What's A Megapixel And Why Should I Care?

The Pixels Determine Just What You Can--And Can’t Do With A Digicam

Reprinted from DIGITAL Photographer, a bimonthly print magazine.

by George Schaub

If you’re in the market for a digital camera, one of the most important specifications to consider is the pixel count the image sensor delivers. There are other indicators of image quality, such as the quality of the lens and the reliability of the auto-exposure system. But the number pixels goes a long way toward telling you what you can and cannot expect when you go to use the image. Now that the new breed of 3-plus megapixel (MP) cameras are just hitting the market, knowing about pixel count, resolution and image end use is more important than ever. As we’ll see, just because a digital camera has a larger pixel count doesn't mean it’s the best choice for the type of image use you have in mind. This knowledge can save you time, memory space, upload and download time and a lot of work later.

            If you look at the specs of some recent models, you’ll notice that they often have a choice of what’s called "Recording Modes." These modes will show you the highest possible image resolution a digicam can capture, as well as optional settings with lower resolutions. The Nikon Coolpix 950, for example, has a 1536x1024-, 1024x768- and a 640x480-mode option. The Agfa ePhoto 780 offers 1024x768 and 640x480 options. Just what do all these numbers mean, and how does the term megapixel get in the mix?  

This photo quality image was captured with a Nikon Coolpix 950 in uncompressed TIFF mode. It's file size is a whopping 10MB!

Defining Mega

            Let’s start by defining what megapixel means. As you probably know, the building block of the digital image is the pixel, short for picture element. This square, or rectangular box, captures and eventually transfers the image information within its space, and when combined with many, many other closely attached pixels, forms the total image. In film, by the way, an image is formed by millions and millions of grains that, when combined, yield an image. Both are illusions that work because of what’s known as Gestalt theory--our tendency to create unified form out of closely allied parts. The idea is the same, but digital builds with boxes while film builds with more randomly shaped grains. This difference gives film and digital their distinctive appearances.

            You can begin to see what a pixel looks like if you greatly enlarge a digital image, just as you can see grain if you put a piece of film under a microscope. Open an image on your screen and keep pressing the zoom key, and you can watch as the image pixellates into boxes. Generally, we want to avoid seeing pixels since they get in the way of the image illusion, but you should try zooming into them just once to see what’s behind the veil.

            An image sensor’s ability to deliver images for different end uses is determined by the number of pixels it contains. Rather than always work with the awkward pixel dimensions, such as 640x480, we can begin to differentiate sensors by using megapixel terms. We do that by multiplying the horizontal (640, or always the larger number) by the vertical pixels. So, for 640x480, we get a bit over 307,000 pixels, or 0.3MP The highest pixel count on the Nikon Coolpix 950 is 1600x1200 pixels, or 1,920,000 pixels, or 1.92MP.

            In Greek, the word "mega" means million, so we say that the Coolpix 950 is a megapixel camera. In short, the higher pixel count delivers more pixels, thus more picture elements and a greater ability to form a better illusion of a picture when a print is made. So, if you want to figure out just what your camera’s megapixel identity is, check the recording mode specs and multiply the vertical and horizontal pixels of the sensor.   

More pixels, better image-  the same grill at 1600X1200 is enlarged and cropped and still shows great detail	Jagged little stairs-  when enlarged, this 640X480 pixel image of a Buick grill begins to loose detail, showing "jaggies"	Enlarged even further individual pixels and stair stepping become evident.

After The Pixel Counting

            Now that we’ve figured out what megapixel means, let’s apply the pixel numbers to some real-world situations. But, before we do, we have one more hurdle to jump, and that’s image resolution. For that, we revert back to our pixel dimensions and introduce a new term-pixels per inch, known in shorthand as ppi. The ppi comes into play when you go to actually do something with your digital image. This may involve Web use or printing. Each application uses a pixel-per-inch equation. In short, ppi is always used in the context of the output device.

            Ppi and image-size relationships are easy enough to figure out. You simply divide the resolution of the output device (your monitor or printer) into the pixel dimensions. Most monitors can show a good-quality image at its resolution of 72 ppi. That’s what you want your Web and e-mail image resolution to be. It saves time and space, and frankly, you don’t need any more than that to show off your images.

            So, to figure out how big your image will be on a 72-ppi monitor, using a 640x480-pixel image, just divide 72 into those numbers. The answer is 8.9 inches by 6.7 inches. If your monitor resolution is 96 ppi (for PCs), the answer is 6.6 inches by 5 inches. If you want to make the image smaller, you can downsize with interpolation, or resampling, a method of tossing out some pixels and still maintaining a good image.

