Serene Life

by garik

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

Are more megapixels good?

Are more megapixels bad?

Do more megapixels increase detail? Do they make my images sharper? On the other side, is there a point that is too much? Do megapixels cause increased noise and other problems? How does print and viewing size come into it?

Widespread Internet-forum wisdom used to be that 6 megapixels was the sweet spot — below that just wasn't enough, but above that, there wasn't much benefit. File sizes got larger, but detail was lost to noise and other problems. The arguments are that cramming too many pixels into a small sensor makes each pixel too small to provide any real benefit, and that higher-megapixel sensors outresolve cheap lenses anyway. (Substitute something like 12 instead of 6 if we're discussing APS-C DSLRs, or 24 for full-frame.)

By 2011, every camera introduced was in the range of 12-18 megapixels. In 2014, a few are still coming in at the lower end of that, but most are in the 16-24 megapixel range. Certainly none are at 6 anymore; even smartphones come with 8 megapixel cameras. Does this offer any real improvement over that old "sweet spot"? Has technology improved to the point where the "wisdom" needs to be updated, or are we all suffering for marketing? Or have we gone past the sweet spot in some ways, but it's okay because of previously-un-argued points. (For example: more noise, sure, but more detail as well.)

Within the 16-24 megapixel common-today range, for the same sensor size, is there any actual benefit to the higher end? How do megapixels directly affect image quality with today's technology? What are the benefits and when do they apply? What are the drawbacks, and when do they apply? How should I adjust my technique (and expectations) based on my camera's megapixel count?

share|improve this question
    
Can we extend this to how it also applies to something like the Nikon D3x in the 24.5 megapixel range? –  rfusca Aug 10 '11 at 14:36
    
@rfusca: absolutely! –  mattdm Aug 10 '11 at 15:14

8 Answers 8

up vote 33 down vote accepted

From a purely theoretical point of view: more megapixels good.

People often talk about how high megapixel sensors were now outresolving most lenses, thus there was no point going higher unless using the very best glass. This is not always true. System resolution is the product of lens resolution and sensor resolution. Thus if you improve one, your system resolution will improve regardless of the other. You do eventually get into diminishing returns, but from a theoretical viewpoint a sensor can't outresolve a lens until diffraction effects take over.

Theoretically for a fixed final output size, noise is independent of sensor resolution. Yes smaller pixels capture less light, therefore the per pixel noise level is higher. But if you resize a high megapixel image to match a lower one, you average pixel values and thus noise is evened out. People regularly complain about noisy high megapixel compacts when viewing images at 100%. But that's a totally unfair comparison.


From a practical point of view: more megapixels not bad

From a practical view the noise situation is more complicated, but evidence I've seen suggests that high MP sensors are not much noisier when compared at the same image size (see above). I'll look up some links.

The situation on resolution is complicated by the fact [most] sensors don't see in colour and thus have a bayer grid which requires an anti aliasing filter. Aliasing is worst when the sampling frequency matches your signal (i.e. image detail) frequency. Increasing the megapixel count faster than increases in signal frequency should improve aliasing, to the point where the traditional aliasing filter can be removed.

There are other practical issues which relate to your ability to extract extra detail from your sensor:

  • The 1/focal length rule no longer applies as you increase megapixels, you need ever increasing stabilisation, and also increasing shutter speeds as subject motion becomes more apparent.

  • Diffraction becomes more of a problem as you increase megapixels as the pixels become smaller than the Airy disk.

  • Data processing and storage requirements are higher.

It's worth emphasising that these are not disadvantages of higher megapixel counts, since you can always downsize your images, and you're not lost anything when compared to a lower megapixel count camera. The exception being in camera data processing, since the camera has to read the whole sensor when shooting stills and somehow process this information.


So how high can you go? I've seen calculations of the diffraction limiting aperture for red light with a 350 megapixel full frame sensor being f/2.8 (green and blue light requiring even larger apertures) so that gives you an idea. Personally I think your returns would get small past a 50 megapixel 35mm sensor, up to a maximum of maybe 75-100. Once you get noticeable diffraction at f/5.6 people are going to become disinterested, and once you have to open up to f/2.8 with a lens that's razor sharp at f/2.8, the megapixel race is over.

