On Why Bigger Isn’t Necessarily Better

southern Iceland in a Super Jeep Weekend Outtakes

When people talk to me about buying a digital camera, there’s one point I stress more than anything else:

Do not let the number of MPs (megapixels) convince you that the camera is better.

Sales people will try and impress you with numbers because most people don’t know what those numbers mean. I’m convinced that many sales people don’t know what those numbers mean, either, but since the MPs are stamped right there on the camera, they will try and tell you more is better in the hopes that you’ll believe them. Don’t believe them!

Here’s a pretty good article that showed up in my Twitter stream that explains why the megapixel count is more than a marketing ploy, it is a scam:

The Great Megapixel Swindle: An Example:
http://petavoxel.wordpress.com/2010/01/19/mp-swindle-example/

Let me give you my own example, comparing two cameras, one with a low megapixel count and one with a high megapixel count. (As you can see by the side-by-sides above, the cameras have slightly different aspect ratios, but it won’t affect the argument, which is megapixels overall.)

southern Iceland in a Super Jeep

This was shot in Iceland with the Pentax K100D, my first DSLR, which I sold last spring (reluctantly, because I loved that camera — the straight-out-of-the-camera JPEGs were great). It is an entry-level 6.1MP DSLR, with a maximum resolution size of 3008×2000 pixels. It was a sunny day, the ISO was relatively low (200), and the focus was for the entire landscape (infinity) so everything should be in focus. Click on the picture to view it at full resolution in a new window.

Exposure: 0.001 sec (1/1000)
Aperture: f/5.6
Focal Length: 33 mm
ISO Speed: 200
Image Width: 3000
Image Height: 2000

Weekend Outtakes

Now, here’s a photo taken by my brother, Allan, in Vancouver with his Canon G10, which is a point-and-shoot with a whopping 14.7 MPs. It wasn’t shot at max resolution (4416×3312 pixels) — which is another beef about marketing, and I’ll get to that in a bit — but the important point is that it was shot at a resolution GREATER than the 1st photo, specifically 3753×2814, or 17.55% bigger. Ignore the colours and contrast, just look at the sharpness and detail. Click on the pic to view it large in a new window.

Note that the ISO (or film speed) is 80, which means the sensitivity is lower and therefore the noise level should be lower than the Iceland shot, for which an ISO of 200 was used. In basic terms, the Iceland shot should theoretically be grainier, or “noisier”, than the Vancouver shot. The Vancouver photo was taken at half the shutter speed of the Iceland photo, but 1/500 is still fast and there should be no shake.

Exposure: 0.002 sec (1/500)
Aperture: f/4.5
Focal Length: 30.5 mm
ISO Speed: 80
Image Width: 3753
Image Height: 2814

I’ve sliced away some parts of both photos to do another side-by-side at 100% (Iceland left, Vancouver right):

See the difference? Which do you find grainier? The Vancouver photo on the right side was shot by the 14.7MP Canon G10, which is more than double the MPs of the Pentax K100D at 6.1MPs on the left side. Now, part of the difference in quality is the lens glass. I am fairly certain that I took the Iceland photo with my Pentax kit lens, the 18-55mm, and a kit lens is the lowliest of all lenses in a brand line. Kit lenses aren’t always terrible, but it’s comparable in quality to the stock stereo you get with your car — for the average person it’s bearable, but people who enjoy music are going to replace it, anyway. On a camera, a kit lens is soon to be replaced, too.

The major factor in the difference is the sensor. If you were to go and read the sensor specs for both cameras, this is what you’ll find:

Canon G10: 1/1.7-inch type Charge Coupled Device (CCD), which is 7.17 mm x 5.31 mm
Pentax K100D: 23.5mm x 15.7mm CCD sensor

Now here is where size does matter! As you can see, the Pentax K100D has a much bigger sensor than the Canon G10 — around 3x bigger. DSLR lenses are bigger because their chip sensors are bigger. It is the one of the reasons why point-and-shoot cameras are more compact, even with a zoom, because it requires less glass to bring light to that little sensor. When companies try and cram more pixels into a small sensor, it actually reduces the quality of the photo.

This article by digitalcamerainfo.com puts it this way:

Fitting more megapixels in the same small amount of physical space means that all the receptor sites on the sensor must be smaller, which means that each site has less light hitting it. Less light per pixel means images that are less clear and sharp.

However, between two different manufacturers making CCDs that are the same size, there’s no way to know which produces better-quality photos using spec numbers because they use different computations when capturing and processing pixels. At that point it is a qualitative difference rather than quantitative. The only way to compare is to compare their photos, and that’s where personal preferences come in. But if you compare any DSLR to any point-and-shoot, the DSLR will always come out on top when the photos are displayed larger — because of the sensor, and because of the glass.

What’s that beef you were talking about?

I mentioned that Allan didn’t take that Vancouver shot at full resolution, which is what I was hoping for in my example. Actually, I couldn’t find any Canon G10 photos in his Flickr collection taken at full resolution. This is actually very common, and most people don’t think about it when they buy a point-and-shoot: if megapixels are such a good thing, why don’t people take advantage of them?

For one thing, the file sizes are way too large. My Nikon D300 is a 12.3MP camera, and the largest photo I can make is 4288×2848 pixels with a JPEG-format file size of around 4-8MBs (depending on what I’m shooting) and that’s plenty. (My RAW files are much bigger, at 9-12MBs.) Why would a point-and-shoot, with a sensor roughly a third of the size, need more pixels than a DSLR if it can only capture a third of the detail? Because it’s a gimmick! You’d never email a full-size snapshot around, and it’s far too big for a web page, Facebook, or any social media site. That size is really useless unless you’re considering enlarging it to hang over your fireplace.

Which brings me to my next point: you’d never want to enlarge a point-and-shoot snapshot, anyway, because even at the size I showed you, which is probably at least 75% of its resolution capacity, it has purple fringing and artifacts. You can barely make out the trees on the mountains. The level of detail is less than ideal; the cost of enlargement would not be worthwhile. The shot itself is good, the colours are set to vivid (that can be toned down in Photoshop), but the detail can’t be increased because Canon’s sensor is too small. It’s about pixel quality, not quantity.

But what about the price?

When you take away the bulk of the DSLR, there’s usually a gripe about the price versus a compact camera. It’s true, they can be pricey, but if you’re an amateur or a beginner, the prices of entry-level DSLRs are actually lower than the higher-end compact cameras. Using the same cameras in my example, I bought my Pentax K100D in May 2007 for about US$500, including the Pentax SMC-DA 18-55mm f/3.5-5.6 kit lens. The Canon G10’s suggested retail price today is US$500. Which means that if you scout for a used Pentax K100D on Craigslist right now, you could probably get one with the kit lens for about half of what the Canon G10 costs currently, and make better-quality photos for much cheaper.

I try and tell people that for the cost of the new compact cameras, they could invest in an entry-level DSLR and even get a lens, but I think people are reluctant to learn how to use a DSLR. I would be willing to teach ANYBODY how to use their DSLR, if that would convince them to make the move.

It’s not that I dislike compact cameras or even Canon, not at all. In fact I owned Canon compact digitals for five years (2002-2007) before I bought my first DSLR, which was a Pentax. I still recommend Pentax for its value, especially for entry-level users. My current DSLR is a Nikon. (As you can see, I’m not particularly brand loyal. There is no need to be except that switching an entire system is expensive due to the lenses and bodies being incompatible between brands; each brand has strengths and weakness across all their lines.)

I learned composition on compacts, which is an important skill regardless of equipment. I still shoot with compacts on occasion, for video and some snapshots. I think it’s very possible to take decent photos with point-and-shoots if the settings are used properly (and displayed small so their flaws don’t show), and I have actually sold photos taken by compacts. But the newer point-and-shoot models mystify most people because they have a lot of bells and whistles. Nobody likes to read the manual (I generally don’t, either, I only use them as reference tools but usually I go online), but the manuals are written by technical writers and they don’t tell you how to take a good photo, they only tell you about settings and specs.

I think this is the longest post I’ve ever written on the subject of photography, but I wanted to take the time to write it out because photographic equipment at the average consumer level is rife with confusion and misleading information. People are bombarded with features that they don’t understand. At a professional level, sales people have less sway with buyers because pros have the technical knowledge and a more watchful eye on the industry. Professionals are pickier about their equipment and view it as a business investment. However, the consumer market is big money and companies like Nikon, Canon, Sony, Olympus, Panasonic, Samsung, Pentax, Casio, Kodak, Konica, and all the rest sell point-and-shoots every day to people who don’t know what to look for in a camera. If you want to compare them, I suggest the side-by-side comparison table feature over at DPReview.com’s buying guide section. There are also comprehensive reviews if you want to go more in-depth, and summary-length versions of the reviews if you just want to get their rating.

My recommendation is to do all your research before you walk into the camera store, then try out a few cameras that are your top picks to see how they feel in your hands, if the controls are in logical places, and if the way you would use it requires digging around in menus versus the settings at your fingertips. Better yet: borrow them from friends or let your friends show you their cameras. It’s also a good idea to read some reviews and ratings, but ultimately, you should look at the photos you already take — portraits? low light? action? landscapes? — and base your search criteria on the style of your shooting rather than gimmicky things like megapixels and dozens of “modes” (eg. snow mode, beach mode, museum mode). Hmmm… I’m getting into another topic (buying), so I’d better stop here.

And if I’ve confused you even more, I take full responsibility and welcome any questions to clear it up!