            Any image software worth its salt will make this easy. Just go into the image size Dialog Box and check Resampling, and type in the size you want. Don’t cut it down too much, because you may throw out too much information. About 25 percent is right. Do this on a copy (Save As) of your image so you still have the original information intact.

            Now, of course, you don’t want to send such an image over the Internet without performing some compression, and that’s where JPG compression and image file sizes come into play. A file size is calculated by first figuring out the pixel count and then multiplying that by three. The "three" is the RGB information (red, green and blue) contained in the image information after the camera has processed the original capture. So, for our 640x480 recording mode, we get a file size of 307K times three, or 921K. Send that to your friends and they’ll never chat with you again. A 50K to maximum 90K file is best to upload on the Internet.

            You can compress the file in either the camera, by choosing a compression ratio when you shoot, or in your image software later. In the camera, just choose a one-eighth (Normal) or one-sixteenth (Basic) compression mode. Using your software, choose JPG as the saving format, and use the Help and Dialog boxes to guide you on your way. Most software programs now have a "Web" option that makes all the calculations unnecessary. 

Playful photo quality - Working with high quality 1600X1200 pixel images permits infinite creative options without fear of quality loss, as with the image above.

Web Readying Your Image

            Now let’s make a Web-ready image from our megapixel camera using every pixel we can. Divide 1600x1200 by 72, and we get a 22-inch by 16.6-inch image! That’s some monitor. But it gets worse. Let’s figure out the file size of our color image by multiplying 1.92 by three. We get a whopping 5.76 megabytes (MB). Try sending that one over the lines, and you’ll get a good night’s sleep in between, and your buddy with the 28K modem will put out a notice to ban you from the Web.

            That’s why camera makers give you a choice of image recording modes and compression ratios, and why software programs allow you to compress the image even further. In short, if all you shoot for is the screen or the Web, you probably don’t need a major megapixel camera.

            But hold on. Just what are those 2MP and 3MP cameras good for then? We’ll see soon enough, when we go to make prints. 

Printing Tricks

            Let’s say you have a printer with a maximum output resolution of 300 dpi (that’s dots per inch). While experience has shown this writer that you don’t always have to print out at maximum printer resolution, let’s use the theoretically best resolution to make our images here. Once again, we get out our calculator to do the math.

            If we make a print from our 640x480-produced image, what size will we get? Divide those numbers by 300 and you get a whopping 2.13-inch by 1.6-inch print. If we drop down to 150 dpi, we double the size, but at 150 dpi we might not like the look of the sparsely dotted image. It’s what’s called the "devil-and-the-deep-blue-sea" syndrome. In practice, I’ve found that I can get away with using a good 300 dpi printer at about 225 dpi, so a 2.6-inch by 2.13-inch print is possible. Printer quality and the way ink or thermal color is laid down have a lot to do with quality here.

            But what happens if we bring in the big gun of 1600x1200 pixels? Then, we get a better-looking print at bigger sizes. Our theoretical limit is about 4 inches by 5 inches, but our 225 dpi yields a 5-inch by 7-inch photographic-quality print. When I say "photographic quality," I mean one that challenges images from 35mm film. In practice, I’ve made some very good 8-inch by 10-prints from the Nlikon Coolpix 950 and others of its class. How about those new big megapixel guns just coming on the market? One sensor we know about as of this writing has a 2048x1536 sensor. Let’s do the math at our theoretical and good quality limits. At 300 dpi we can get a straight 6.9-inch by 5.12-inch image. At  225 dpi, we get a 9.10-inch by 6.8-inch image. And those are photographic-quality prints. There’s a kicker to all this. Remember when I mentioned resampling to downsize an image for the Web? Well, it works both ways. You can also resample up, to a degree, and in effect add pixels to the image. The software does this by sampling adjacent pixels and filling the space with what it considers logical tones and colors. In this way you can get even greater print sizes than our calculations. Yes, you can do this with our little old 640x480-pixel image, but think of what this means for the big megapixel counts.

So, now when you choose a digital camera, or even a recording mode on the digicam you own, keep the end use of the image in mind. The pixels tell the tale and help determine just what you can--and can’t--do with that image. And, as we go through this stage of a mere 3MP into greater and greater megapixel counts, the sky is the limit for the kind of print quality we’ll get from our digital cameras. At this pace, we’ll probably see another leap in pixel count before long. My printer, and I, can’t wait. 

George Schaub is the author of The Digital Darkroom: Black-and-White Techniques Using Photoshop, $29.95, published by Silver Pixel Press, Rochester NY

______________________

DIGITAL Photographer is printed bimonthy. It is chock full of digital camera reviews and articles to help you get the most from your digital camera. Right now, Apogee Photo readers can get two years for the price of one by just clicking here and filling out the information