If the viewer is optically challenged, the viewer will prefer the coarser, more aliased image. The person with an eye for natural detail will appreciate the higher density, and this image will. also, empirically contain more information about the real world detail.
Maybe you're in the wrong conversation (or most of the rest of us are) - you're talking about artistic decision. I don't think anyone here is trying to say you don't have a right to have a coarser, more pixelated final image. Most of us are talking about what an accurate representation of reality is captured, and it is true that a higher-res capture gives you more options (although it can not exactly duplicate any arbitrary lower density, as those have artifacts relative to their pixel resolution).
Erk, that's a typo. Last phrase is supposed to be "it's that the sensor can't pick up the ADDITIONAL sharpness you get from less diffraction."
Very well put. This is all so clear to me that I am just baffled by this anguishing over diffraction every time a higher resolution camera comes out.
Don't forget the noise. A sensor with twice the MPs will have 1 stop higher pixel level noise, and that noise (or variation, in the light/signal itself) will (also) to some degree mask the extra detail that was the whole reason for the increased MP count.
Maybe reviewers like to show off their technical vocabulary ... but they often say things that aren't true, or easily lead to misunderstanding.
Nope! Sounds logical, but it doesn't happen. The extra noise of breaking down into smaller pixels doesn't hide any detail; not unless the "read noise per unit of area" increases, and you're looking at the shadows.
Yes I agree there and the reason for my understanding of diffraction is that since I started using the MP-E 65mm I find that I have to use apertures well into the diffraction limited zone to get enough depth of field. I've found that with careful sharpening it is possible to get these diffraction softened images looking pretty sharp. As you point out an image lacking in aliasing is easiest to sharpen. Here's an example. It's not perfect because I only did it quickly for web posting and I need to fix a bit in the lower left corner, but it gives an idea of what I have been working on. I use a mixture of wide pixel radius sharpening for micro contrast and smaller pixel radius for detail. This one might be a touch harder than ideal, but as I say it was processed quickly.Maybe reviewers like to show off their technical vocabulary ... but they often say things that aren't true, or easily lead to misunderstanding.
Diffraction is just a very minor additional softness to an image, easily sharpened if there isn't a lot of noise. In fact, diffraction can be thought of as two slightly softened versions of an image, blended. One, at 80%, is just the central cone caused by the diffraction. The cone part has a very narrow radius. The "rings" part of the diffraction has a wider radius, and can easily be mostly removed in effect with a wide-radius USM.
I would not recommend to anyone to back off on the f-stop if they want the DOF, just because of diffraction, especially when noise is low. In fact, an image totally lacking in aliasing is the safest one to sharpen. Any image capture system where a black-to-white transition takes less than three pixels exclusive is broken and artifacted, anyway. We accept aliasing only because of the coarse captures and display.
--
John
That's because sharpness is about media and expectations, not about real resolution.
Wow, that eye is one tough subject! I'm amazed it doesn't look more aliased. Probably due to a total lack of aliasing in the capture, combined with weight-conscious sharpening.
I didn't say that the (more noisy) higher MP image would have less detail than the lower MP image, just that the extra noise also is an 'issue' that should be considered when discussing pixel density, together with diffraction, camera shake, lens blur, etc.
If converters automatically sharpened more for higher f-stops, I don't think most people would complain, except for the higher noise at high ISOs.
The beginning of my approach is to use very softly diffused wrap around flash lighting which allows me to apply far stronger sharpening than you could normally get away with. Normally flash lighting produces too much contrast and it limits the amount of sharpening you can apply. To me diffraction is just a normal optical effect to be dealt with. It's inescable at high magnifications because at none diffraction limited apertures there simply isn't enough depth of field to get a reasonable photo of a 3-D subject. The way some people talk about it you'd think your camera would explode if you dared to use an aperture in the diffraction limited zone.Wow, that eye is one tough subject! I'm amazed it doesn't look more aliased. Probably due to a total lack of aliasing in the capture, combined with weight-conscious sharpening.
--
John
The bottom line is that the ONLY known potential IQ issue with increased pixel density is the failure to maintain or decrease read noise per unit of sensor area.
Here is a very good and accuracte article about diffraction and diffraction limits.
