Plugged In: More megapixels not always worth more money

By Joe Kashi, for the Redoubt Reporter

The megapixel wars have resumed, with Canon leading the attack, followed by Sony, Nikon and Pentax. Does any of this make the slightest difference to most users? As a practical matter, no. Here’s why:

Nikon and Sony first “broke” the megapixel truce, shipping Nikon’s large D800/D810 digital SLR series and Sony’s compact A7R mirrorless cameras, both built around Sony’s excellent, 36-megapixel sensor. Canon responded with its 50-megapixel 5DSr model, basically a standard 5D Mark III body with a higher-resolution sensor.

Canon further upped the ante by announcing ongoing development of a 120-megapixel dSLR for normal use, and a 250-megapixel model for surveillance and scientific use.

That’s a quarter-billion megapixels. Only a few months ago, Canon’s top-end professional cameras used sensors containing about 22 megapixels. Canon insisted, truthfully, that this was more than adequate for nearly all professional and amateur needs.

More recently, Sony’s newest A7R Mark II mirrorless camera includes what Sony describes as an optimally sized, 42-megapixel sensor. Pentax’s recently announced full-frame model will likely also use this new Sony 42-megapixel sensor, recently described by DXO as having the best all-around characteristics of any digital imaging device. I expect that Nikon’s next-generation D8XX cameras will likely use this sensor, or a variant of it.

A very high-resolution sensor has one potential advantage, assuming nearly perfect lenses and favorable situations. While one can reduce, when desired, the level of detail and sharpness of a very high-resolution image through later post-processing, the opposite is not true.

You can’t enhance detail that’s not captured in the first instance by a lower-resolution image. Theoretically, the higher-resolution sensor may provide greater flexibility for some circumstances where hypersharpness and crisply rendered detail may be needed for practical reasons or might be artistically appropriate.

That said, anything beyond 50 or 60 megapixels seems wasted for most nonsurveillance or nonscientific uses. The basic resolution of a 50- to 60-megapixel sensor already allows you to make 30-inch-wide images at 300 dots-per-inch, noninterpolated resolutions. With that much resolution, assuming excellent lenses, you can make fully sharp images without any interpolation software filling in the blank spots and smudging fine detail.

Very few people have either the ability to print images larger than what’s directly available with 50 megapixels or the space to display them. Very few people have, or can afford, the superquality lenses needed to take full advantage of such high-resolution sensors. Without lenses capable of such high resolutions, the sensor’s maximum theoretical resolution is meaningless.

Perhaps more fundamentally, our eyes are not really capable of discerning significantly higher dots-per-inch resolutions. One well-publicized test made several years ago by The New York Times found that people could not discern whether very large prints were made with 5-, 10- or 16-megapixel cameras.

Similarly, greater than 16-megapixel resolution is wasted if you’re primarily making images for web and social media posting, or making final prints not enlarged beyond the 13-by-19-inch maximum size typical of better, consumer-grade photo printers.

Even if you are making very large prints, 30-inches-wide or larger, you’ll only notice a practical difference when you’re using exceptionally sharp lenses and careful technique to make images of static subjects near optical infinity.

I found this to be true in my own experiments just a few weeks ago, comparing normal, 16-mexapixel images made with an Olympus E-M5 II with the same images taken immediately afterward using that camera’s 64-megapixel, high-resolution RAW mode.

I tried the super-resolution mode because I was trying to accurately capture the subtle yet intense colors of autumn fireweed foliage, something that tends to falter with most digital cameras. In addition to 64-megapixel resolution RAW files, the new Olympus’ high-resolution sensor-shift mode captures the primary Red Green Blue color components much more truly than the traditional Bayer color filter arrays used by nearly all other digital cameras.

Bayer arrays use software to make a sophisticated guess at each pixel’s true color, rather than separately and precisely measuring every pure color component. As a result, areas of intense color or unusual light tend to be rendered somewhat inaccurately by digital cameras relying solely upon Bayer arrays and software interpretation of color and fine detail.

  • Hint: To fully realize the color capabilities of any digital camera, you’ll need to hardware-calibrate your monitor and printer to each other, and also use X-Rite’s color-checker Passport color reference and software combination to directly calibrate your camera and post-processing program to each other.

While the 16-megapixel images appeared nicely sharp at 100 percent magnification, basically equivalent to a very sharp, 12-by-15-inch print at 300 dpi resolution, the 64-megapixel shots frequently had noticeably limited sharpness at the 100 percent magnification needed for making larger 30-by-24, exhibition-grade prints at the same 300 dpi print resolution.

The gear was not the problem. I had previously tested it indoors with flat test charts and knew that it could produce super-resolution images under optimum circumstances. No, my problem arose from using super-resolution camera gear under real-world conditions.

At the relatively short, 5- to 6-foot distances between my camera and the crimson fireweed, there simply wasn’t enough depth of field for front-to-back, super-resolution sharpness of important subject areas. Not even at f/7.1. A slight breeze caused subject motion that wasn’t visible to me but was visible at 100 percent magnification of 64-megapixel images. The heavy tripod and remote release I used could not counteract that seemingly minimal subject motion degrading sharpness.

I used very small, precisely set autofocus points, but the autofocus often focused on some sharp line in the background, despite precise adjustment. That would not be noticeable with the 16- to 24-megapixel sensors, but is definitely noticeable at higher resolutions.

It’s worth noting that contrast autofocus used by compact-system cameras and dSLR cameras in live-view mode is inherently more accurate than the phase-shift autofocus used by dSLR cameras and their optical viewfinders, suggesting that dSLR cameras would likely have even greater autofocus difficulties in such situations.

  • Hint: There’s a technical solution to depth of field limitations on the horizon. Firmware releases for some prograde cameras now feature a focus-stacking mode in which the camera automatically takes eight or more shots at slightly different focus points, then combines them into a single sharp image file with greater depth of field. This can already be done with post-processing software if your camera has a “focus-bracketing” feature, but that requires some post-capture work. Both approaches require a static subject and tripod-mounted camera.

If you do use a super-resolution digital camera, then for best results you’ll need to buy some superlatively sharp lenses (a real expense), use your camera on a heavy tripod with a remote release, avoid subjects that may show subtle motion or require great depth of field, and use your lenses stopped down to their optimum apertures.

Good technique and personal discipline are obviously necessary to gain any consistent benefits. It reminds me of the slow, methodical techniques required for tripod-mounted, large-format film cameras many years ago.

If you’re able and willing to work in a slow, disciplined manner and must make 30-inch or larger, exhibition-quality prints, then super-resolution cameras may work well for you. Otherwise, I doubt you would find a leap beyond 40 or 50 megapixels worth the hassle and cost.

Local attorney Joe Kashi received degrees from MIT and his law degree from Georgetown University. He has published articles about computer technology, law practice and digital photography in national media since 1990. Many of his articles can be accessed through his website, http://www.kashilaw.com.

1 Comment

Filed under photography, Plugged in

One response to “Plugged In: More megapixels not always worth more money

  1. jack

    thank you for the information. nicely written.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s