Home > smugmug > The Megapixel Myth – You're getting ripped off.

The Megapixel Myth – You're getting ripped off.

February 8, 2007

Megapixels don’t matter. There, I’ve said it. Throw your stones, rotten vegetables, or what-have-you.

David Pogue has an article in the New York Times today about the Megapixel Myth. His heart is in the right place, and he comes to the correct conclusion, but he’s fairly sparse on the details and the proof is a little hard to grok without actually being there.

News.com ran a story this week quoting Chris MacAskill, my father and SmugMug’s President. He comments on how the word megapixel is a marketers’ dream, and he’s right. Anyone marketing a camera based solely on megapixels is ripping you off. But their article wasn’t clear enough either. So here it is, as clear as I can make it:

What you *really* want are “better” pixels, not more of them. Contrary to popular belief, adding more pixels to the same size sensor isn’t going to help you very much. In fact, it may hurt. Why? It’s reallly quite simple: A limited amount of light gets through the lens and hits the sensor in your camera. It’s really quite small. At some point, if you cram more pixels into the same tiny space, those pixels aren’t picking up enough light to be useful. Instead, they’re making your photo noisier and reducing the quality.

The real way to get better images is to have a bigger sensor (so it can capture more light) and a better lens (so more light gets to the sensor). It’s that simple.

Don’t believe me? NASA’s Spirit Rover has a 1 megapixel camera and it takes better photos than any 8 megapixel camera you can buy at Best Buy. MSNBC has a great article on how it works and sample images. See for yourself.

Finally, SmugMug has printed more than 3 million photos for very discerning customers. We publish the reasons why photos get returned. The number of returns for “not enough megapixels” is at or near zero.

So how do you figure out what camera to buy since megapixels don’t matter? Pogue is right on the money with this one: read reviews at sites like dpreview.com.

Categories: smugmug
  1. Douglas
    February 8, 2007 at 1:45 pm

    The photographer, camera lens, and exposure settings have more of an effect on quality than does a camera’s MP sensor.

    A general example…
    If you have a 10′ x 10′ room and you cover it with 10 1′ x 1′ tiles you can place 100 tiles. If each tile was a color you could create a limited amount of detail.
    If in the same 10′ x 10′ section you fill with 20 6″ x 6″ tiles you can place 400 tiles. Which adds twice as much details.

    Therefore, to increase the resolution perceptibly you need to quadruple the pixel count. Yes, there is a difference in detail between a 3 MP camera and a 12 MP camera if all other things are relatively equal. But no so much between a 5 and a 7 MP camera.

    The distribution of dots on an inket print makes it harder to evaluate, but I’ve foiund it to be noticable on higher resolution monitors, especially when photo retouching.

    Hope this info is helpful!

  2. February 8, 2007 at 1:48 pm

    This is going to sound like a spammy comment, but I swear it’s not.

    I have been trying to make this point for about 2 years now on my website.

    I have included tons of comparison pictures that very clearly show how lens quality, sensor size, and things like image stabilization, do MUCH more to increase the amount of detail captured by a photograph than does the number of pixels.

    So thanks for helping to communicate the truth that more megapixels are not always better. But if you want to see it communicated with images (which I personally think is more effective), then check out




  3. February 8, 2007 at 2:02 pm

    My immediate response is always frustration with these kinds of articles. I *know* that you have made generally factual statements, but the premise is untrue from the perspective of a Systems Engineer who was a commercial photographer for eight years. Let me explain –

    When I compare pixel counts, I try and base it on equal image creation circumstances; dpreview does that as well – their test images are remarkably consistent and well shot. There is a visible difference in image sharpness and detail even with minor pixel count differences (ie, 6.3 MP vs 7 or 8 MP) that starts to become obvious when you begin cropping, assuming all other variables are at least similar.

    I understand what you mean, but simply because most people will accept image quality doesn’t mean it’s ‘good’. I mean, millions of people used (and loved) 110 film for years, and it sucked really really badly, even at those small enlargement sizes.

