15
\$\begingroup\$

Have we reached the point where where the megapixel race is more about the race, of having more then the other guy, than about image quality?

Only a few years ago 6MP was touted as the optimum number of MP that you needed to take really good pictures.

But lately, like most technology, MP have been jumping over each other in leaps and bounds.

Nikon recently released the d800 with an (in my opinion) insane 36.3MP. But fine, the d800 is a pretty high end camera, easy way to drop a few grand. But then they also just released d3200, which aimed at being an entry level 'learner' DSLR, with 24.2MP. That's twice as much as d5000 I bought two years ago.

I know that more MP is good. Higher MP = sharper image. But at what point do these increases in sharpness become negligible at best, and increases in MP count serve nothing more then bragging rights?

When you consider people have been taking gorgeous photographs for decades, that some amazing pictures were taken on early DSLRs with less then 10MP, how often is 36MP really going to be useful?

\$\endgroup\$
7
  • 1
    \$\begingroup\$ This is exactly what I was trying to get at with Do megapixels matter with modern sensor technology?. \$\endgroup\$
    – mattdm
    Apr 24, 2012 at 2:39
  • \$\begingroup\$ All I can say is that even when not viewing crazy large or 100% crops, you really can see some extra detail with the D800. \$\endgroup\$
    – rfusca
    Apr 24, 2012 at 5:23
  • \$\begingroup\$ Pixel size is more relevant to compare than megapixels, so take the square root of the number of pixels. Now you're comparing a pixel size of 3.2 with the early DSLRs, which is apparently fine, to the "insane" d800's 6 \$\endgroup\$
    – Matt Grum
    Apr 24, 2012 at 8:36
  • \$\begingroup\$ @MattGrum: I'm confused about that last statement. When you say "pixel size", do you mean pixel pitch? If so, the D800 has a pixel pitch of about 4.6 microns. Relative to other cameras: 7D = 4.3, D7000 = 4.8, 5D III = 6.2, 1D X = 6.9, D3s = 8.4. The D800 has a pixel pitch smaller than pretty much all other sensors except the 7D (and, once released, the D3200, which will have a pixel pitch of about 3.8 microns.) I've arrived at these numbers by dividing the physical height of the sensor (say 24mm, 15.7mm, 14.9mm) by the rows of pixels. I'm not really sure where the square root comes into play. \$\endgroup\$
    – jrista
    Apr 24, 2012 at 17:00
  • 2
    \$\begingroup\$ @MattGrum: Ah, yes, totally agree with you there. :) The difference between "linear" pixel count and "area" of pixels. I've made that argument so many times recently on other forums...its a concept people really don't get. Maybe we could use a blog post on the subject... \$\endgroup\$
    – jrista
    Apr 25, 2012 at 1:07

8 Answers 8

16
\$\begingroup\$

Megapixels are Necessary!

The megapixel race is certainly not "unnecessary". Consistently throughout the last decade, progress has been made on the megapixel front while consistently increasing image quality. The anecdotal adages would have you thinking that was impossible, but there are quite a few technological and fabrication improvements that have made lower noise, greater signal-to-noise ratio, and increased dynamic range possible despite shrinking pixel areas.

I think the advent of the 36.3mp Sony Exmor sensor currently used in the Nikon D800 is an exquisite example of what low-level technological improvements can do to lower noise and increase dynamic while still allowing significant increases in image resolution. As such, I think the D800 is a superb example of why the megapixel race is most definitely not over by any means.

As for whether it is just bragging rights? I doubt it. Better tools can always be used effectively in the hands of a skilled artisan. Higher resolution and more low-ISO dynamic range have some specific high value use cases. Namely, landscape photography and some forms of studio photography. The D800 is in a very unique spot, offering near-medium format image quality in a package approximately 1/10th the cost. For some studios, there is no substitute for the best, and they will use $40,000 digital medium format cameras as a matter of providing the right perception to their customers. For many other studios, however, and for many landscape photographers, the D800 is a dream come true: loads of megapixels AND high dynamic range.

No, the megapixel race is most definitely not over, and it is certainly not unnecessary. Competition on all fronts produces progress on all fronts, and that is only ever a good thing for the consumer.


Potential for Improvement

To go a little deeper than my conclusions above, there is more to the story than simply that competition on all fronts is good. Technologically, physically, and practically, there are limitations that will indeed restrict the potential gains as we continue to increase sensor pixel counts. Once we have reached those limits, useful gains at reasonable cost will have to be made elsewhere. Two areas where that can occur would be optics and software.