Larger formats allow more megapixels before diffraction sets in (at a given f/stop) however depth of field is shallower at the same f/stop, requiring you to stop down more for depth of field, so there appears to be no intrinsic advantage when it comes to diffraction (though it's easier to make lenses that are sharp at the diffraction limiting aperture for a larger format).

The existence of 80 megapixel medium format cameras points to the fact it would be possible, diffraction wise, given good enough glass. Though as users of such cameras point out how difficult it is to utilise 80MP this points to it being a good practical limit, if not a theoretical one.

share|improve this answer
    
The 1/focal length rule not applying would only really occur when pixel peeping as well - right (so the 'logic' is the same as the noise argument?)? Also, if you've got a larger megapixel camera with the intention to be able crop or in order to make larger pictures, the noise argument definitely still applies? –  rfusca Aug 10 '11 at 15:23
1  
@rfusca The main point is all these things (noise etc.) are not disadvantages of a higher MP camera, as some claim. You wouldn't say a 20MP camera is worse than a 10MP because of extra camera shake, because said camera shake goes away upon resizing to 10MP. They are however impediments to getting a better image out of a higher MP camera - so they are relevant, as you say, if you buy a high MP camera in order to crop more or print larger. –  Matt Grum Aug 10 '11 at 15:46
1  
ah, i'll take it like that then :). So the final conclusion is basically...more megapixels isn't worse than less, but its not necessarily better in a practical sense. –  rfusca Aug 10 '11 at 15:54
1  
Very long post... then I have lots of comments. ;-) “Theoretically for a fixed final output size noise is independent of sensor resolution.” +1 because most people miss this important point. –  Edgar Bonet Sep 23 '11 at 13:03
    
“System resolution is the product of lens resolution and sensor resolution”. By “resolution” you probably mean MTF. However, sensor resolution is not well described by an MTF: it's rather a hard cutoff at the Nyquist frequency. Therefore, in this case system resolution is the smallest of lens resolution and sensor resolution. However, when you add the low-pass filter to the equation, the product of MTFs becomes relevant again. –  Edgar Bonet Sep 23 '11 at 13:04

In general, more megapixels are better. However, there is more than one factor in the sharpness of a camera's images. For example, if you have 500 megapixel images, the lens will still limit the sharpness of the images to something far lower in resolution. Many point and shoot cameras have enough megapixels and cheap enough lenses that the limited factor is clearly the lens rather than the number of pixels.

Some minor disadvantages of large numbers of megapixels are slower transfer from the camera to computer, and larger files on the computer and memory card. You can generally adjust the image size smaller on a camera, but this may not affect Raw images.

Other than image size inconvenience and additional cost, too many megapixels won't hurt anything.

share|improve this answer

A different spin on the more megapixels question is not "is the edge to edge image clarity better" but "is there something I could do with the extra bits"? One thing I'm seeing more and more is the flexibility to repurpose images by cropping simply because a cropped image still has sufficient resolution for many, if not most, purposes.

And... if/when lenses match sensor resolution for the higher-end sensors (I'm not convinced all lenses are outstripped by the top DSLR sensors), then you will probably be happy to have the extra resolution.

Echoing the "more pixels equals slower to save to media" sentiment, this can be a problem when shooting action and in at least one other (edge) case: hand-held HDR.

share|improve this answer

I have long been of the opinion that for normal users, more megapixels gives you no 'real world' benefits.

The pros of more megapixels:

  • Allows you to print larger, without loss of detail.
  • Allows you to crop an image without loss of detail. ( <-- Personally one I use a lot )
  • Allows you to really pick out lots of small details in review.
  • Records more information in a scene.

The cons of more megapixels:

  • Increased information captured in a scene can mean that in areas of contrast such as the edge of something, so much information is captured that it does not quickly 'change' from one to the other. This can mean upon reviewing images, that they look softer, and not as sharp as photos from a lower-megapixel camera. My experience of this is the 8mp EOS 350D vs. the 18mp EOS 7D. The 7D is technically superior in every way but even with the same lenses, the photos taken with the 350D look sharper.
  • More megapixels = larger file sizes, especially RAW. So more disk space may be necessary, as well as your CPU working harder (/longer) to load them.
  • Some lenses will start to show flaws with such high resolution cameras, so to get the best out of the camera, you'd need to invest in more expensive glass.
  • Low light performance can start to suffer, as the pixels are packed more densely onto the sensor, leading to more noise on longer exposures.

So, unless you are taking a photo with the intent to blow it up to A1 size, or beyond (advertising hoardings, etc), then you simply do not need the megapixels. For regular viewing on-screen or printing for a family album, lower megapixel (8-12) will be more than enough and give you sharper results.

Note that these sentiments are based only on my own observations in real-world usage. They are not scientific in any way...

share|improve this answer
    
Another highly technical disadvantage of packing more sensor elements into a fixed size package is that they can start to interfere, causing trouble. And of course more elements means more heat generation (unless technology allows for either better heat dissipation or lower power consumption). –  jwenting Apr 26 '12 at 9:23

Sometimes more pixels is bad.

You want better pixels, not more of them, depending on the sensor size. You need big enough sensors that you can capture sufficient numbers of photons.

While sensors are getting smaller, Moore's law and all that, photons are not.

share|improve this answer

there is an excellent on line publication put out by Nikon that addresses this issue very well. it is titled 'nikon d800/d800e technical guide'. it describes how to best use/shoot these two 36mp cameras. although directed to a specific camera the point of the article is clearly made through comparison photos and corresponding detailed text that addresses whatever level you are at in you photographic experience. although you might expect it to be a "commercial" for Nikon, it is not. very good unbiased article.almost, but not quite, it really made clear the fact that some of us may have to learn how to shoot a high mp camera.

tom hill

share|improve this answer
    
This is kind of a teaser, not really an answer. What conclusion do you draw from the article? –  mattdm Dec 6 '13 at 14:05
    
If the main point of your answer is in some external document, you should at least link to it. As it is, this answer by itself doesn't provide much information. –  Olin Lathrop Dec 6 '13 at 14:30
    
my response was intended as a comment on the subject, not an answer. I don't know if anyone has an answer that all would agree on. the link to the site is the title I provided. I always thought it was better to provide as much info as possible so as to make as well informed opinion as we can. tom –  tom hill Dec 15 '13 at 23:13

I fully agree with Matt's and Steve's answers, but I think one also needs to consider the enormous advantages of having a higher resolution when doing post processing on images. In general, more megapixels will yield much better images if you try to get the most out of post-processing (provided, of course, you are not comparing a bad camera with a large number of megapixels to a good camera wih a smaller number of megapixels).

Pictures taken with the best camera by good photographers will often have a bit more resolution than is justified by the image quality. Also, as Matt points out, even a perfect shot with a perfect lens would have unsharpness due to diffraction if the resolution is high enough. So, you can question if increasing the resolution by, say, a factor of ten would be of any use if typical pictures already show unsharpness on length scales longer than the pixel size.

Let's consider Mike's point about unsharpness in areas of contrast. Suppose that the brightness changes rapidly from one value to another value over 4 pixels and with ten times the resolution this change would then have happened more gradually over 40 pixels. Given the latter image, I can calculate the point spread function much more accurately as I have ten times the number of data points. This would allow me to sharpen the image using deconvolution much more accurately. Deconvolution will generate artifacts, the more accurate you can do the computations the better the balance between recovered details and spurious artifacts are.

Another application is to correct for overexposure when the brightness value of small parts of the image is clipped to the maximum (also in the raw image). Without the benefit of multiple pictures with different exposures, you cannot construct a HDR image. But what you can do is consider the edges of the overexposed area in the picture and then with the local point spread function (which one can try to calculate from nearby high contrast areas), you can calculate the correct brightness (also gradients and higher order derivatives) just into the overexposed areas.

Obviously this will work much better the more pixels you have to work with and when the image isn't perfectly sharp. While the results of such an exercise will be quite limited ( you won't be able to restore details well inside the overexposed area), I think just getting the color right of an ugly white overexposed area can make this worthwhile.

share|improve this answer

It seems that no-one have touched on the issue of light-sensitive area. Sensors can either be front- or back-illuminated and this will result in different effects when increasing the number of pixels.

Front-illuminated sensor

A Front-illuminated sensor will have transistors and electrical paths on the light-sensitive side of the sensor. These components will cover parts of the sensor, and reduce the light-sensitive area. Adding more pixels means more transistors and reduced light-sensitive area.

Smaller light-sensitive area results in lower performance.

This can be mitigated somewhat by using microlenses.

Back-illuminated sensor

Since the transistors for each pixel and the electrical paths are not on the same side as the light-sensitive area, back-illuminated sensors will have the same light-sensitive area, even thought the number of pixels is increased.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.