So how big DO you print? And why it matters!

I only ask because a lot of the pointless discussions on this forum seem to centre around minor perceived or real issues that are only visible at 100% magnification on a monitor.

Personally I only care about what prints look like, because web posting can make anything look good from 3mp upwards.

Not only are monitors very low resolution (so you are seeing the effect of pixelation as well as other issues) but if you zoom out you are still viewing an image at approx 100 DPI (for a 23" 1980 width display). You are not seeing a print.

At 50% you are still looking at an image which is 100dpi on a monitor, but about the same size as an image printed at 200 dpi. Can't see the issue any more? Well you will see it even less in a 200dpi print because you will not be looking at a pixelated image, and it will be reflective not transmissive.

So how big is a 200 dpi print on a 16MP image? Answer 24.5" X 16.3". This is good enough for A2 paper (23.4" X 16.5") assuming you are using enough border to frame it. Do YOU have an A2 printer? No? Perhaps you have an A3+ photo printer? In that case you are printing at 260dpi  good enough for high quality publishing.

So can a camera perform well at 100% but less well in print? Sure...so, what looks bad in an A2 print at 200 dpi? Generally things that are plainly visible at a macro level:

Over sharpening (or the wrong radius of sharpening) makes halos extremely visible and detail crunchy. This can actually be exaggerated by printing. Something many cameras do in JPEGs by default (Oly specifically) which is why on all other cameras I am forced to shoot RAW.

Corner softness and vignetting can ruin a large landscape print, and poor in-out of focus transition can ruin a good portrait. Too much auto-correction of vignetting can also enhance noise.

Distortion and CA of course. Removing distortion in PP will also reduce resolution.

Colour and macro contrast, noise, banding, shadow blocking, shadow noise or blown highlights also scream at you when looking at a large print.

And NOTHING looks worse than aliasing and moire, something I have wrestled with with every Bayer camera I ever owned and which makes digital cameras look "digital" in a large print.

Same for noise and NR. A little luminance noise can be ignored because downsizing removes it. The killer is chroma smearing - patches of colour variation on dark areas - something the Em5 is NOT so good at at high ISO (though better than many other CSC cameras)....

http://www.dpreview.com/reviews/olympusem5/16

Have a look at the chroma blotching on the grey patch at ISO 800/RAW. Downsizing may remove the grain, but the blotching will still be visible. In JPEG, the Oly simply applies very obvious NR which will ruin detail in larger prints. Chroma NR will also change the colour balance.

Any camera with an AA filter will produce mush in place of detail when it reaches the resolution limit. It will look fine up to that point, but very smooth beyond it which - for grass or foliage of brickwork - looks very unrealistic. While necessary to eliminate moire on a 2X2 array, the Xtrans sensor does not have an AA filter, so detail at the resolution limit does not suddenly become mushy. In fact it continues to produce false detail below the resolution limit, which gives the appearance of greater detail because details beyond the limit do not become artificially smooth.  This is not real detail, but it does improve the appearance of the detail that is there (see Foveon).

If you wonder about that, here is DPreviews resolution assessment of the Xpro1...

http://www.dpreview.com/reviews/fujifilm-x-pro1/18

Look what happens below Nyquist. This indicates there there IS detail in those shots, which you can achieve with careful sharpening, but that the detail does not decay to mush past its limit.

Compare that with the Sony A99 here....

http://www.dpreview.com/reviews/sony-alpha-slt-a99/15

See a massive difference? Only in RAW does the A99 compare, and even then with lots of colour moire. At that's a 24MP FF sensor.

Yes, some of the default images chucked out by review sites look oddly disappointing, but these are not PP'ed in any way, unlike the resolution test above. With moderate PP, the Xtrans shots look a lot better, plus they have a notable absence of some of the more pressing issues listed above and better subject isolation than MFT cameras as well.

So, if your measure of a good camera is based on 100% pixel peeping while ignoring the more visible issues (above) then you are not a photographer, just an obsessive gadget freak engaged in a c*ck waving competition.

Or, of course you may be a sensor or RAW processor designer, in which case this is a perfectly reasonable course of action, but that only accounts for about 1% of the posters on this site (and most of them are probably talking BS - hard to tell isn't it?).