Technological Limitations

Technologically, there are distinct limits to how much you can improve IQ. The primary source of image degradation in sensors is noise, and there are a variety of electronically introduced forms of noise that can be controlled. I think Sony, with their Exmor sensors, is very near to reaching technological limits, if they have not already. They have utilized a variety of patents to reduce sources of noise production at a hardware level directly in their sensors. Key sources of controllable noise are dark current noise, read noise, pattern noise, non-uniformity noise, conversion (or quantization) noise, and thermal noise.

Both Sony and Canon use CDS, or correlated double-sampling, to reduce dark current noise. Sony's approach is a touch more efficient, but both use essentially the same approach. Read noise is a byproduct of amplification due to fluctuations in current through the circuit. There are a variety of patented and experimental approaches to detecting voltage variation in a circuit, and correcting it during amplification, to produce a "more pure, accurate" read result. Sony uses a patented approach of their own in Exmor sensors, including the 36.3mp one used in the D800. The other two types of pre-conversion electronic noise are pattern noise and non-uniformity noise. These are the result of discontinuities in circuit response and efficiency.

Pattern noise is a fixed aspect of each of the transistors used to construct a single sensor pixel and the electronic gates used to initiate read and signal flush. At a quantum level it is near impossible to make every single transistor exactly identical to each other, and this produces a fixed pattern of horizontal and vertical lines in sensor noise. Generally speaking, pattern noise is a minor contributor to overall noise, and is only really a problem in very low SNR regions or during very long exposures. Pattern noise can be relatively easy to remove given you approach the problem correctly. A "dark frame" can be constructed by averaging multiple samples together to create a pattern-noise template that can be differenced with a color frame to remove pattern noise. This is essentially how long-exposure noise removal works, and it is also how one can manually remove fixed pattern noise from long exposures. At a hardware level, fixed pattern noise can be mitigated by burning in a template that reverses the effects of FPN such that the differences can be added/subtracted at read time, similar to CDS, thereby improving the "purity" of pixel reads. A variety of experimental approaches to burning in FPN templates, as well as more abstract approaches, do exist today.

Non-uniformity noise, often called PRNU or Pixel Response Non Uniformity, is the result of slight variations in the quantum efficiency (QE) of each pixel. QE refers to a pixels ability to capture photons, and is usually rated as a percentage. The Canon 5D III, for example, has a QE of 47%, which indicates it is efficient enough to regularly capture 47% of the photons that reach each pixel. Actual per-pixel QE may vary by +/- a couple percent, which produces another source of noise, as each pixel may not capture the same number of photons as its neighbors despite receiving the same amount of incident light. PRNU changes with sensitivity as well, and this form of noise can become exacerbated as ISO is increased. PRNU can be mitigated by normalizing the quantum efficiency of each pixel, minimizing variation between neighbors and across the entire sensor area. Improvements to QE can be achieved by reducing the gap between photodiodes in each pixel, introduction of one or more layers of microlenses above each pixel to refract non-photodiode incident light onto the photodiode, and the use of backlit sensor technology (which moves a lot or all of the read wiring and transistors behind the photodiode, eliminating the chance that they might obstruct incident photons and either reflect them or convert them to heat energy.)

Thermal noise is noise introduced by heat. Heat is essentially just another form of energy, and it can excite the generation of electrons in a photodiode much like a photon can. Thermal noise is caused directly by the application of heat, often via hot electronic components such as an image processor or ADC. It can be mitigated by thermally isolating such components from the sensor, or by actively cooling the sensor.

Finally there is conversion noise, or quantization noise. This type of noise is generated due to inherent inaccuracies during ADC, or analog-to-digital conversion. A non-integral gain (a decimal gain with whole and fractional part) is usually applied to the analog image signal read from the sensor when digitizing an image. Since an analog signal and gain are real numbers, the digital (integral) result of conversion is often inconsistent. A gain of 1 would produce one ADU for every electron captured by a pixel, however a more realistic gain might be 1.46, in which case you might get 1 ADU per electron in some cases and 2 ADU per electron in other cases. This inconsistency can introduce conversion/quantization noise in the digital output post-ADC. This contribution to noise is pretty low, and produces a fairly fine deviation of noise from pixel to pixel. It is often fairly easy to remove with software noise reduction.

