Megapixels How many for 17 X 22 print?

[Edalongthepacific]

I have noticed that some newer high-end cameras use full frame sensors with enhanced light gathering characteristics. However, these are "only" 12 megapixel sensors. Does better light gathering ability translate into superior dynamic range? In enlargement printing, would you rather have greater dynamic range or just more pixels? Can quality really outpace quantity when it comes to preserving detail?

[ErikKaffehr]

Hi,

In short:

DR would generally be better with large pixels.
How many pixels are needed depends on eyesight and viewing distance.

It seems that many photographers regard 12 MP as enough for normal size prints.

Prints, themselves have very small DR, having a density range of normally .05 - 2.3 which is 7.5 stops. So DR is not relevant for print, having a good DR can be most helpful in preparing a picture for printing, however.

It seems that both noise and DR are improved with downsampling.

In real world the issue may be more complex. It seems that those having both DSLRs and MFDBs see advantages of MFDBs over DSLRs event in small prints. Although
both have more than 12 MPixels. It's less than obvious where the advantage comes from. My guess is that MFDBs need less resolution (a lower linear frequency) for a certain size of detail in print. Modern MF-lenses are pretty good, so they probably have higher MTF (edge contrast) for a given size of detail. With a DSLR the MTF for a certain size of detail would be the same regardless of the number of pixels (if compare FX to FX or APS-C to APS-C).

I would recommend this article from Ctein: http://theonlinephotographer.typepad.com/t...-be-enough.html

Or my own writing here: http://83.177.178.7/ekr/index.php/photoart...xels-do-we-need.

In short it could be said that the eye can detect very fine detail, but is most sensitive to edge contrast on larger detail like perhaps 0.2 mm. So with different MPixels you are at different parts of the "Contrast Sensitivity Function" curve of the eye.

Finally, few of the pixels remain with sloppy technique. Use a good tripod, MLU, focus exactly and don't stop down to much, else the pixels are going to be wasted anyway.

Best regards
Erik

I have noticed that some newer high-end cameras use full frame sensors with enhanced light gathering characteristics. However, these are "only" 12 megapixel sensors. Does better light gathering ability translate into superior dynamic range? In enlargement printing, would you rather have greater dynamic range or just more pixels? Can quality really outpace quantity when it comes to preserving detail?

[Gemmtech]

How did we ever make 16 x 20 prints with 6mp sensors?     I have some very nice prints from my old Canon 10D, however I will agree my Nikon D300 and D700 produce sharper prints.  I still believe the composition is more important.  There's no doubt 65 MP produces a technically better image than a 12 MP but who cares?  It's like televisions, for years we all had CRTs and they looked good, then 720 and it went from there to now they are coming out with 2160 and 3D.  

Bottom line:  You can make nice 24 x 36 prints with a 12mp camera, and I'll agree the print will get technically better with a 24mp sensor, or 39mp or 65mp sensor, but you'll never notice if it's not there!    

[Jonathan Wienke]

Can quality really outpace quantity when it comes to preserving detail?

Image quality can be thought of as (pixel quantity) * (pixel quality). They are equally important; if either is lacking, the image as a whole will suffer. A 12-megapixel DSLR beats a 12-megapixel digicam because the quality of each individual pixel is better with the DSLR. But you can't take the concept to extremes and say that 2 megapixels of perfect quality will beat 12 megapixels of OK quality. You want as much of both quality and quantity as you can afford.

[madmanchan]

It rather depends on your image content. Low-res images of fluffy clouds tend to enlarge to whatever size you want ...

[MrIconoclast]

I have a very nice print that is about 20x24 taken with a 7 megapixel P&S camera.

It's the photographer, not the camera.

[Jeremy Payne]

Does better light gathering ability translate into superior dynamic range?

Compare a D700 to a D300 on DxoMark, both 12MP ... one FF, one cropped ... the answer is: "Yes"



http://www.dxomark.com/index.php/eng/Image...brand2%29/Nikon

[bjanes]

Compare a D700 to a D300 on DxoMark, both 12MP ... one FF, one cropped ... the answer is: "Yes"



http://www.dxomark.com/index.php/eng/Image...brand2%29/Nikon

It may be too obvious to comment on, but at base ISO the DRs of the two cameras are equal. At higher ISOs the larger pixel camera has an advantage. Also, the slopes of the ISO vs DR curves are different. Based on the number of photons collected, each doubling of the ISO should result in 1 stop less DR, and this is the case for the D300. However, for the D700 the curve is flat at base ISO and the slope approaches -0.5 at about ISO 700. Below ISO 700, the D700 loses 1 stop of DR at the high end for each doubling of ISO because fewer photons are collected but gains DR in the shadows because of reduced read noise at low ISO. If the D700 had a better analog to digital converter and other electronics, one could achieve a higher DR and this would be approximated by extending the slope 0.5 line back to the Y axis.

If you look at the DXO DR plot for the D3x, the slope is nearly -0.5 throughout the ISO range, showing that read noise does not vary significantly with ISO. This has some practical implications for shooting in dim light where exposure is constrained by shutter speed/aperture considerations. Under these conditions with the D700, one will get better results by shooting at ISO 700 rather than at base ISO and increasing exposure in the raw converter; the read noise decreases with increased ISO up to about 700; increasing the ISO beyond 700 decreases highlight headroom and has little effect on read noise; one would be better off increasing the exposure in the raw converter rather on the camera. With the D3x, one could almost leave the camera at base ISO and increase the ISO in the raw converter. This is what is done with some medium format backs.

For more details see Emil Martinec's discussion and look at Figure 15a. When the read noise approaches the minimum, little is gained by increasing the ISO.