Concerning Megapixels

(This is another in a series of articles based on posts I shared elsewhere. This one is based on a reply to a post concerning how important it is to move to a newer, improved sensor with higher photo site density. The immediate question had to do with how often the improvements would be significant enough to be seen, and the writer had correctly pointed out that there can be advantages to higher “MP count” when making very large, high quality prints.)

It is useful to try for a realistic understanding of how and when a higher MP sensor may show its advantages. This post tries to not take a position on brands and models, but rather to lay out a comparison of some relevant technical stuff — from which we can all draw our own individual conclusions.

There is a point below which you would be hard pressed to tell the difference between prints made from 22MP and 36MP cameras and above which you might be able to.(1) For example, virtually everyone would agree that the difference is typically completely invisible in small web images, and virtually everyone would agree that it could be visible if you closely inspect a print that is six feet wide. Since we could debate just where the boundary is — and, frankly, it is somewhat subjective — you could pick any point on the print size scale that you want and the principles will be the same.

Some Comparisons

Just for fun, let me use completely arbitrarily use two print sizes and base the comparisons on the 22MP Canon 5D Mark III and the Sony/Canon 36MP sensor cameras.

First, using native 300 ppi resolution from the original file you can produce the following print sizes without interpolation:

Sony/Nikon 36MP sensor – 16.4″ x 24.5″
Canon 5D III 22MP sensor – 12.8″ x 19.2″

In the real world, we regularly print much larger and very beautiful prints from such files. One common “rule of thumb” suggests that you might want to be cautious about upsizing to the point that your original ppi resolution falls below about 180. (Many feel that you can uprez to larger sizes and interpolate to a minimum of about 180.) At that resolution, you could create the following print sizes from an uninterpolated original file at 180 ppi:

Sony/Nikon 36MP sensor – 27.3″ x 40.9″
Canon 5D III 22MP sensor – 21.3″ x 32″

And, since none of these are absolute limits, it is important to point out that people have been making beautiful larger prints from 21MP and 22MP sensors… so you could clearly go a bit larger on either sensor than what is shown here, with careful uprezzing, depending upon subject, depending upon the print media (e.g. – canvas versus traditional papers), and so forth.

What Does It Mean?

In all cases, the maximum size isn’t really the issue as much as the relative difference between them. With that in mind, how much larger does the near-doubling of MP from 22 to 36 get you in a print that based on the same ppi “resolution” in your post-processing software?

If all of these cases, to the extent that sensor resolution is the controlling factor(2), you can make a print about 20% wider/taller with a 36MP sensor as compared to what you can make with a 22MP sensor at the same resolution.

22MP and 36MP sizes
22MP and 36MP Sizes

This is potentially significant, in and of itself, to a small subset of DSLR photographers who regularly print in the 24″ x 36″ and larger range and who notice problems in their prints related to sensor resolution that either limit the usage of their prints or prevent them from producing larger prints. To the extent that other factors that also limit print magnification are well controlled**, those who feel that they cannot print beyond about 24″ x 36″ with a 22MP sensor image might now feel confident printing at 30″ x 40″ with roughly the same sensor resolution coming from the 36MP sensor.

These are facts, but the interpretation of facts and their significance is always a subjective matter. From my perspective, it is clear that the newer higher MP sensors coming from Sony (and included in cameras from several manufacturers) are advancing the potential quality of image capture from various systems and formats that use them, and among those advances are the potential for larger print sizes.

Subjective Conclusions

Photographers differ about what this means to us individually. For example, the hypothetical photographer I mentioned above — who currently makes a lot of very big prints and is objectively limited by sensor resolution in his/her ability to produce large enough prints — might find that the 20% increase in linear print dimensions from the 36MP sensor is compelling enough to invest in another camera or even switch to a different brand, since it can allow somewhat larger print size. Another photographer who understands the ways in which the newer sensors improve on the older sensors but never prints at those very large sizes, may realize that more MP would not make any significant difference in his/her photography. A third photographer (I’m in this category) understands why improvements are good, believes that the manufacturer of the gear he/she currently uses will “get there” before long, and thus chooses to continue to use the current equipment while waiting.

In the end, while the technical theory is interesting and useful, what matters is the photographic results, and at these sizes those results pretty much necessarily will be evaluated in the form of prints, making it imperative that look beyond the math and the specifications and ground our evaluation in comparisons of real world photographs… which, unfortunately, we cannot really directly do using the medium of the web.

Footnotes:

(1) I’m part of a print review group that meets regularly to share and critique member’s photographs. Members use a wide range of gear including Canon and Nikon DSLRs (including the D800e and D810), Fujifulm and other mirrorless cameras, MF and LF film, medium format 80MP back systems… and one member has recently been using the monster Seitz scanning back camera! The member who shoots the 80MP back digital MF system (who “downsized” to this from LF a few years back) brought some carefully made test images comparing sections of 30″ x 40″ prints from the 80MP back and his new D810 to our most recent meeting and challenged us to tell the difference. There were differences, but they were so small as to be essentially completely insignificant.

(2)  Some photographers overlook the fact that sensor resolution itself isn’t the whole story when it comes to making very large prints from very small originals. With greater magnifications — far beyond what we attempted with film enlargement — a range of factors that affect image resolution become more and more critical. These include camera stability — going beyond just using a solid tripod to making sure to use electronic shutter initialization and allowing the camera/lens/tripod system “settle” before exposure— and much more critical attention to optical issues including the quality of the lens, the accuracy of focus, and choice of aperture.

It is also true the print resolution is not the only potential advantage from higher MP sensors. For example, to the extent that noise is an issue in your prints, the size of the noise “grain” will be smaller with the higher MP sensor. There is also the potential to improve the quality/smoothness of tonal gradients with higher MP counts. And as newer sensors continue to improve noise handling, higher ISO performance, and dynamic range, the potential quality of image capture improves in other ways than simply resolution.

G Dan Mitchell is a California photographer and visual opportunist whose subjects include the Pacific coast, redwood forests, central California oak/grasslands, the Sierra Nevada, California deserts, urban landscapes, night photography, and more.
Blog | About | Flickr | Twitter | FacebookGoogle+ | 500px.com | LinkedIn | Email

Text, photographs, and other media are © Copyright G Dan Mitchell (or others when indicated) and are not in the public domain and may not be used on websites, blogs, or in other media without advance permission from G Dan Mitchell.

Join the discussion — leave a comment or question. (Comments are moderated and may not appear immediately.)

This site uses Akismet to reduce spam. Learn how your comment data is processed.