Missing Pages: Megapixels
The Missing Pages column is a collection of all of the information that should have been included in your camera’s Owner’s Manual—but somehow got left out. This is a hybrid assortment of short articles that delivers the know-how you need to derive the maximum enjoyment—and creative expression—from your equipment.
It’s sort of a juiced-up User Guide for creative people who are not necessarily technical. Each part will teach you how to use one of the camera features or functions that you previously ignored or left set on Auto. And each will include a Creative Project so that you can try some scripted experimentation.
We will explain complex technical subject matter a way that everyone can understand. And if you happen to be a technical expert yourself, we’re including “Nerds Only” sidebars just for you. That way you can dig in deep—or just straddle the edges—of the technological stuff. It’s your decision.
Installment VII: Megapixels
Definition: Citing megapixels (MP) is a popular shorthand way to express the resolution of a digital camera in terms of how many millions of individual pixels (picture elements) are present in the image at time of capture.
Note the significance of the phrase “…at time of capture” in the definition above. Resolution can be occultly increased through interpolation, a process that analyzes actual data and makes intelligent guesses to increase the quantity of information. The number of physical pixels doesn’t change. The camera’s signal processing engine fills in the blanks based on complex mathematical algorithms.
After an image is captured you can change the number of pixels that it contains, sometimes without even realizing it. When we Resample an image we change its pixel dimensions. Upsampling involves adding pixels, while Downsampling requires discarding pixels. Both procedures degrade image quality, sometime significantly. People often do this indiscriminately when resizing images to print them. Below is an extreme example of an image that’s been degraded by Upsampling.
Resolution has a horizontal component and a vertical component. An image that is 1200 pixels tall by 1800 pixels wide contains 1200×1800 or 2,160,000 pixels. That’s about 2-megapixel. One megapixel is not actually one million, although it’s close. 1MP = 1024 × 1024 = 1,048,576, so we accept with nearly 5% inaccuracy because, well, it really doesn’t matter just as long as everyone counts them and labels them the same way.
Incidentally, an image that measures 1200×1800 is said to have an Aspect Ratio of 2:3, which is the same as a 4×6 inch print.
Many camera specifications show one number for Total Pixels and a lower number for Effective Pixels. The first, larger number indicates how many pixels can be counted on the entire surface area of the CCD or CMOS imager, including the edges that are not used. The second number is the only one that’s meaningful, as it indicates how many are actually recording data.
How many megapixels do you need?
How many megapixels do you really need? It all depends on what you want to do with the image. Unfortunately, digital cameras are lumped into categories based on the number of megapixels they can capture, and uninformed consumers make erroneous conclusions about image quality based on that single label. Having more pixels does not always mean better images. These images were shot with a humble 3-megapixel camera.
Having more megapixels can offer some distinct advantages, however. Higher resolution images produce larger prints without loss of quality. If you make 13×19 inch prints, for example, you need all the resolution you can get. The second advantage, one that’s all too often overlooked, is that you can crop smaller portions of a high-res image and still retain high quality.
On the other hand, if you are producing images for a website, or mostly print 8×10 or smaller, a nice 6-to-10 megapixel camera will serve you fine. Like many photographers, however, you may cling to the notion that someday you’ll want to turn your images into gigantic poster-sized prints. In that case, shoot large RAW images now and you won’t regret it later—that is, unless you run out of hard drive space. Read on.
High resolution comes at a high price. More megapixels mean larger files. Large files make for longer download times and can really fill up a memory card fast. Ditto your hard drive and backup externals. It’s not unmanageable, but you should be prepared. Think in terms of terabyte drives in your RAID configuration and you’ll be okay.
Do not confuse Resolution with Compression.
The Compression variable is often shown on the camera’s menu under the heading “Quality.” Camera makers essentially hide the squeeze factor by giving the various compression ratio levels clever aliases like Fine, Standard or Extra Fine. In truth, the compression ratio varies from as much as 20:1 to as little as 2:1 (plus, of course, uncompressed). It should be easy to understand that the more you squeeze the more you lose in terms of data. Some folks turn up the compression while keeping the resolution at the max setting to save file space. Not good. It’s far better to keep the compression as the lowest setting and reduce the number of megapixels.
More megapixels can in fact be a very bad thing if the additional pixels are crowded onto the same size CCD. In other words, some 6-megapixel image sensors can outperform 10-megapixel image sensors with the same physical dimensions because the 10MP sensor obviously has less space between the pixels. This space is called “pitch” and as it decreases the noise level increases. Increasing the number of pixels means increasing the pixel density, which in turn almost always means noisier images.
There are notable exceptions, and some camera makers have devised ingenious ways to diminish (if not entirely neutralize) noise. But there is a downside to aggressive noise reduction, too: less sharpness. Few DSLR cameras and darn few compacts achieve an acceptable compromise. That’s another reason why it’s important for the creative photographer to understand how the Resolution, Compression and Noise Reduction settings interrelate on their particular camera.
It’s very encouraging that the latest crop of popular compact cameras that are aimed at enthusiasts (Canon PowerShotS90, Panasonic Lumix DMC-LX3, Fujifilm F70-EXR for example) have put the emphasis on image quality and noise reduction instead of sheer marketable megapixel power. Camera manufacturers do listen to their customers who demand high quality images, not inflated numerical specifications.
Every digital camera I’ve ever seen allows the user to “dial down” the resolution. The Canon EOS 5D Mk II, for example, offers setting of 21.0, 11.1 and 5.2 megapixels when shooting JPEG and similar settings for RAW+JPEG. The Creative Assignment for this installment has three parts.
First, with your camera on a tripod, set the highest resolution and lowest compression ratio that your camera offers and shoot a colorful but stationary subject. Do this in bright light so that image noise does not become an issue. Repeat this process for every resolution setting your camera offers, but keep the compression ratio the same throughout. Since different cameras use different nomenclature for Resolution and Compression you’ll have to check the Owner’s Manual. If you’re shooting a Nikon D5000, for instance, you’ll set Image Size sequentially at Large (4288×2848), Medium (3216×2136) and Small (2144×1424) to get 12.2, 6.8 and 3.0 megapixel images.
Next, repeat the above, but set the compression ratio to the highest level (i.e., most compression). Again using the Nikon D5000 as an example, you have the option to set the Image Quality JPEG Fine for 1:4 compression, JPEG Normal for 1:8 and JPEG Basic for 1:16. For this exercise, use the 1:16 setting.
The third step is done with Photoshop (or Aperture, Lightroom or even the browser that came with your camera). Open two images that were shot at the same resolution setting but at the extremely different compression settings. Then, with both images open, zoom 100% (Actual Pixels) and critically examine the differences. If you fail to see any difference, make an appointment with a qualified ophthalmologist. At the very least you can expect to see the high resolution/low compression appear much better than the low resolution/high compression. How much better depends on your camera and, to some extent, the subject. A shaggy dog looks pretty good no matter what setting you use, but a person’s face—full of details and skin texture—will change dramatically based on how you have these two parameters set.
Bonus Assignment: borrow (or buy secondhand) a digital camera that has a VGA (640×480) resolution option. Many current compact cameras, including the Canon PowerShot SD1200 IS, offer VGA resolution as an option. Confirm that the camera is set at the low-res setting and carry it with you the next time you go out to photograph. You can use your “real” camera, too, but be sure to shoot every interesting subject at VGA. When you download and examine the images you will be amazed how good they can be—provided you don’t try to make them bigger than they should be.
For Nerds Only
The first breakthrough 5-megapixel camera, based on a Sony CCD, was the Minolta DiMAGE 7, shown at the PMA Show in February of 2001 and introduced to the media at bi-coastal press conferences on May 23, 2001. Precisely 90 days later, Sony introduced their own 5-megapixel camera, the Cyber-shot DSC-F707. Ironically, when Minolta abandoned the digital camera market after being acquired by Konica, it was Sony who purchased all of the Minolta digital camera patents, technology and intellectual property (and hired quite a few Minolta engineers, as well).
Got film? Dedicated film scanners in the 4000 dpi range can create 4000×6000 pixel digital images of extraordinarily high quality. If you’re one of us who began taking pictures long before the digital age, consider a film scanner to resuscitate your collection of negatives and slides. Don’t settle for a multi-function flatbed scanner that also does film part time. If you want the best quality, buy a film scanner that is optimized for scanning film.
Text ©2009 Jon Sienkiewicz.
Text ©2009 Jon Sienkiewicz.