Making the most out of your megapixels

Perhaps you've been following my series of posts about the so-called megapixel myth. So far, I've confirmed the ugly truth that if you're in the market for a digital camera, then there's a good chance you'll never need more than 3 or 4 megapixels from your camera.
Written by David Berlind, Inactive

Perhaps you've been following my series of posts about the so-called megapixel myth. So far, I've confirmed the ugly truth that if you're in the market for a digital camera, then there's a good chance you'll never need more than 3 or 4 megapixels from your camera. Most people are better off saving their money (instead of opting for the most megapixels) or using it to get more bang for their buck in other parts of the camera. For example, in the amount of optical zoom. Or, an additional lens (if the camera supports lens interchangeability). Or maybe a decent tripod.

About the only people who need more than four megapixels are people who significantly crop and/or enlarge their photos. And no, an 8x10 isn't a significant enlargement of an original photo. It would be a significant enlargement if the original was cropped down to something that's 10 percent of the original photo's size and then what's left gets enlarged to 8x10. But most people don't do this with their pictures.

I do. In fact, in my attempt to get some great shots of some local bald eagles in flight (and perched), I'm learning about how to get the most out of the 6 megapixels that my Nikon D70 affords me. At the end of the day, given how I am cropping and enlarging, 6 megapixels may not be enough. But my hunch is that there are other things I can be doing to get more out of what I have.

Case in point? This past weekend's eagle watching. Here, in succession, are three different views of the same picture. Do the different images demonstrate the limitation of 6 megapixels or not?

The first is the 3008x2000 pixel original JPG-based photo that was created with my Nikon D70, resized into a 469x312 frame so it can fit in this blog. I do my photo resizing in Adobe's Photoshop CS. But, it doesn't matter what you use to resize your photo. If most of your pictures are going to end up online for sharing (either in a photosharing service like OPhoto, Webshots or Flicker, or in a blog), keep in mind that a lot of pixels get thrown out when you resize a 3008x2000 pixel image to get it down to 469x312. Here's that first version:


The above photo was taken with 750MM of optical zoom from about 400 yards away. The camera was mounted on a monopod and compared to last week where I was letting the camera automate everything (and got some blurry but decent results), this week, I took some matters into my own hands.

To minimize blur, I knew that I needed a faster shutter speed. After all, we're talking about a bird in flight whose movements (read: distance from the lens) were unpredictable while I was following it with the lens. I also wanted the most amount of light. So, I set the ISO (film speed) to 500 (it was at 200 last week). A film speed of 400 seemed low to me (a bad instinct) and 800 seemed high. The D70 offered 500 and 640 which is why I picked 500. Then, I put the camera in aperature priority mode to push the lens to its maximum aperature setting (the lowest number: f/6.3) to get the most light in. The resulting shutter speed was 1/2500 second. Looking at this number, I probably should have dropped to ISO 400. Not only might the resulting shutter speed have been fast enough (probably around 1/1250), pictures are less grainy as the film speed comes down and the extended shutter time would have let a bit more light in which might have revealed some more of the eagle's detail. But 200 was too slow. Also, instead of relying on auto white balance, I picked a pre-programmed setting for sunlight. Here's a cropped version of the above photo demonstrating the megapixels at work, to some extent.


If you look closely, you can see the eagle's wings are outlined by a brighter blue coloring. That's a result of Photoshop's unsharpen mask at work. It's working hard to pick out lines and sharpen them up a bit, but there simply isn't enough raw data (pixels) for Photoshop to work with. Why? Well, one critical mistake I made was I shot at 3008x2000 (the largest dimension the D70 is capable of) with camera's JPG quality setting on fine (vs. normal or basic). The D70's fine setting programs the camera to use the least amount of compression (1:4) when converting the image it captures into a JPG file. At the normal setting, JPG compression is 1:8 and at the basic setting, it's 1:16.

The net result is that even though 3008x2000 pixels adds up to approximately 6 million pixels, the resulting JPG file in no way contains the original information that was captured for each of those 6 million pixels. JPG compression was at work in the camera. And even though I picked fine, that's still not good enough if I wanted to make the most out of my 6 megapixels. For that, I should have been shooting in the camera's raw mode. Versus the 3000x2008/fine mode which generates files that are anywhere from 1 MB to 3MB in size (depending on level of detail), the D70's raw mode reliably generates files that are 5.5MB in size (the downside of which is they'll eat up memory a whole lot faster). So, forgetting how much less data a photo editor like Photoshop has to work with when lossy JPG compression is involved, here's what happens when you crop and enlarge and there simply aren't enough original pixels to work with:


Now the final question. Beyond shooting in the camera's raw mode, what else could I have done? Here are some thoughts:

  • Used a much better zoom lens. The Sigma that I use is a good lens. But part of what we're seeing here is the difference between a sub-$1000 lens and a $3,000 to $5,000 lens (what I couldn't afford). Not only are the optics going to be better, but my guess is that the difference between 150 feet and infinity on the focus ring isn't about a 1/8 inch (in other words, a mere tap of the ring can change the focus range by 100 feet, making granular focusing difficult).
  • I might also try the camera's built-in timer for the shutter release. It's bad enough that the camera is moving as I track an eagle in flight with the lens. But, on a monopod in particular (vs. a tripod), the odds that depression of the shutter release will result in unwanted movement of the camera are much greater than the chances of unwanted camera movement when the timer is used.
  • Get a digital SLR with more megapixels. All of my lenses are Nikon lenses (most of which were purchased for a film body). So, I'm pretty much married to Nikon. Nikon has many digital SLRs now. But the D80 is the next step up in terms of megapixels, offering 10.2 megapixels with image dimension of 3872 x 2592.  At around $1000 for the body, the D80 costs the same as the D70 did when I first bought the latter. The next step up is the $4700 12.4 megapixel D2Xs, a bit rich for my blood.

Have your own suggestions? I'd love to hear them. Finally, when I saw this through the lens at 400 yards, I thought I had another eagle in my sights (it's hard to see in the viewfinder, especially with my bad vision). But thanks to a few megapixels, it's clear that this was no eagle. It was a red-tailed hawk:

Editorial standards