My recent post on microshake and micro blur has generated a lot of very positive feedback and I thank folks for their grace and courtesy. This article discusses steps to take when you've done all you can and believe that the issue is the combination of lens and camera. It get's a little deep, so be brave and read on... So you've decided that your hard earned lens is not as sharp as you think it should be. You might be right. Certainly if you are paying for Canon L glass, or Nikon FX glass, or Zeiss glass, your investment is not insignificant and you want the best return possible. On the other hand, that $200 do everything, or cheap "nifty-fifty" are built to deliver a price point and while they can be quite good, let us not kid ourselves about the depth of their optical excellence. Because that pond is quite shallow and often murky.
Regardless, you think your lens can do better. What can you do?
If you are not technical, or don't have computers and cables and lighting and lots of time, take your lens to the manufacturer for service. Some lenses have long warranties since a lens is typically either good or bad. In Canada, Sigma does up to 7 years, Nikon does 5 years, Tamron does up to 7 years. Canon still has a bad case of crania-rectal inversion and only offer one year. Warranty aside, the service professionals who do the work for these companies are typically very talented folks who want to do great work and they can help.
Over the years, I have learned that even the best lenses don't always perform to their top grade. The measure of Fit helps determine how accurately a defocused lens, refocuses time and again on the same target at the same distance. Some lenses just don't and there may be little you can do about that. In order to make AF very fast, focusing helicoids ramp very quickly and that can result in inconsistency in focus performance. The different focusing methods, contrast and phase detection will also impact focus success as will the available light and the reflectance of the focus target.
I have owned and still do own lenses that I just didn't believe were doing their job properly. Up until recently I have used the Spyder FocusChecker as a high level route to dial in focus adjustments. What is that you ask? More powerful cameras often offer something called AF micro adjustment, a means to fine tune a particular lens on a particular body. Check your manual to see if your camera does this and if not, you are back to the manufacturer because you aren't going to be able to go further on your own.
Recently, I did some research into computer assisted focus checking. One company Reikan, stood out with their product called FoCal, which as you might gather stands for Focus Calibration. The Pro version of the software sells for 70 British Pounds (so not cheap) and you are definitely involved in the process of installation and operation. While there are fully automatic modes, there is a dependency on your camera being able to have the AF micro adjustment controlled remotely. Nothing I use can do that so there is a lot of hands on when doing a calibration.
It's very good software, but a calibration is an investment of your time. You need to print off their calibration target on non-reflective paper, preferably using a high quality inkjet to reduce reflections, although I have been successful using laser printed targets. You don't have to, but I adhered my targets to foamcore using photo spray glue to keep them from bending and twisting. Then you put it on a wall that is evenly lit with reasonable light levels. I hung mine from a justin clamp off a boom and used a pocket level to make sure it was square to the vertical and not twisted.
Next you put your camera / lens combination on a tripod, level it up and move the combination so the centre focus point is centred on the target. Reikan says to remove battery grips if you use them before doing a calibration. I used my trusty RRS TC-24L with BH-40 ball head to hold the cameras. If the lens I was calibrating had a foot, I used it to mount to the tripod.
Now connect the camera via USB to your computer and launch the software. It launches and you discover your camera. This worked very well, although the first connection was slow. Once done, you do a Target Setup where the software takes over your camera and feeds a Live View image to the computer to tell you if your camera is properly oriented to the target. It's very easy to do, but is fidget work. Obviously the distance from the camera to the target will vary with the lens being tested. When I was doing my Sigma 120-300 I nearly ran out of room to separate the camera from the target.
Once you are aligned, you can use the Fully Auto mode and hit start. My cameras cannot adjust AF micro adjustment automatically so as the software makes a series of images, you are instructed what changes to make to the AF micro adjustment settings. It's easy but time consuming. Depending on the aforementioned lens fit, you may get by with as few as six AF changes, or do more. The most I had to do was sixteen. In that space the camera is making exposures at the different settings and the software is analyzing the results to produce a sharpness curve to find the optimum adjustment. Once the proper adjustment is found you simply save it in the camera and that is done.
Now the software should be able to produce a PDF report showing the outcome of the testing and calibration. This didn't work consistently for me, with the program hard crashing more than half the time. Since I had already loaded the corrections on the camera and noted them on my iPhone, this crashing was more annoying than critical but I hope that Reikan improves things in the next version.
I did my testing on a Mac, as long time readers would expect, and the software requires installing the open source Mono framework. When I was at Novell, we had brilliant people working on Mono so even though you may not have heard of it, it works pretty well, but do note that being open source, there are LOTS of versions out there so stick with the one recommended by Reikan.
In the span of a couple of hours, I did calibration on four lenses. I did the Sigma 120-300/2.8, the Canon 85/1.2L, the Canon 70-200/2.8L IS II, and the Canon 28-300/3.5-5.6L IS. The only lens of the group whose sharpness I had ever questioned was the Sigma and while it had a pretty aggressive micro adjustment outcome, it was in bounds. The 85 and the 70-200, had never been a question, but both benefitted from a micro adjustment, particularly at the short end of the 70-200 (a focal length I nearly never use). The 28-300 had been adjusted using the Spyder Focus Checker, and while after doing so I stopped questioning the lens, the outcome after doing the FoCal calibration is quite a bit better.
The software comes in different levels of capability. I bought the Pro version because it can test lenses longer than 400mm, and can also test sharpness across the range of apertures in a lens. We've all heard about the aperture with optimal sharpness, now you can find out which one is best for your camera and lens.
I'm going to offer to do calibrations for my students and members of my photography club for the cost of time only because I think the cost is hard to justify if you don't have a lot of lenses to do. I'm a bit of a gearhead and being able to do all my lenses will make the software worthwhile, and I'm prepared to invest my own time in my gear. The general user may find FoCal overkill, but I will say that it is a powerful solution to a really annoying problem, and may avoid you having to send a lens in for service.