blinds to keep reflections of the camera and tripod out of the pictures. The Sony does have a much better live view image than the Nikon, making focusing more accurate and less unpleasant, but I’ll have to live with that.
For me, the long pole in the tent is the Sony’s shutter shock. I’ll be using lenses from 50mm to 135mm, so shutter-induced vibration is a real concern. Perhaps it can be moderated, but my work with the Sony a7R and the Nikon D800E over the last several months have proven to me that, in circumstances where sharpness is a key photographic quality, I’ll be struggling for the best setup even with the Nikon and its myriad ways to provide delay between the mirror-up and/or shutter-cock operations and the release of the first curtain, and why not start with the smallest vibration forcing function and then do everything practical to deal with it?
Therefore, my tentative plan for this blog the next few weeks is as follows:
- Report on the new oscilloscope testing procedure, with some a7R examples.
- Produce an overall assessment of the a7R based on the testing that I’ve already done..
- Start work on testing the D800E’s vibration blur through a range of longish (say, 1/8 through 8 seconds) shutter speeds.
- Develop strategies to ameliorate the D800E’s vibration blur under the circumstances that I’ll be facing in the firehouses.
- Compare how well I can do in sharpness with available light and electronic flash.
- Make the decision whether or not to use electronic flash in the firehouse series. Advantages are (unknown at this time) sharpness advantages. Disadvantages are complicated and time-consuming setup and takedown, need for AC power, and the fact that I’d need an assistant.
- Perform sharpness testing on the candidate lenses and pick on for each required focal length.
- Decide what to do about stacking. My initial experiments have been frustrating. The motorized tracks are more prone to vibration than conventional mounting. Using available light, exposures need to be made before the light can change much. Making many photographs of each setup is time consuming and discourages experimentation.
- Make and post some actual pictures. The sooner I can get to this point, the better, but not at the expense of having to go back and reshoot everything if I discover that I could have made things better.
- If and when there develops a body of conventional wisdom on dealing with Sony a7R shutter slap, test it.
All this is subject to change as I learn more.