  4. February 8, 2007 at 2:06 pm

    you know I have been trying to tell people this for years. the key is to get a big sensor and a great lense. that will produce amazing photographs

  5. February 8, 2007 at 2:06 pm

    One very important fact should jump out from the original NY Times article “One print had 13-megapixel resolution; one had 8; the third had 5. Same exact photo, *down-rezzed twice*, all three printed at the same poster size.”

    That’s like saying “We took Yao Ming and Muggsy Bogues, trimmed them at the ankles and knees down to 5′ and they both played basketball just as well!”

    The point is that down-rezzing is throwing away that additional data. If he were to reprint the same series at the *maximum* DPI each picture could support you’d get a different result.

    For what it’s worth though, Nikon’s fastest digital shooter, the D2Hs only outputs 4.0mp files (sacrificing size for sheer speed) – good enough for most newspapers and magazines. For most uses, you can get away with lower megapixels – and if you don’t know *why* you need extra megapixels than you likely don’t need them at all. But for those of us that know why they need it, Pogue’s psuedo-scientific reasoning to try to convince us otherwise is somewhat annoying.

    If you’re printing 20×30 prints, a 12Mp D2Xs will print a lot better than a 4Mp D2Hs – even though for all intents and purposes they are the same camera (same sensor size, lenses, body, etc.) – but yes, the 4.0Mp D2Hs likely takes better shots than any point-and-shoot 7 or 8Mp camera.

  6. David
    February 8, 2007 at 2:16 pm

    @ Steve

    More megapixels will mean a ‘better’ picture – with more detail – on a same size sensor only if you are on a tripod, and use a longer exposure time for the picture.

    Now, if you are talking the difference of a full sun shot of 1/500th versus 1/750th, you won’t notice the difference in the picture quality. However, if you’re taking picture in soft light and require 1/50th vs. 1/30th its going to make a huge difference.

  7. February 8, 2007 at 2:22 pm

    Problem is, megapixels is a neat, easy-to-remember spec. My daughter emailed me last month, “Dad, I heard megapixels aren’t important any more; so what spec do I look for instead?” And I replied, “Ummmmm…” When the manufacturers print the sensor size or lpi of the lense resolution in large clear letters in their ads, them megapixels will properly get ignored.

  8. February 8, 2007 at 2:34 pm

    Thank you for this information, at my work, we were debating this very topic for our termite and wdo inspectors who wanted cameras to also document their inspections with pictures and we couldn’t come to a consensus for what we needed in a camera.

  9. February 8, 2007 at 2:43 pm

    Wow how amazingly silly. The quality of an image is a result of multiple factors and Megapixels is one of them. It is true that if your CCD/CMOS is tiny adding more pixels won’t help, it is true that if your lens sucks, adding more pixels won’t help. But it is also true that if you have a large sensor and a good lens, megapixels matters.

    Lens quality, Sensor size, pixel density, sensor sensitivity are the technical factors that go into a good image. It is true that a 4 Megapixel Camera phone isn’t going to make an image much better than a 2 Megapixel Camera Phone. But an 8 Megapixel Canon Rebel sure will, and a 31 Mega Pixel H3D-31 will make you drool.

  10. February 8, 2007 at 2:59 pm

    I found this article to be inaccurate and even a bit careless piece of work, like an attempt to hastily impress army of down-to-earth heaps of amateurs.

    It was useful to lace their boots down on the ripp-offs by manufacturers of many point-n-shoot cameras who claim the useless tens Mega Pixels at pretty small camera sensors.
    Still it lacks integrity to be the really useful source of information.

