[civiletti]

24mp in APSC is higher pixel density than 36mp in full frame. All arguments on performance and tech level based on MP count alone are bogus beyond file size and write speed.

[ElJamoquio]

I'd rather have 50 MP than 16 MP. But either will do.

[ChristianRock]

When I saw the D4s specs, something Ken Rockwell says sprung to my mind... amateurs care about megapickles, pros care about getting the shot they need when they need it.

It doesn't matter if you like the guy, I believe he is right about this one

[mecrox]

24mp in APSC is higher pixel density than 36mp in full frame. All arguments on performance and tech level based on MP count alone are bogus beyond file size and write speed.

I was really thinking more about dollars and cents. It makes no sense for folks to spend too much just because they think 36+ mpx is somehow a "must-have". It isn't, except in special cases, and It's not worth spoiling other aspects of one's life in pursuit of this chimera. Besides, the real key is good lenses, imho. The dollars and cents stuff doesn't apply quite so much to high-mpx APS-C because prices there for many items are almost a commodity thing nowadays. No one who's said anything about my pics has ever asked how many mega-pixels were involved. That said, I'd have worried for their health if they had. It's soooo easy to get sucked into all the hype, though.

[civiletti]

I was really thinking more about dollars and cents. It makes no sense for folks to spend too much just because they think 36+ mpx is somehow a "must-have". It isn't, except in special cases, and It's not worth spoiling other aspects of one's life in pursuit of this chimera. Besides, the real key is good lenses, imho. The dollars and cents stuff doesn't apply quite so much to high-mpx APS-C because prices there for many items are almost a commodity thing nowadays. No one who's said anything about my pics has ever asked how many mega-pixels were involved. That said, I'd have worried for their health if they had. It's soooo easy to get sucked into all the hype, though.

You are correct that people should not spend money based on specs they do not understand. And for many the Pentax Q would suffice. Specs that make a difference in one's photography are worth paying for. I care little for high frame rate, high iso performance, or how sexy a camera looks. I do care how much detail I can get in an image, how much dynamic range I can capture in one exposure, and how well I can see focus. For me, a 36mp FF sensor is definitely worth paying for.

[TomekP]

No it's not, we're far from that point.

That is simply incorrect.

You should consider this article Do Sensors “Outresolve� Lenses? - but you must note that there is only considered diffraction limit. For lenses open below f/8 there is a problem of different aberration witch significantly lowers the theoretical resolution (which shows the next link).
Or do a quick look at the Lens scoring versus Optical Metrics - DxOMark database - no lens reaches the 36Mpix count on d800.

[dtmateojr]

Here we go again. It's GAS vs Physics No lens can fully utilize a 36mp sensor. I doubt if any can outresolve 24mp even.

[Rondec]

When I saw the D4s specs, something Ken Rockwell says sprung to my mind... amateurs care about megapickles, pros care about getting the shot they need when they need it.

It doesn't matter if you like the guy, I believe he is right about this one

Of course that is true. Megapixels effects cropping ability or ability to print larger. Basically the D800 (36 megapixel full frame sensor) correlates to a 16 megapixel APS-C sensor, like the K5 has. The only negatives I can see with the D800 have to do with file size and because of that, slower processing times (either in camera or out of camera) and slower frame rate.

The D800 would not be a great sports camera, but for landscape it is the closest thing you can get to medium format without having a medium format.

In the end, format size, megapixels are all secondary to light, subject, composition, and lens used.

24 megapixel shot with DA 15




16 megapixel shot with DA *55

[ElJamoquio]

The K-5 has the same pixel density as the D800 and D800e. The K-5 gives me more resolution and dynamic range than any previous camera I have used. The argument against a 36mp FF Pentax for landscape and nature images is disproven by experience.

Experience and experiments.

I love stitched-together pics from the K-5. 36 MP FF would be great. Heck, 50MP would be better.

[jsherman999]

You should consider this article Do Sensors “Outresolve� Lenses? - but you must note that there is only considered diffraction limit. For lenses open below f/8 there is a problem of different aberration witch significantly lowers the theoretical resolution (which shows the next link).
Or do a quick look at the Lens scoring versus Optical Metrics - DxOMark database - no lens reaches the 36Mpix count on d800.

It doesn't need to resolve all 36MP with that P-Mpix number to get a higher system MTF score on 36MP vs. less MP. Read the first link you gave more carefully, and there's a ton of discussion of this on places like dpreview, which I feel like I'm re-hashing every time this question comes up. Read this thread, for example.

"...One misconception about DxO's Perceptional MPix metric is that it represents a resolution limit. It does not. It represents a threshold in the high-frequency roll-off of the MTF. The roll-off might be more or less gentle or steep, but it is not a hard threshold. And in fact, there is abundant information coming through the lens beyond the P-Mpix number that could be exploited in digital signal processing. So an upgrade from 24 to 36MP -- or beyond -- can still make a difference." -Luke Kaven

"...As a result, I can say unequivocally ... even the lowly 50mm f/1.8G will continue to increase in photographic resolution with further increases in sensor resolution. The "magic" lenses will just outpace it (and the other "average" lenses) by about 2:1, when flatness of field is taken into account." - Scott McMorrow

"... the D7100 results are indicating that the point of diminishing returns is quite far away. At some point I will confirm the DxO results with my own measurements, but for now, the projected optical center resolution of most lenses is on the order of 300 lines/mm, based on the D7100 results, and the stellar ones are upwards of 500 lines/mm. Diminishing returns begins to occur when the resolution of the sensor reaches or surpasses the optical resolution of the lens, such that the combined system resolution is sqrt(2) of the resolution of the individual elements. For the average lens, the crossover seems to occur around 75 MP, with the total resulting system resolution being 53 MP (caveat: basing this on the measurements of DxO that have been processed in unpublished ways)." - Scott McMorrow

Also, the DxOmark 'score' there you reference is performing some edge and corner weighting to arrive at the number, so it's not really reporting absolute resolution (DxOmark talks about this, but people still misconstrue it.)

