I’ve been talking a lot recently about camera acclimation testing, and getting to know your camera. In that same vein I want to talk for a moment about personal minimums for image quality. I’ll be honest, I never thought a lot about how far I can push the quality before the results became unacceptable.
Moreover, I need to be right up front about this. I can’t tell you what you should consider the worst quality you’ll accept. What I can do is point at some areas that I’ve found illuminating or instructive and maybe that can lead you to figuring things out a bit better.
My first real exposure to just how far I could take things was back when I was shooting waterfalls in Alaska at ISO 1250 and 1600. I figured I wasn’t going to get anything worth keeping, but I had no idea if or when I’d have the opportunity again, so I shot anyway. I printed one of those images just for kicks. The quality isn’t great at nose limited distances, but it was perfectly fine at more reasonable viewing distances. Shocker! ISO 1250–1600 on the 5D mark III isn’t as unusable as I had though it was.
One of the problems I think I’ve had with judging what is acceptable is an artifact of Lightroom and computer displays. Because displays have a fixed relatively course spacial resolution, they can’t render finer and finer details when fed with higher resolution source material. As a result, software like Lightroom requires you to zoom in to 1:1 magnification to see the effects of sharpening. Otherwise, the effects, which are happening at the pixel level, would be averaged out due to scaling.
A result of this is that 100% magnification becomes the zoom where I judge sharpness and quality. However, the image is massively magnified compared to what it would be in most cases as a print.
For example, my monitor is a 24“ 1920×1200 display ( 20–7/16” x 12–3/4″ and about 96 PPI). It displays all of 2.3 MP worth of information at a resolution far lower than you’d want for a print of the same size and viewing distance. If I were to imagine my monitor as a window into a print of some size, with a 23MP camera like my 5D mark III, I’d be looking at a small part of a 60 x 40 inch print.
In effect, this means that I’m sitting less than an arms length away from an image that I’d never view anywhere near this close under normal conditions. As a result, I’m seeing details far more magnified than I would under normal viewing conditions.
Going the other way, fit magnification doesn’t help since there’s so little detail that you’re not going to remotely see exactly what a print would look like either.
This monitor and magnification situation also has an impact on judging acceptable motion blur and sharpness.
This is another example of an image that fails the 100% magnification test. However, as a print, at a reasonable viewing distance, it holds up just fine.
My point here isn’t that we shouldn’t lower our standards to accept poorer images. However, if we don’t really understand what your absolute floor is for quality you may be missing out on opportunities. Then again, maybe this has been obvious to everyone but me, and I’m just coming to the party late.