From time to time Thom takes on statements made on other Internet sites and tries to provide the correct context for them.
"If you have twice the reach and/or can handhold shots, you get a lot more winning shots much more easily." —comment about Olympus 150-400mm f/4.5 lens.
A lot to parse in this seemingly simple statement. I'll start with the un-obvious: what the heck is a "winning shot"? This particular comment came about as the result of someone asking how to get an acceptable image of a reticent bird. Somehow that escalated to "winning shot." The implication is that a "winning shot" only is a tight one of a bird. While I'm an advocate of getting the composition (crop) right in the field, I've seen really great shots of a bird taken with a lot less focal length than even 300mm (effective).
What tends to happen is this: people want to replace access (being close to something, like a bird) with focal length. That brings up all kinds of other issues, including what's happening in the air between you and the subject. And let's not forget perspective. There's no one right answer, and the rightest answer is not "the longest lens you can find."
"Thanks to [X-Trans] technology, our cameras don’t have moiré or false color." — Fujifilm executive in dpreview interview
Simply not true. This statement and its authenticity has been examined by many for years now. You can still produce color artifacts with X-Trans sensors, it's just not as easily provoked and has different characteristics than Bayer color artifacts. As I pointed out in my early X-Trans camera reviews, the original approach to converting X-Trans raw data tended to produce color smudging on small detail. More recently, raw converters have found ways to reduce the visibility of said smudging, but X-Trans still has the same problem as Bayer: near adjacent color data is being used to interpolate missing color data. There are always consequences of interpolation, typically false color and/or antialiasing.
I understand the problem, because it's a marketing problem. How do you say "our approach is somewhat better." Not perfect, but in some modest yet meaningful way better. Let's just say that the visibility of color artifacts is minimized in Fujifilm's approach. But as a marketing statement, that's pretty weak. The majority of your potential customers won't know what you're talking about, and those that do will want a measurement. But lying about the difference? Sorry, that's a terrible approach, and Fujifilm needs to stop repeating that falsehood, immediately.
Worse still, the same executive goes on to say "our medium format 100MP sensor resolves 99% of subjects, so there’s no moiré or false color, so no need for X-Trans." Also not true (including that 99% resolution claim). Yes, with more pixel density—and that 100mp medium format pixel density is basically no better than the APS-C or full frame state-of-the-art right now—the visibility of color artifacts would be reduced, particularly if you compare 8x10" prints from a 100mp medium format camera against a 24mp APS-C camera. But no, it doesn't go away completely as Fujifilm keeps suggesting.
Thing is, by lying you lose credibility with the very customers that would be interested in an actual reduction of artifact visibility. Why the heck would you ever do that? This is a "just believe me" approach that ultimately gets you into trouble, because it will turn out some day for a customer that you shouldn't have been believed.
As much as I like the Fujifilm products, Fujifilm needs to stop misleading customers about them. These aren't just misstatements, they are outright falsehoods. Funny thing is, later in the interview one of the Fujifilm executives correctly points out a key difference: "our biggest point of differentiation is color reproduction." Yes, very true for many serious photographers. Moreover, it's an aspect related to the thing they're misleading people about. Not only are the Fujifilm film simulations different in what the other camera companies are doing, but the presence of color artifacts does tend to be lower, as well. Why can't we just get that wrapped into a strong, accurate marketing statement?
"Nor does cropping an image change its depth of field, whether the crop is the sensor, or Photoshop. For example, capturing an image at 300mm @ f/4 is identical in the APS-C portion whether the camera records a full-frame capture or an APS-C crop-capture." —assertion about depth of field on another site
The pixels don't change when you crop, that much is true. But depth of field (DOF) as a concept—it's not a fact, as actual "focus" is always in a infinitely narrow plane (which may be curved), so what we're always talking about with depth of field is perception of acceptable focus—and that has to do with how the final image is presented. To repeat: DOF is a concept that items outside the actual focus plane might be perceived as in focus by someone in the final reproduction. Zeiss's original formula, which generated the Circle of Confusion values that are so prevalent throughout photography, is based upon magnification of the capture to a common presentation size, and how that presentation size is viewed.
So if your intent is to produce, say, 13x19" prints, you'll be magnifying the crop-sensor pixels more than the non-cropped pixels. And in doing so, the so-called Zeiss approach asserts that you trigger a change in perception of acceptable focus. This is the most commonly held DOF concept in use.
Now there are plenty of other theories about DOF—again, not facts, but theories—that speak to something other than what Zeiss does. The two most prevalent have to do with the size of the aperture opening through which the light moves, and the size of the photosites on the image sensor. In both cases, those theories are trying to get to "the smallest thing you can record." If that thing is within some acceptable "focus"—and remember, we're usually dealing with interpolated Bayer data—then you have "depth of field" that's "acceptable".
It doesn't help that the DOF discussion more often than not devolves into comparison of m4/3 and full frame, which have different aspect ratios.
There's a reason why we talk about "equivalence" so much. As with almost everything in photography, we have to balance different factors against each other in considering what to do to capture an optimal set of image data.