The removal of electronic forms of noise has the potential of improving the black point and black purity of an image. The more forms of electronic noise you can eliminate or mitigate, the better your signal to noise ratio will be, even for very low signal levels. This is the major front upon which Sony has made significant progress with their Exmor sensors, which has opened up the possibility of true 14 stop dynamic range with truly stunning shadow recovery. This is also the primary area where many competing sensor fabrication technologies are lagging behind, particularly Canon and medium format sensors. Canon sensors in particular have very high read noise levels, lower levels of QE normalization, lower QE overall, and only use CDS to mitigate dark current noise in their sensors. This results in much lower overall dynamic range, and particularly poor shadow SNR and shadow DR.

Once all forms of electronic noise are mitigated to levels where they no longer matter, there will be little manufacturers can do to improve within sensors themselves. Once this point is reached, then the only thing that will really matter from a per-pixel quantum efficiency standpoint is pixel area...and with near-perfect electronic characteristics, we could probably stand pixels sizes considerably smaller than the highest density DSLR sensors today (which would be the Nikon D800 with its 4.6 micron pixels, the Canon 7D with its 4.3 micron pixels, and eventually the Nikon D3200 with 3.8 micron pixels.) Cell phone sensors use pixels around the 1 micron size, and have demonstrated that such pixels are viable and can produce pretty decent IQ. The same technology in a DSLR could go even farther with maximal noise reduction, so we really do have a long ways to go.

Physical Limitations

Beyond technological limitations to the perfection of image quality, there are a few physical limitations. The two primary limitations are photon noise and spatial resolution. These are aspects of physical reality, and are things we really don't have much control over. They cannot be mitigated with technological enhancements, and are (and have been) present regardless of the quality of our equipment.

Photon noise, or photon shot noise, is a form of noise due to the inherently unpredictable nature of light. At a quantum level we cannot exactly predict what pixel a photon might strike, or how frequently photons might strike one pixel and not another. We can roughly fit photon strikes to a probability curve, but we can never make the fit perfect, so photons from an even light source will never perfectly and evenly distribute over the area of a sensor. This physical aspect of reality produces the bulk of the noise we encounter in our photographs, and amplification of this form of noise by the sensor's amplifiers is the primary reason photos get noisier at higher ISO settings. Lower signal to noise ratios mean there is less total signal range within which to capture and amplify photons, so a higher SNR can help mitigate the effects of photon noise and help us achieve higher ISO settings...however photon noise itself can not be eliminated, and will always be a limitation on digital camera IQ. Software can play a role in minimizing photon shot noise, and as there is some predictability in light, advanced mathematical algorithms can eliminate the vast majority of this form of noise after a photo has been taken and imported in a RAW format. The only real limitation here would be the quality, accuracy, and precision of the noise reduction software.

Spatial resolution is another physical aspect of two dimensional images that we have to work with. Spatial frequencies, or two dimensional waveforms of varying luminosity, are a way of conceptualizing the image projected by a lens and recorded by a sensor. Spatial resolution describes the scale of these frequencies, and is a fixed attribute of an optical system. When it comes to sensors, spatial resolution is a direct consequence of sensor size and pixel density.

Spatial resolution is often measured in line pairs per millimeter (lp/mm) or cycles per millimeter. The D800 with its 4.3 micron pixels, or 4912 rows of pixels in 24mm of sensor height, is capable of 102.33 lp/mm. Intriguingly, the Canon 7D, with its 3456 rows of pixels in 14.9mm of sensor height, is capable of 115.97 lp/mm...a higher resolution than the D800. Similarly, the Nikon D3200 with 4000 rows of pixels in 15.4mm of sensor height will be capable of 129.87 lp/mm. Both the 7D and D3200 are APS-C, or cropped-frame sensors...smaller in physical dimensions than the full-frame sensor of the D800. If we were to keep increasing the number of megapixels in a full-frame sensor until they had the same pixel size as the D3200 (3.8 microns) we could produce a 9351x6234 pixel sensor, or 58.3mp. We could take this thought to the extreme, and assume it is possible to produce a full-frame DSLR sensor with the same pixel size as the sensor in the iPhone 4 (which is well-known to take some very good photos with IQ that, while not as good as from a DSLR, is more than acceptable), which is 1.75 microns. That would translate into a 20571x13714 pixel sensor, or 282.1mp! Such a sensor would be capable of 285.7 lp/mm spatial resolution, a number that, as you'll see shortly, has limited applicability.