  11. February 8, 2007 at 3:01 pm

    To me, the megapixel race is like the horsepower races you read in car magazines. “Oh really? That Ferrari has 450 horsepower? I guess that means this other car with 350 horsepower must suck.” Meanwhile, 98% of the people reading those magazines are driving 12 year old Nissan Sentras and make under 20 grand a year. Meaning it has no bearing on reality. Sure, to professional photographers, more megapixels are probably a positive attribute in a camera. But if they’re professionals, they wouldn’t buy a camera based on megapixels anyway. Now, when my mom goes into Best Buy and some dickweed tries to sell her an 8 megapixel camera because the 2 megapixel one she has (and knows how to use and won’t call me about) isn’t good enough to take some shots at Disney World, then we have a problem. My advice? Let’s kill Best Buy!

  12. J. D
    February 8, 2007 at 3:24 pm

    *cough* bullshit *cough*
    For consumers yet .. thats about it.

    If you are shooting on a 35mm digital camera it must an approved camera from this list: Nikon D200, Nikon D2X, Canon EOS 30D, Canon EOS 5D, Canon EOS 1D MK 11, Canon EOS 1Ds, Canon EOS 1Ds MK 11. All medium format backs (e.g. backs by Phase One and Leaf etc) produce sufficiently high quality images to be accepted by us.

    Don’t see anything under 8MP on that list …


    The key is you want a good deal of mega pixels on a big enough sensor to be able to crop.

  13. MD
    February 8, 2007 at 3:28 pm

    Like most statements, the real answer is “yes and no”.

    If you’re doing termite inspections, more megapixels helps because you can blow up the picture to look at details – which is the job. If you’re doing posters, same thing… There was an modern artist in the 70’s who would do portraits about 10 feet square in excruciating detail (using 8×10 film for the original to work from).

    My first digital was the Fuji4900Z, which was 2.4Mp upsized to 4Mp. 4×6 prints from it I described as “painfully sharp”. 20×30 posters – not. However, publishers will also point out that what’s acceptable for a screen is rarely sufficient for a commercial printed product. 8×10 at 200dpi is 3Mp or so, the lower limit for commercial photography?(8x10x300dpi, 7Mp) Coffee table books or weddings? There’s a reason the pro’s used Hasselblads and view cameras.

    One point made in the article was “more pixels on same size sensor”. Yes, unless the technology improves, noise increases. Similarly, if the lens construction does not support the correct sharpness, no real benefit. I have yet to see a real discussion on (a) whether the noise issue is a permanent physics problem, or just the limit of the technology, waiting for a technical breakthrough; (b) whether a given camera’s lens construction is “not sharp enough” for the megapixel amount – is a brnad’s lens “circle of confusion” for a focused picture larger than a pixel sensor?

    What also is not mentioned much, is how much massaging the camera does to the original before it produces a finished JPG; especially newer, higher megapixel noisy cameras.

    I’ve never used Smugmug, but the online services I have used either warn you or prevent you from printing at too low resolution.

    Regarding Spirit Rover – if all it takes to beat an 8Mp camera is a 1Mp sensor, why wouldn’t consumer cameras do that trick – with 8Mp sensors – to give 64Mp-quality pictures. The article claims “Imax quality”; really all it is is quality lenses, and a large (hence, low noise) sesnor. I presume the “Imax” part comes from stitching, something anyone with a tripod and software can do.

    The problem is, beyond a certain point, the quality of lens is irrelevant. If it focuses so the circle of confusion is say, less than half a pixel sensor, your picture is as sharp as it can get. A bigger sensor is good, but means you need a much bigger lens. Too low light (ISO 800 or so) or too long exposure (seconds) creates noise, so you want to collect a lot of light in a short time.

    Overall, the real answer is always more complicated.

  14. February 8, 2007 at 3:40 pm

    With one shot of my Olumpus SP320 (7.1 mpxl) I can outdo 98% of other point and shoot photographers because I took the time to learn my camera and its capability. Sure, I’ve seen Canon 3.2 mpxl cameras take good photos, but can they crop and edit them to the same extent? I’m not a chip engineer, but I say megapixels make a difference in a quality sensor with a reasonable lens. It is intuitive to me that with equal lenses, more megapixels gives better image resolution and sharpness, and probably uses more battery power to boot. Hmmm. Now there’s a possibility — maybe more pixels uses more batteries. How about that, pink bunny?

