Viva la Resolution!

How much memory to take on a trip is a decision that all digital photographers face at some point, pro and amateur alike. If you're a pro, the answer is to buy a bunch of memory and charge it to the client or write it off. If you're a traveler on a budget, the answer isn't that simple. You have to buy smart and make your choice based on quantity vs. quality. 

A digital newbie will say, "I love my new digital camera and the manual says that on this setting I can fit 22,000 pictures on one 4GB card!”  While that might be technically true for 640x480 VGA images, it's usually followed at a later date by, “This camera sucks! The pictures looked ok on the LCD, but they're are all fuzzy when I have them printed!”  

I’ll try to explain the how and why of quality vs. quantity and afterwards, give my suggestions on how to make an informed decision on what's best for you.

Quality and resolution:

 In a digital camera there are two major factors that affect quality. These are image size (resolution) and compression (quality). I will continue my preference for non-technical explanations and leave out the gory details about photosites, Bayer arrays and the worst of lossless vs. lossy compression. I will also leave RAW format out since most compact cameras don’t offer it and every manufacturer has their own. If you want more tech, Google those words and brace yourself!

Image Size (resolution):

Digital cameras capture light on a sensor made up of tiny light sensitive dots that roughly translate into pixels on the final image. Simply put, the resolution of a camera sensor is the number of  dots, or pixels, side to side multiplied by the number of pixels up and down. If that number equals a million, the sensor’s resolution is one megapixel (MP). Marketing departments have made the megapixel a household word and it is the measuring stick that we now use to measure the ability of a camera to capture detail. More dots = more resolution = more detail. With some exceptions, this holds true. (You may want to read my article on camera choices.)  

Let’s turn those dots, or pixels into something easy to relate to. Your computer monitor’s resolution is a good place to start, since you will probably use it the most to view your pictures. Monitor resolution is also measured in pixels. A 14” laptop screen typically has an 1368 pixels x 768 pixels (1.05 MP) resolution. A 20” LCD will usually be 1600 x 900 (1.44 MP). If you work with a 24” widescreen display, your images will be displayed at 1920 x 1200 (2.3 MP). The much-lauded Retina display on the iPad 4 is 2048 x 1536 (3.15 MP). If you are lucky enough to own one of the newer hi-resolution 27" or 30" graphics monitors, they display a whopping 2560 x 1600 image (4.1 MP). Whopping, by the way, for a monitor. For a camera, it's hard to find a new one under 10 MP anymore.  

Prints are another common destination for digital photos. Regardless of the mechanical resolution, most printers work with files that are sized to print at 240-300 DPI, or Dots Per Inch (a leftover term from offset printing…the use of computers makes PPI or Pixels Per Inch a more common reference nowadays.) At 300 PPI, the individual dots on the print are beyond the resolution of the unaided human eye, so prints made at this resolution appear as smooth, “photo quality” pictures. What this means is a 4” x 6” standard print needs to be 1,200 pixels high by 1800 pixels (2.16 MP) to appear clear and sharp like a print made from film. An 8” x 10” print needs to be 2,400 high by 3,000 wide (7.2 MP). Now we have some reference points and begin to see why 10MP or more can come in handy.

Quality

No, the quality setting doesn’t mean you can set your camera to only keep good pictures, though I’m sure somebody’s working on it! The “quality” of a digital image refers to how much the software in the camera compresses the image before it stores it. The image format used by virtually every digital camera today is JPEG (pronounced jay-peg - your picture files will have a .jpg extension). JPEG originated as an ISO (International Standards Organization) standard for the compression of digital images in 1990 and was written by the Joint Photographic Experts Group, hence the name.  The JPEG in use today is not quite the same, but the name stuck. I guess it was easy to pronounce.  

To understand image compression, I suppose we have to dip our toes into the Ocean of Tech and take a quick look at how digital images are stored. Each pixel is represented in an image file as a number that tells the computer what color it is and how bright it should be. In an un-compressed image like a bitmap (.bmp) or a TIFF (.tif), each pixel has a place and is reproduced whenever the file is opened. The code is written to the file as “pixel code” x “number of pixels”, so all 10 MP image files, for instance, will be exactly the same size. This will become relevant in a moment.  

In a simple, understandable world, bitmap file code would go like this:  
Row 1, Pixel 1 = 234; Row 1, Pixel 2 = 236; Row 1, Pixel 3 = 236; Row 1, Pixel 4 = 245;…
Row 2, Pixel 1 = 234; Row 2, Pixel 2 = 234; Row 2, Pixel 3 = 236; Row 2, Pixel 4 = 245;…
Etc., etc, etc.
 

In a JPEG, the encoding software saves space by combining similar pixels together and eliminating the need for coding each pixel separately. Like this:  
Row 1, Pixel 1 = 234; Row 1, Pixel 2-3 = 236; Row 1, Pixel 4 = 245;…
Row 2, Pixel 1-2 = 234; Row 2, Pixel 3 = 236; Row 2, Pixel 4 = 245;…
Etc. x 3…
 

As you can see, even in my tinker-toy representation of the code, the system saves space, a lot of it! Woo-hoo…something for nothing! Well, not quite. The example above only combines identical pixels (lossless). In the real world of compression, the software combines similar pixels to save even more space (lossy).