The real question is whether such resolution in a DSLR form factor would be beneficial. The answer to that is potentially. The spatial resolution of a sensor represents an upper limit on what the entire camera could be possible of, assuming you had a corresponding lens capable of producing enough resolution to maximize the sensor's potential. Lenses have their own inherent physical limitations on the spatial resolution of the images they project, and those limitations are not constant...they vary with aperture, glass quality, and aberration correction. Diffraction is another physical attribute of light that reduces the maximum potential resolution as it passes through an increasingly narrow opening (in the case of a lens, that opening is the aperture.) Optical aberrations, or imperfections in the refraction of light by a lens, are another physical aspect that reduces the maximum potential resolution. Unlike diffraction, optical aberrations increase as the aperture is widened. Most lenses have a "sweet spot" at which point the effects of optical aberrations and diffraction are roughly equivalent, and the lens reaches its maximum potential. A "perfect" lens is a lens that does not have any optical aberrations of any kind, and is therefor diffraction limited. Lenses often become diffraction limited around roughly f/4.

The spatial resolution of a lens is limited by diffraction and aberrations, and as diffraction increases as aperture is stopped down, spatial resolution shrinks with the size of the entrance pupil. At f/4, the maximum spatial resolution of a perfect lens is 173 lp/mm. At f/8, a diffraction limited lens is capable of 83 lp/mm, which is about the same as most full-frame DSLR's (excluding the D800), which range from about 70-85 lp/mm. At f/16 a diffraction limited lens is capable of a mere 43 lp/mm, half the resolution of most full-frame cameras and less than half the resolution of most APS-C cameras. Wider than f/4, for a lens that is still affected by optical aberrations, resolution can quickly drop to 60 lp/mm or less, and as low as 25-30 lp/mm for ultra fast wide angle f/1.8 or faster primes. Going back to our theoretical 1.75 micron pixel 282mp FF sensor...it would be capable of 285 lp/mm spatial resolution. You would need a perfect, diffraction-limited f/2.4 lens to achieve that much spatial resolution. Such a lens would require extreme aberration correction, greatly increasing cost. Some lenses do exist that can achieve nearly perfect characteristics at even wider apertures (a specialized lens from Zeiss comes to mind that is purportedly capable of about 400 lp/mm, which would require an aperture of about f/1.6-f/1.5), however they are rare, highly specialized, and extremely expensive. Its a lot easier to achieve perfection around f/4 (if the last several decades of lens production are any hint), which indicates that the maximum viable, cost-effective resolution for a lens is about 173 lp/mm or a touch less.

When we factor in physical limitations into the equation of when the megapixel race will be over, we find that (assuming near technological perfection) the highest cost-effective resolution is about 173 lp/mm. That's about a 103mp full-frame or 40mp APS-C sensor. It should be noted that pushing sensor resolution that high will only see the benefits at an increasingly narrow band of aperture around about f/4, where lens performance is optimal. If the correction of optical aberrations becomes easier, we may be able to achieve higher resolutions, pushing 200 lp/mm, but again, such resolutions would only be possible at or near maximum aperture, where as at all other apertures the overall resolution of your camera will be lower, potentially far lower, than what the sensor itself is capable of. Significantly outresolving the lens leads to perceptual issues, namely the perception that photographs taken at other than the ideal aperture appear soft, lacking sharpness.


So when does the megapixel race end?

Answering this question is not really something I believe anyone is qualified to answer. Ultimately, its a personal choice, and will depend on a variety of factors. Some photographers may always want the potential that higher resolution sensors can offer at ideal aperture, so long as they are photographing scenes with increasingly fine detail that necessitates such resolution. Other photographers may prefer the improved perception of sharpness that is achieved by improving the characteristics of lower-resolution sensors. For many photographers, I believe the megapixel race has already ended, with around 20mp in a FF DSLR package is more than enough. Further still, many photographers see image quality in an entirely different light, preferring frame rate and the ability to capture more frames continuously at a lower resolution paramount to their success as a photographer. In such cases, its been indicated by many Nikon fans that around 12mp is more than enough so long as they can capture 10 frames a second in sharp clarity.

Technologically and physically, there is still a tremendous amount of room to grow and continue making gains in terms of megapixels and resolution. Where the race ends us up to you. The diversity of options on the table has never been higher than today, and you are free to choose the combination of resolution, sensor size, and camera capabilities like AF, ISO, and DR that fit your needs.