  15. Who Cares
    February 8, 2007 at 3:42 pm

    Thanks for that information

  16. Edo
    February 8, 2007 at 5:10 pm

    Actually NASA’s Spirit Rover camera requires three images to produce one color picture, so those pictures have 3 megapixels.

  17. Solo
    February 8, 2007 at 11:05 pm

    It’s not surprising that the megapixel term was hijacked by the marketing dept. More of something gives you some advantage. And like everything computer related, you know the next generation will be better.

    They’ve been very successful to associate more MP = better, because more MP = newer gen.

    But it is not the only market that will focus on just ONE measure. Computers CPUs used to fight the Megahertz war, even though speed does not scale that well with the clock. The car industry is fighting the Horse Power war, even though who really cares if you reach 60 MPH in 7.5 seconds instead of the competitor’s 8.2.

    And so on. It’s what the marketing department does: simplify as much as possible for anyone to understand without to have to educate anyone.

  18. February 9, 2007 at 12:23 am

    blog.forret.com, I’d go with that…

    I did a pro Photoshop course a few years back with a glamor photographer (he was the tutor, not a student). He had a 4Mpx professional Canon for general shoots (a 1D I think it was) but with a decent chunk of glass out front (like $6k lenses that kind of decent) he could take photos that were perfect for the kinds of magazines to which he was contributing and could also be blown up to A3 size on a dye sub print without any fear of pixelation or artefacts being visible. With a quality sensor and a very high quality lens you will never need more than 4Mpx, unless you’re cropping in tight

  19. eas
    February 9, 2007 at 12:24 am

    What many people here defending more megapixels seem to be missing is that the lenses on small cameras don’t even have enough resolving power to make use of the resolution offered by high megapixel cameras. Check the review at DPreview.com for examples.

    The result is that your 10mpixel photos don’t have any more information in them than the cheaper 6 or 7MP with the same lens. Go ahead, crop them, blow them up. It doesn’t make any difference. The photos you get would have the same info as if you’d cropped and/or resized the lower resolution photos. Actually, the photos from the lower res camera might actually have more information because they’d have a better signal to noise ratio.

  20. carey
    February 9, 2007 at 12:52 am

    I was just wondering why you posted a digg about your blog that was about another website? why not just digg the original article and then blog your digg??

  21. February 9, 2007 at 1:04 am


    If you read my post, the entire point was that David Pogue (and News.com in a separate article) wasn’t telling the whole story, and not nearly clearly enough.

    If I had thought his post was worth digging, I would have dugg that. Since it wasn’t, I thought it was worth commenting on and clarifying the issue, given that I’m an expert on the situation.

  22. February 9, 2007 at 2:10 am

    I agree with Vance. This article takes a bit of truth, mixed with a lot of balderdash, resulting in a confusing mess that is partially true, and partially not.

    While it is true that megapixel alone does not determine how good a camera is, it is also absurd to ignore it completely. Like it or not, there is a strong correlation between megapixels and picture quality.

    I work in a photo shop, and see a lot of good and bad photos. Whenever we check, the worse photos are usually low resolution (

  23. February 9, 2007 at 5:26 am

    It is very similar to all the spam and ads for penis extenders.
    There is a small return rate as few wish to admit or even recognize that they have been ripped off.

  24. Paul Knowles
    February 9, 2007 at 1:09 pm

    So many words. so little understanding.
    Physics really does make a difference.
    You look for the diffraction limit of the objective,
    and pick the sensor accordingly.

  25. Cindy L
    February 9, 2007 at 3:08 pm

    CCD SHAPE also matters. You can keep your squares, I prefer my Fuji “Honeycombs”. Best technology on the market with their aspherical lens and faster processor in my opnion.