Sort of like this:  
Row 1, Pixel 1-3 = 235; Row 1, Pixel 4 = 245;…
Row 2, Pixel 1-3 = 235; Row 2, Pixel 4 = 245;…
Etc. x 3…
 

When the file is opened after compression, pixels 1-3 which were averaged to a middle value are all displayed as that average value, losing their original value. This sounds really bad using my little example, but in practice, an image file supports millions of color levels and the combination of dozens of levels is usually undetectable. The exception to this happens when you edit an original file and re-save it several times. Opening and closing to view a JPEG file does nothing, so don't worry about killing your pictures by looking at them. If, however, you edit and save the file, it will be recompressed and data will be lost each time. After several edits and saves, even a high quality JPEG can exhibit degradation. I always edit a copy even when using Photoshop which has an option for lossless compression.. Another seeming oddity of compression is that images with a lot of similar adjacent areas, such as a scene with a solid color background will compress more than an image with a wide variety of colors and details. That's why your frame counter may show 50 shots left and you get 35 -70 more. Your camera isn't broken, it is just averaging your file sizes and giving you a guide to go by.  

The reason you have different quality settings is so you can decide how much image quality you want to trade for file size reduction. The higher the quality setting, the more of the original image info is retained and the larger the file. As you reduce the quality, less and less of the original image info is retained and the smaller the file gets.

In this example I used a small section of a photo taken with a 6MP DSLR camera and saved as the highest quality JPEG. The cut-out was saved at highest, medium and lowest quality settings in Photoshop to illustrate the effect of compression. 

Highest:

Standard:

Lowest:


Hmmm… not much difference is there? Lets look really close, as if you wanted to make a 11 x 14 enlargement from a 5MP original:

Highest:

Standard:

Lowest:


At this point, the image degradation becomes noticeable. At the lowest setting, the colors are showing blotchiness, there's noticeable pixelization and diagonal lines are becoming jagged. That means that there is, after all, a moral to this story... 

The Scenario: 

  You have purchased a 10MP compact digital camera and are ready to travel, see the world, and bring it all back home to show friends and family. You want to make a DVD slide show of your best shots and some prints for the scrapbook. To do this you need to take enough memory, or “digital film” to get all those once-in-a-lifetime shots.  Your camera is a good one that has many menu options for setting resolution (image size) and quality (compression).  

The Question: “How much memory should I bring?”  

The Answer: “It depends…”  With all of the variables in image processors and manufacturer’s “standards”, there really aren’t any reliable constants you can go by. You really do have to learn a little about your camera before you can SWAG an answer.

Resolution settings:

 The largest file your camera will record; ALWAYS! If your camera is a 16MP compact, shoot at 16MP. If you think that a 10MP setting is good enough, then you should have bought a 10MP camera and saved some money. You can reduce the size of a 10 MP picture to 640 x 480 to e-mail it, but enlarging a 640 x 480 image to print a 5 x 7 will always look bad. (If you're thinking that the good guys on CSI can really get portrait quality images of a suspect from a convenience store security camera, then I have a whole list of things that I would like to sell you!) Making a small image bigger always loses quality. Always.

Quality settings:

As you can see in the examples above, the difference between the highest and standard quality is not that great. In almost all cases, a print made at the highest quality image setting in most cameras will be virtually indistinguishable from a print made from a standard quality image. Unless you are a working professional who doesn't use RAW and may need to make enlargements of any picture at any time, standard compression is very useable and will save a lot of space. That is not to say to never use the highest quality. Personally, I always use the highest setting. If you can afford the space, it is the best choice. Remember, you can always re-compress files to save space where the highest quality isn't needed, but it is a one-way-street! Once compressed, the image data that was discarded to achieve greater compression is gone forever!  

Best practices:

•  Set your camera to take the largest image available.

•  Use high or standard quality (but never lowest). 

•  Consult your manual to see how many pictures will fit on your chosen size of card or simply plug the card into the camera and see what the frame counter says. Better yet, crank off a bunch of pictures for a real world test. I say "a bunch" because you want a realistic average since file size can vary by as much as 50% due to compression. If you take 50 pictures of everything around the house, including scenes with a lot of solid colors and some with a lot of small details and different colors, you should get a good working average.

•  Buy enough memory to take 1½ to 2 times the number of photos you think you'll take. If you plan to take 50 pictures a day (keep in mind that that would be high by film standards and in general, low for digital) and you're going on a 7 day vacation, plan on 350 pictures x 2 = 700. Test for average file size with your camera on your chosen settings, do the math and buy accordingly.

•  When in doubt, buy too much. $20 for a fast 8GB card at home beats $39 for a cheap card on a cruise ship or at a resort...if it is available at all. Memory is as common as film now but the cost per GB can go up by a factor of four (or more) if you need more while traveling.

•  Don't forget that memory is reusable and, in the long run, cheap! So you spent $100 on memory this trip. Next time, you already have it and you can start averaging the cost downward, trip after trip.

•  Buy quality memory. Agonizing over spending the extra $20 on good memory after you just plunked down $500-$1000 on a camera is to me, the definition of false economy. Spending $10 with free shipping for a no-name 16GB card on eBay seems like a good deal  vs. $20-$25 for the same size SanDisk Extreme, Lexar Platinum or another major brand. Getting a "Card Not Useable" error on your LCD while on a $20,000 trip to Antarctica eats up those savings in a hurry. Keep in mind that eBay maintains a large number of guides on how to spot counterfeit memory card sellers since they make up the majority of eBay memory retailers.

•  Buy a protective case. Most memory cards are tiny and can get lost easily. A $10-$15 case is cheap insurance and can help keep track of card status. Full cards go face down in my case and empties, face up.

•  Here's the most important tip so far: You're on a trip with a camera...don't forget to HAVE FUN and TAKE PICTURES!

Dave