\$\endgroup\$
7
  • \$\begingroup\$ Once we get to a stage of being able to take an image for a 14x48 foot billboard at 300dpi with 2400mm equivalent digital zoom, I can't see the race ending before this, and it may continue afterwards. As far as I can tell this equates to 14*12*300 * 48*12*300 * (2400/35)^2 / 1,000,000 = 40,950,638 megapixels. If you left off the digital zoom requirement this would still be 8709 megapixels. At 8709MP a full frame sensor being 36mm across would have a pixel width of about 208nm. 2012 Intel CPUs use 22nm technology. \$\endgroup\$ Jan 16, 2014 at 7:32
  • \$\begingroup\$ ...continued. I realise that visible light waves are larger than this at about 390nm ~ 700nm. But we still have a fair way to go before this it critically limiting. \$\endgroup\$ Jan 16, 2014 at 7:48
  • \$\begingroup\$ I am not sure what you mean about digital zoom. That is basically enlargement in post, and it wouldn't get you anywhere remotely close to 300ppi at 14x48 feet. I mean, you could certainly do that...but there isn't any point in doing it...you would just have massively blurry image detail. Might as well stick to printing at 15ppi. As for the pixel pitch, once they reach 700nm, they are filtering red light. By 550nm they are filtering green light, and by 460nm they are filtering blue light. There won't ever be 208nm pixels for visible light. \$\endgroup\$
    – jrista
    Jan 16, 2014 at 8:21
  • \$\begingroup\$ Regarding where pixel sizes are today...the next generation of small form factor sensors will be using 0.95µm pixels...that is 950nm. The next generation after that is probably 825nm, after which we reach that wavelength limitation...I don't think we will see 700nm pixels in any sensor. Granted, these pixels aren't going to be used in FF or APS-C sensors for a long time to come, but technologically speaking, we already are getting pretty close to the megapixel limit (referring to pixel pitch.) Finally, it isn't really logical to apply CPU transistor sizes to pixel pitches. Intel uses 22nm... \$\endgroup\$
    – jrista
    Jan 16, 2014 at 8:26
  • \$\begingroup\$ ...transistors. Pixels are different. Pixel area is critical to light gathering capacity, which directly relates to noise levels. A 22nm pixel is simply illogical. Sensor transistor sizes are already getting pretty small. Canon still uses 500nm, but the last generation used 180nm transistors, and newer generations are using 90nm and some even 65nm. The next stops for sensor transistor size are 45nm and maybe 32nm (although I really don't expect to see 32nm in use until 825nm pixel pitch, if we even see it there, as it isn't necessary with BSI.) \$\endgroup\$
    – jrista
    Jan 16, 2014 at 8:31
7
\$\begingroup\$

Storage/speed issues aside, having more megapixels is going to make absolutely every single picture you take better. Maybe only a little better in some cases, but that sounds like something worth having to me.

If you've ever had an image suffer from Moire (colour banding patterns):

Maze artefacts:


(source: gol.com)

Aliasing:

http://cdn.asia.cnet.com/i/r/2005/dc/39095631/4.jpg

Colour fringing, false detail, lack of colour detail or any other demosaicing artefacts then your problems would be solved if you had more megapixels.

Eventually I see 80-100 MP DSLR sensors, at this point you're not going to want to store every pixel every time but a reduced resolution downsampled RAW mode, like Canon's mRAW will provide you with an image with exceptional colour detail similar to what is achievable with Foveon sensors but at much higher resolution.

\$\endgroup\$
7
  • 1
    \$\begingroup\$ One note about Canon's s/mRAW. I used those formats for a couple months after buying my Canon 7D. While they are called RAW, they are a very far cry from the actual native RAW format from a post-processing perspective. When processing an mRAW file, I noticed remarkable limitations on how far I could push exposure, saturation, toning, etc. around compared to the native raw. In many cases, the mRAW failed miserably when trying to recover highlights or lift shadows. Even with a 100mp sensor, I would still always prefer the native RAW, as pre-interpolating the pixels imposes many limits. \$\endgroup\$
    – jrista
    Apr 24, 2012 at 17:09
  • \$\begingroup\$ "Colour fringing, false detail, lack of colour detail or any other demosaicing artefacts then your problems would be solved if you had more megapixels." I always assumed colour fringing is produced by the lens, not the sensor: how would higher sensor resolution solve this? Wouldn't it make it "worse" instead, i.e. pushing the limits of lenses so artifacts and general optical defects are more visible? \$\endgroup\$
    – MattiaG
    Apr 25, 2012 at 16:23
  • \$\begingroup\$ @MattiaGobbi: He is referring to demosaicing artifacts, which includes a form of color fringing that results from very basic demosaicing algorithms, not the color fringing produced by lens aberrations. \$\endgroup\$
    – jrista
    Apr 26, 2012 at 4:00
  • \$\begingroup\$ @jrista - Thanks, I'll have a look into this. I can't but think demosaicing in its basic form is supposed just to make the image softer, as three pixel out of four in the final image have, in a way of speaking, color which is the average of the colors of surrounding pixel. This also accounts for low color accuracy on edges. Could more complex artifacts be generated by algorithms intended to enhance sharpness and color within the demoisaicing process? \$\endgroup\$
    – MattiaG
    Apr 26, 2012 at 12:42
  • \$\begingroup\$ @MattiaGobbi: The purpose of demosaicing is not to make the image softer...its to interpolate the individual color channels from a bayer sensor into RGB pixels. There are quite a few demosaicing algorithms. One of the most common is AHD demosaicing, which is a weighted algorithm that eliminates most color fringing and produces pretty sharp results. There are a variety of other approaches as well that are used in open source RAW editors and astrophotography tools that are either faster, more accurate, designed to extract as much detail as possible, etc. \$\endgroup\$
    – jrista
    Apr 26, 2012 at 19:02
2
\$\begingroup\$

I'm not absolutely disagreeing with what others have said, but the answer in part depends on what you most value. I am most interested in high ISO low noise performance with pixel resolution being important but secondary. Others have very different priorities. I have an A77 24 MP APSC camera which is at about the leading edge of APSC mp performance, but noticeably behind some APSC cameras in areas that I care most about.

After having looked at results from the D700, D3, D3s, D3x, 5DMkII, 5DMkIII, A800 and D4 my conclusion is that at present the megapixel race has run ahead of high ISO performance and that at present for my purposes the "best performing" camera is the October 2009 released Nikon D3s. According to numbers nothing else quite matches it, and according to how I understand performance actually works in the real world, nothing else comes close.


The following sort of material tends to produce flame wars. I am attempting to simply describe what I see. Other peoples' eyes may work differently :-).

I am personally disappointed in the D800 and it's 36 mp sensor. I had been hoping for something that was a clear head and shoulders above the D700 and that might gently unseat the D3s.

The DXOMark sensor ratings low-light ISO assessment

is by no means the ultimate guide to how well a camera does in such situations in real world conditions but is a good guide to what may reasonably be expected. The rating states an ISO setting at which the camera just passes 3 minimum requirements.

The 4 year old D700 has a DxO sensor low ISO rating of 2303 ISO and the D800 is rated at 2853 ISO. The new D4 is rated at 2965 ISO and the once and still king of this measure is the (becoming legendary) D3s at 3253 ISO. BUT these ratings are adjusted to a standard 12 mp image size with the ISO rating on test being scaled by a factor of square_root(megapixels/12 megapixel). Conversely to get what they saw on test you scale the raing DOWN by sqrt(12/mp). So the D800 with 36 mp is a factor 0f sqrt(36/12) = sqrt(3) = 1.732 higher on the reporting chart than actually measured. So they measured it as 2853/1.73 =~ 1650 ISO. The justification given for the scaling is that the 'noisiness' in an image is mathematically reduced by downsampling due to averaging of information in adjacent cells. In theory scaling by a factor related to sqrt(megapixels) makes sense. But, when looking at images I'm not convinced. They are saying that a camera with grater absolute noise to signal ratios per pixel but more mp will produce an improved result when downscaled. Maths says yes. Eye brain system says that the effect is far less than the scaling suggests. I could probably dig up the specific examples that I drew these conclusions from some while ago, but this is subjective and there are enough comparisons around to allow each to find their favourite version.

The EOS 5D MkII (NOT III) has a DXO ISO rating of 1815 against ISO 2303 for the D700. But comparison of images of identical scenes taken in identical lighting conditions with equivalent lenses at high ISO settings and converted to the same image size shows an extremely significant difference between the two. So great that I'd not consider a 5DkII for even this reason alone.

I haven't yet seen enough of D800 output to be adamant of the conclusions, bu what I've seen indicates that a second hand D700 is liable to be a very attractive and possibly superior choice if low light and high ISO performance is your priority. And a D3s is head and shoulders better again.


A superb "must read" article. Complements JRista's excellent rply.
Noise, Dynamic Range and Bit Depth in Digital SLRs

Also refers to:

IRIS - free image processing software with astronomical photography bias - but useful for much else.

Free IMAGEJ image processing software from US NIH

\$\endgroup\$
3
  • \$\begingroup\$ Can you provide a link which shows 5D2 and D700 images of identical scenes taken in identical lighting conditions with equivalent lenses at high ISO settings and converted to the same image size? I find it hard to believe the difference is "extremely significant" \$\endgroup\$
    – Matt Grum
    Apr 24, 2012 at 12:22
  • \$\begingroup\$ @MattGrum - I'll try and find the images that convinced me that the D700 was my ultimatetarget (if we ignore the D3s). I have been waiting for a D700s or whatever so the D800 is a vast disappointment. Marvellous toy but not the next step towards "see in he dark" that I was hoping for. Sony will have 2 x FF out later this year and one should use the D800 sensor more or less so there may be some hope for the other. BUT Sony have a very poor record with high ISO noise compared to Nikon with the same sensor. My A700 was < D300 until Rev 4 software. \$\endgroup\$ Apr 24, 2012 at 13:00
  • 3
    \$\begingroup\$ Also there's a point that's frequently missed when this issue is discussed and that is that you can apply stronger noise reduction to high megapixel images without artifacts. This is because noise if much finer grained and falls in between the details instead of obscuring them. If straight downsampling averaging improves noise by a factor of 1.73, a sophisticated noise reduction scheme ought to be able to do a lot better. For a fixed amount of incoming light increasing megapixels gives you more information (about where the light falls) even if per pixel noise is higher. \$\endgroup\$
    – Matt Grum
    Apr 24, 2012 at 13:31
2
\$\begingroup\$

I used to think that MPs are overrated, until I did an experiment with oversampling. Inspired by the audio sampling rule of thumb to sample twice the frequency you need. 22K waves are sampled with 44k, but if you draw the figures you will only get the wave is it is in perfect phase. You can also risk sampling only zeroes. You need at least 4x oversampling to get the wave and its shape (it my be a sawtooth or a sine, you cant know with 2x the sample rate). Professional audio gear sample internally 192khz and then downsample to 48k or 44k.

I found that the same goes for photos - if you want to end up with a 1024x768 image, the best frequency you can achieve is where every second pixel is dark and the alternate every second pixels are bright (lets call it a texture). If you grab the image at 1024x768 you might miss the phase of that texture, or it may simply be blurred away due to the "true system resolution" being to low or bayer demosaicking will certainly screw it up. So you need to grab at least a 4096x3072 image without taking bayer demosiacking into account, so I would go for double that to account for the bayer, ie. 8192x6144.

The downsampling should be a better one than bilinear or bicubic to get the benefit. A sinc based filter is best, e.g. lanzcos.

1:1 vs oversampled then downsampled with lanczos:

Oversampling

\$\endgroup\$
2
  • \$\begingroup\$ Good points. Note that since a picture is a 2D image increase in MP is the square of the oversampling rate. So 2X oversampling is 4 times as many MP, and 4X oversampling is 16 times as many MP, 8X oversampling is 64 times as many MP. \$\endgroup\$ Jan 16, 2014 at 7:40
  • \$\begingroup\$ I know. note that I do not (unlike most ppl) count resolution in MPs. I work with cameras in many different aspect ratios (e.g. 1x12000, which is then a 0.012MP camera but it was most better resolution in one axis than a 4:3 36MP camera). you can see this in my resolution examples. \$\endgroup\$ Jan 16, 2014 at 11:06
1
\$\begingroup\$

I used to think the megapixel race was kind of silly, until I realized that high-end 36 MP cameras, make low-end (but perfectly usable) gear much more affordable. If someone needs to buy a camera that can product billboard size prints, great! Meanwhile, the rest of us get much take great pictures (for our modest needs) on our iPhones and prosumer Nikons.

\$\endgroup\$
2
  • \$\begingroup\$ The iPhone 4 and some of the recent Androids take amazing photos. I expect them to eat the p+s market completely in a few years. And I expect them to eat into the low end superzoom/dslr market. The good news is that Moore's law holds so our better APS-C DSLRs will continue to get better. \$\endgroup\$ Apr 26, 2012 at 22:01
  • \$\begingroup\$ Moore's law counts in optics too? I mean The "digital" part, where Moore's law might work, only starts inside the camera body. \$\endgroup\$ Mar 16, 2013 at 9:04
0
\$\begingroup\$

I'll give you a short and useful answer (I hope)

Lots of the answers given before me have great info so don't dismiss them

But to answer the question of: How often is 36MP going to be useful? Depends on your situation Amateur, who never prints and only displays digitally. Never.

Amateur who prints sometimes. Occasionally, if occasionally printing larger than A4

Pro, for various resaons. Quite often

For people who never print or don't go larger than poster size you won't ever see any usefullness in anything more than 10-12 and it does have downsides E.g. when shooting RAW (you do all shoot RAW don't you??)image sizes on the 21MP 5DmkII are around 24Mb, I've been told the image sizes on the D800 are around 30Mb Which can fill cards up very quickly So if you get a good 10-12 MP camera and don't print bigger than poster, you will get triple the number of images on a card and won't be able to tell the difference than if you had spent a huge amount more on the D800

I hope this helps

\$\endgroup\$
5
  • 3
    \$\begingroup\$ How about cropping? E.g. an amateur without pro-level ultra-telephoto lens. Wouldn't megapixels help? \$\endgroup\$
    – Imre
    Apr 25, 2012 at 21:13
  • 1
    \$\begingroup\$ I'm with @Imre here...more megapixels is of paramount importance when you can't afford the $10,000+ lenses necessary to get the kind of reach you require to capture the photographs you need. Cropping is the only alternative, and a camera like the D800 offers some astounding cropping capabilities. As for space...space is cheap. You can get 128Gb of CF space for a couple hundred bucks, which is less than 10% of the cost of the D800 itself. Relatively speaking, 30mp photos is a small price to pay for the IQ and cropping ability you get. \$\endgroup\$
    – jrista
    Apr 26, 2012 at 19:09
  • \$\begingroup\$ Robert Capa famously said "If your phots aren't good enough, you're not close enough" Cropping after the fact is never a replacement for learning to frame correctly in the first place. \$\endgroup\$
    – Richard
    May 23, 2012 at 15:05
  • \$\begingroup\$ Cropping after the fact is never a replacement for learning to frame correctly in the first place. Unless you are shooting wildlife it will be uncommon for you to need a lens longer than 200mm and there's dozens of lenses out there at that focal length or less for quite cheap. Working in this industry I've only ever used a lens longer than 200mm in two circumstances(for Formula 1, where for safety we couldn't go closer and wildlife) The most common are 50mm, 85mm and 100mm, so 24-70 and 70-200 will cover all \$\endgroup\$
    – Richard
    May 23, 2012 at 15:14
  • 1
    \$\begingroup\$ Yes, unless you're shooting wildlife - which is exactly what many amateurs like to do. \$\endgroup\$
    – Imre
    Apr 10, 2013 at 16:02
0
\$\begingroup\$

I just got my D800E which I moved to from a D200. I have measured 100 lppm with this thing using autofocus with a sigma 24 1.8 at f4. I haven't printed anything yet as I have had it only 2 days. I was able to excite moire with it shooting a test target, but it was visible only on the monitor, CaptureNX2 eliminated it with a low demosiac setting. I have a 55 micro nikkor which looks sharper, but it really can't be better than 100 because of the sensor. The big advantage is of course that 100 lppm is spread across a FF sensor and that is a lot of real image real estate. Finally I can shoot without having to compose to the frame so tightly. I can even shoot 645 or square - it will be a great freedom for my style where I like to frame for the subject. or at least that's what I am hoping

\$\endgroup\$
-3
\$\begingroup\$

No one has been taking gorgeous digital pictures for decades. At the turn of this century, many folks thought that film was far superior. These days, that argument has been settled.

It is not true that more pixels means sharper image, there are limits due to diffraction of the lens that provide a limit. Of course, if you use a bigger sensor, you can avoid that issue for practical sensor, which is why a lot of pros are now moving past 35mm (Full frame) and onto 6x4.5 images.

Often the megapixel count is just marketing fluff, to sucker in folks who don't know better. But sometimes more is better.

Its a more complex topic than your question's biases suggest.

\$\endgroup\$
1
  • \$\begingroup\$ What you're saying about diffraction is sort of true. Roger Cicala over at lensrentals.com have a nice blog post with numbers that show the (small) effect of diffraction. \$\endgroup\$ Mar 16, 2013 at 10:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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