  26. Bob
    February 9, 2007 at 10:00 pm

    This article is wrong. The Mars Rovers do not take better pictures than a digital camera. They are not as sharp, and the are essentially black and white with a single-color filter. Maybe the pictures you are thinking of are the mosaics made of many 1mp images.

    There is a point where the pixel resolution surpasses the lens resolution, but it does not make the image worse. It just doesn’t make it significantly better.

    In general, a larger, high quality lens will utilize a higher pixel density.

  27. February 10, 2007 at 11:16 am

    That’s why I’m waiting for the price to come down a bit to upgrade my Canon XTi to a Canon EOS 5D soon

  28. Dr. Kenneth Noisewater
    May 25, 2007 at 9:55 am

    Wow, discussion of the Megapixel Myth and no mention of Foveon or Sigma?


    I don’t work for them, but I find their tech pretty darned interesting…

  29. February 21, 2008 at 11:53 pm

    NASA’s Spirit Rover is providing a lesson to aspiring digital photographers: Spend your money on the lens, not the pixels.

    Anyone who has ever agonized over whether to buy a 3-megapixel or 4-megapixel digital camera might be surprised to learn that Spirit’s stunningly detailed images of Mars are made with a 1-megapixel model, a palm-sized 9-ounce marvel that would be coveted in any geek’s shirt pocket.

    Spirit’s images are IMAX quality, mission managers say.

    The word pixel is derived from the term “picture element.” A pixel is the smallest dot of information that goes into making a digital image. One megapixel is a million pixels set up in an array equal to 1,000 by 1,000.

    Intuitively, more pixels means higher resolution. That’s generally true on a display screen. But when capturing images, where a pixel is more properly called a sensor, the count is just one of many factors that control quality.

    Seeking perfection

    The technology used to make Spirit’s Panoramic Camera, or Pancam, is essentially the same as what goes into a Casio or Pentax digital camera.

    But the Pancam’s lenses — there are two, which provides stereo imaging capability — are crafted more finely than anything you’d probably want to plunk down a Visa for. And the light-capturing chunk of silicon, called a charged coupled device, or CCD, was manufactured with no tolerance for the minor flaws that are inherent in mass-produced consumer cameras.

    Perhaps most important, the sensors on Spirit’s CCDs are bigger, explained Patrick Myles, director of corporate communication at the Dalsa Corporation, which built the CCDs for all of the rover’s cameras (Spirit has nine altogether, including hazard avoidance cameras and a microscopic imager).

    A Sony DSC-F717, with a street price of around $600, has 5.2 million sensors (or 5 megapixels) on a chip that is 8.8 by 6.6 millimeters (or .35 by .26 inches). The Pancam has just a million sensors spread across a chip that’s 12 by 12 millimeters — nearly a half-inch square.

    Each tiny Pancam sensor, measured in microns, is nearly four times as big as those on the Sony.

    In the consumer market, which Dalsa does not target, 5-megapixel cameras often use the same size CCD as a 3-megapixel camera. More pixels are simply crammed onto the same-size chip.

    “The pixels themselves get smaller,” Myles said. “This has an impact on image quality.”

    Why? For one thing, smaller pixels are less light-sensitive.

    Also, the lens quality might not support the additional pixels. As the receptors get smaller, a higher quality lens is needed to properly focus light onto each pixel. So where each pixel ought to capture different light information — say perhaps a subtle shading change on the subject’s cheek — the same information can get spread across several pixels after passing through a lower quality lens.

    20-20 vision

    The Pancam was conceived at Cornell University and built at NASA’s Jet Propulsion Laboratory. Dalsa, based in Waterloo, Ontario, Canada, makes cinema-quality video components and other high-end imaging devices and was called on to make the CCDs for the Pancam and the other cameras on Spirit and its twin rover, Opportunity.

    “They are the world’s highest performing chips in terms of light sensitivity and chip quality,” Myles said in a telephone interview earlier this week.

    Overall, how does a Pancam stack up to the typical 5-megapixel camera you might purchase at Best Buy?