Also, here's another estimate.

"... While near Gigapixel resolutions are feasible with FX from a theoretical point of view, my impression is that FX is "good for" resolutions around 100 MPixels (their eventual sweet spot). I assume higher than 100MP resolutions to become the speciality of larger than FX, or fixed lens FX, cameras. As both will go down in price too." -Falk

.

[Rondec]

It doesn't need to resolve all 36MP with that P-Mpix number to get a higher system MTF score on 36MP vs. less MP. Read the first link you gave more carefully, and there's a ton of discussion of this on places like dpreview, which I feel like I'm re-hashing every time this question comes up. Read this thread, for example.

"...One misconception about DxO's Perceptional MPix metric is that it represents a resolution limit. It does not. It represents a threshold in the high-frequency roll-off of the MTF. The roll-off might be more or less gentle or steep, but it is not a hard threshold. And in fact, there is abundant information coming through the lens beyond the P-Mpix number that could be exploited in digital signal processing. So an upgrade from 24 to 36MP -- or beyond -- can still make a difference." -Luke Kaven

"...As a result, I can say unequivocally ... even the lowly 50mm f/1.8G will continue to increase in photographic resolution with further increases in sensor resolution. The "magic" lenses will just outpace it (and the other "average" lenses) by about 2:1, when flatness of field is taken into account." - Scott McMorrow

"... the D7100 results are indicating that the point of diminishing returns is quite far away. At some point I will confirm the DxO results with my own measurements, but for now, the projected optical center resolution of most lenses is on the order of 300 lines/mm, based on the D7100 results, and the stellar ones are upwards of 500 lines/mm. Diminishing returns begins to occur when the resolution of the sensor reaches or surpasses the optical resolution of the lens, such that the combined system resolution is sqrt(2) of the resolution of the individual elements. For the average lens, the crossover seems to occur around 75 MP, with the total resulting system resolution being 53 MP (caveat: basing this on the measurements of DxO that have been processed in unpublished ways)." - Scott McMorrow

Also, the DxOmark 'score' there you reference is performing some edge and corner weighting to arrive at the number, so it's not really reporting absolute resolution (DxOmark talks about this, but people still misconstrue it.)

Also, here's another estimate.

"... While near Gigapixel resolutions are feasible with FX from a theoretical point of view, my impression is that FX is "good for" resolutions around 100 MPixels (their eventual sweet spot). I assume higher than 100MP resolutions to become the speciality of larger than FX, or fixed lens FX, cameras. As both will go down in price too." -Falk

.

I do think there comes a point in real-life shooting where you see diminishing returns -- not because of lack of lens resolution, or diffraction, but because you just can't hand hold your camera steady enough. At 24 megapixels on APS-C you already have to be pretty careful with technique. I can't imagine shooting 50 megapixels on APS-C and trying for a pixel sharp image.

[jsherman999]

I do think there comes a point in real-life shooting where you see diminishing returns -- not because of lack of lens resolution, or diffraction, but because you just can't hand hold your camera steady enough. At 24 megapixels on APS-C you already have to be pretty careful with technique. I can't imagine shooting 50 megapixels on APS-C and trying for a pixel sharp image.

50MP on aps-c (!) would require higher shutter speeds than we're used to using for handheld if you want absolutely pixel-sharp results, but it will never look any worse than lower-MP in any case and may often look better even with some acuity lost to blur. Consider this comparison, taken from a dpreview thread on this:

[kentishrev]

I'm sure for many of you professionals and high-end amateurs the difference between 16 and 24 MP will make a difference. For me, looking at entry-level DSLRs last year I was torn between the Nikon D3200 and Pentax K-30. I went with the Nikon for a couple of reasons, the key one being the 24MP count.

Subsequently, I've realised the D3200 did not do half the things I wanted (even as a newbie), although the image quality is fantastic. I moved to the K-30 and haven't looked back. For newbie photography like mine, the drop to 16MP seems to make no difference to image quality, which remains fantastic. The experience of the camera is, however, much greater.

[jsherman999]

I'm sure for many of you professionals and high-end amateurs the difference between 16 and 24 MP will make a difference. For me, looking at entry-level DSLRs last year I was torn between the Nikon D3200 and Pentax K-30. I went with the Nikon for a couple of reasons, the key one being the 24MP count.

Subsequently, I've realised the D3200 did not do half the things I wanted (even as a newbie), although the image quality is fantastic. I moved to the K-30 and haven't looked back. For newbie photography like mine, the drop to 16MP seems to make no difference to image quality, which remains fantastic. The experience of the camera is, however, much greater.

But if you were given the choice: K30 @ 16MP or K30 @ 24MP, same price, maybe slightly lower FPS for the 24MP version, what would you choose?

[kentishrev]

But if you were given the choice: K30 @ 16MP or K30 @ 24MP, same price, maybe slightly lower FPS for the 24MP version, what would you choose?

if given the choice, at no cost, then I guess 24MP, since it doesn't make things worse. The point I was trying to make was that it seems to a brand newbie to matter, only to find it might not really. Now, if Pentax put the 24MP into the K-30 in the first place, they might catch the first-time buyers, since this is the only thing that let them down in my first decision.