    “There really isn’t any comparison,” Myles said.

    NASA officials say the camera shows what a human with 20-20 vision would see on the surface of Mars. But anyone who has zoomed in on a distant rock in one of Spirit’s color pictures would have to wonder if perhaps Superman’s vision might be a better comparison.

    Experts argue endlessly about what the human eye can actually see, however. Comparing human vision to what a camera captures “is really up to great speculation,” Myles said.

    NASA’s analogy, Myles explained, is “probably a bit of marketing spin. It helps people visualize the quality.” The height and breadth of a Pancam image is roughly equal to what a person would see, taking into account peripheral vision. And the Pancam has a human perspective. It sits atop a mast on the rover, 5 feet (1.4 meters) above the surface.

    Myles said the actual image quality probably exceeds human capabilities, especially after the image is processed and a computer is used to provide a zoom function.

    Use your Mouse to Explore the Mars Rover Spirit 360-Degree Panorama
    Image: NASA/JPL/Cornell University
    Created by David Palermo of WorldVR.com
    Presented by collectSPACE.com

    Tricks with light

    The Pancam does not make a color picture directly. Instead, it records light versus dark in shades of gray. As with other CCD cameras used in high-end astrophotography, such as on the Hubble Space Telescope, a series of filters are applied to gather multiple images that are then blended together.

    In the most basic application of this process, three images are gathered of a scene, one each recording red, green and blue light. Those are then put together with special software to create a color picture.

    A consumer digital camera uses a single coated filter to make the transition from photon reality to electrons and then digital information.

    Additionally, the Pancam swivels 360 degrees around and 90 degrees up or down, so that individual scenes can be stitched together to create a view of the rover’s entire surroundings. The pictures are expected to reveal important geologic details about rocks, and they’re also used for navigation and to pick distant science targets.

    Modern Ansel Adams

    Spirit’s pictures are said to be three times sharper than those of the 1997 Mars Pathfinder mission or the 1970s Viking landers.

    The Viking missions, as well as the Voyager missions to the outer planets, used technology similar to antiquated television vacuum tubes. CCD technology was first developed in 1969, but it took decades before arrays were big enough to be useful.

    Much of the research that ultimately led to today’s commercial digital cameras was funded by NASA. A first major step was in developing an 800 by 800 pixel array — less than a megapixel — which is what’s in the Hubble telescope.

    The Pancam results so far have mission managers ecstatic. Cornell astronomer James Bell, who led the development of the camera, called the first Spirit pictures “absolutely spectacular.”

    Nobody has argued with him.

    In fact, Steven Squyres, a Cornell professor who directs the rover science team, called Bell “the Ansel Adams of the Space Age.”

  30. December 7, 2008 at 10:32 am

    free mlm lead

  31. Bazm
    August 27, 2009 at 2:40 am

    My 5Megapixel Camera from Canon with their awesome Glass 10X macro lens beats the pictures taken by my previous 10megapixal no-name cheap camera that broke at the drop.

  32. Bazm
    August 27, 2009 at 2:41 am

    correction 12X optical lens 🙂

  1. February 8, 2007 at 3:16 pm
  2. February 8, 2007 at 5:34 pm
  3. February 8, 2007 at 6:35 pm
  4. February 8, 2007 at 7:18 pm
  5. February 8, 2007 at 8:05 pm
  6. February 8, 2007 at 10:10 pm
  7. February 8, 2007 at 11:23 pm
  8. February 9, 2007 at 12:28 am
  9. February 9, 2007 at 12:31 am
  10. February 9, 2007 at 3:53 am
  11. February 9, 2007 at 12:54 pm
  12. February 11, 2007 at 1:22 am
  13. February 14, 2007 at 12:54 am
  14. February 15, 2007 at 2:48 pm
  15. February 20, 2007 at 6:34 pm
  16. March 19, 2007 at 12:43 pm
  17. July 30, 2008 at 10:23 am
Comments are closed.
%d bloggers like this: