Digital Faqs - September 2001
By Michael J. McNamara
September 2001

Frequently Asked Questions

 

How fast is Digital Film? Who has the fastest memory cards out there? And will faster memory cards make a difference with your camera?

Since digital cameras burst onto the camera scene over 10 years ago, the amount of misinformation and confusion surrounding them has been unbelievable. We've done our best to elucidate the facts and explain many of the terms and technologies used in digital photography, but there are still quite a few readers' questions that pop up on a regular basis. Some are too complex to answer in our "Contact Sheet" section, so we've created this new Digital FAQs column to address them in greater detail. If you have a burning digital photography question, don't be afraid to ask us for the answer!

Q: MEMORY MADNESS
I don't understand why the new 64 megabyte "Digital Film" card I just bought doesn't work as advertised. It claims to be a faster 12X speed CompactFlash card, so shouldn't I notice some improvement in low light—like less camera blur and greater range to my flash—since I'm using faster digital film?

A: Believe it or not, this is a popular question at the camera sales counter and in chat rooms. And at first, it sounds like a logical question. After all, if a faster film (such as an ISO 800 print film) improves performance in low light, shouldn't faster "digital film" do the same? Nope!

While the term "Digital Film" has been used to market certain brands of memory cards and is thrown about wildly in digital camera circles, "digital film" is really a misnomer. But in its defense, the term sounds a lot better than "removable storage card", and in some respects helps ease the transition to digital photography by providing a familiar reference. In truth, removable storage cards are not photosensitive, and aren't even a part of the imaging system in your digital camera. They have no effect on the relative light-sensitivity of your camera. They are merely solid-state data storage devices.

Here's how it works: When light hits the photosensitive pixels of a CCD or CMOS sensor, an analog voltage that varies with the amount of light landing on the pixel is generated. This voltage is converted to digital data by an A/D converter, processed into color and image information, and then sent to the memory card as digital data. Only the CCD is exposed to light, and unlike a film-based camera where you change film speeds to suit the lighting conditions, you can't adjust the light sensitivity of a CCD sensor without boosting background noise. Unfortunately, increasing sensitivity degrades image quality much more noticeably than switching to a higher ISO film. It's in the process of storing the image data on the card, or reading it back for viewing on the LCD monitor, that "digital film" speed is used as a benchmark to indicate the potential write speeds of a storage card. The same benchmark is used to rate the potential write speed of a CD-RW drive.

A 1X-speed storage device such as a memory card or CD-RW drive can record 150 kilobytes of data per second to its memory (regardless of the memory media used), while a 12X device can record up to 1.8 megabytes per second to its memory. That means that under ideal circumstances, a 12X CF card should be able to record several images per second from a typical 2.1MP or 3.1MP camera until the storage capacity of the card is reached. For example, let's say you loaded a 12X-speed 64MB CF card into a 3.1MP camera that stores an image in Fine JPEG compression at 800 KB (.8 MB) per image. Theoretically, you should be able to capture 80 images at more than 2 frames per second (filling the 64 MB card in 35 ½ seconds). On the other hand, a 4X-speed memory card would only be able to capture a frame every 1 1/3 seconds, taking 106 seconds to shoot the same number of images. In short, using a faster digital film is the way to improve camera performance and eliminate that frustrating lag time between shots.

  When 12X isn't 12X: If storage card speed was the only factor in the equation, faster digital film would make for faster shooting. But it doesn't, just as using a 16X CD-RW drive with the wrong type of CD-R disk or on a really slow computer slows the recording speed. In actuality, a faster speed memory card might not affect the recording speed of your camera at all. Or it may increase it slightly, or even significantly. It really depends on the age of the camera, its processing hardware, and whether or not it has an internal memory buffer. In single-shot mode, most digital cameras begin to write image data to memory as soon as the analog information from the CCD is converted to digital data, processed into color and pixel information, and compressed to shrink the data. This takes up to 5 or six seconds on some lower-priced or older cameras, so using a faster memory card may have little effect on speeding up the lag time. Even on newer cameras with faster processing circuitry, the camera drivers may not be up-to-date or able to utilize the faster speed of the card. In this case, you might notice a small increase in speed—if you're lucky. Hopefully, now that faster cards are available, camera manufacturers will provide compatible drivers that take advantage of the improved card speed.

In the meantime, many new cameras use much faster processing hardware and include a RAM memory buffer that can be filled up rapidly with image data. Since data is first stored in the RAM buffer, lag time between still shots decreases significantly. Most cameras with a RAM memory buffer also let you shoot several images in a row in burst mode. For example, the Olympus E-10 can capture 4 frames in about 1.3 seconds, while the Canon G1 captures 8 frames in 8 seconds. As soon as RAM buffer on these cameras is full, the camera stops firing and you have to wait a few seconds until buffer memory is transferred to the memory card. On the E-10, it took 27 seconds to capture and write all the data from four shots (taken at 1.3MP res in HQ setting) to a 340MB, 12X Lexar Media CF card; 29 seconds using a 512MB Simple Technology Compact Flash (CF) card (speed unknown), and 29 seconds using an Olympus brand 32MB SmartMedia card. Even a three-year-old 4MB CF card turned in a 28-sec time. On the Canon G1, the time it took to capture and store an 8-frame burst varied from 23-29 seconds.

Should you buy a memory card just because it claims to be faster? Only if you want to ensure that the memory card doesn't add to the lag time on your particular model, or if you're able to check out a sample card to see if it helps before you buy. Most new storage cards are relatively fast. Other features that factor into buying a particular brand, include price and USB compatibility. As cameras grow in resolution and file size per image, and more are 12X (or higher) compatible, the card speed might become more of a factor. But for now, most lag times can be primarily attributed to camera design and processing hardware. Unfortunately, you can't change those without buying another digital camera.

Q: RESOLUTION AND PIXELS
How do you relate the number of pixels captured by a digital camera to the size of the print you can make from that camera? I'm interested in buying a $300 point-and-shoot digital camera with 1.3MP resolution in order to make 8x10-inch photoquality prints. Surely, a camera that costs that much should be able to deliver better quality than a single-use or low-cost film camera!

A: Having tested hundreds of digital cameras over the years, as well as nearly every major film-based camera and new film, we can confidently state that a 1.3MP camera doesn't come close to matching the image quality of a standard single-use camera, let alone a great one-time use or decent P/S. Sure, you can get a 3x5-inch print that looks pretty good from several 1.3MP cameras, and that may satisfy many amateur photographers. And a 1.3MP image looks pretty good on a monitor and may also be acceptable as an 8x10-inch print when framed and viewed from a distance (not at arm's length). But when you enlarge the detail in these images even slightly past a 4x6-inch print size or compare the detail in side-by-side digital and film-based 5x7-inch enlargements, the superiority of the film-based image is clear. It's not until you shoot with 2.1, 3.3, and 4MP cameras that the distinctions become less and less evident with size.

 

 

Contrary to some misconceptions, there's no film in "digital film!"

We have always tried to maintain the highest standards when it comes to rating image quality, and when digital cameras came along, we didn't change our standards to suit their lower resolution. We continue to base our analysis of image quality, whether the image is shot digitally or on film, on a number of factors that are important to advanced amateur and professional photographers. These include image resolution, color accuracy and saturation, skintone reproduction, shadow and highlight detail, contrast, sharpness, and graininess (or noise, in the case of digital).

We use a number of methods to determine all of the above. With digital cameras, we measure the lines of resolution a camera can capture from a standard IT-10 test target that is evenly lit by two strobes at 45 degree angles to the target. With film-based systems, the test target is a standard USAF 1951 resolution target and we list the resolution in terms of line pairs per millimeter. The IT-10 target is designed for digital camera testing and gives us resolution figures along horizontal, vertical, and diagonal directions, as well as a number of other image-quality factors, such as noise and contrast. Film-based systems usually have similar resolution in most directions and capture far more detail than can be measured by an IT-10 test target, so we usually give one figure for resolution. As we recently pointed out in our "7 Film Scanners Tested" article, the best SLR and lens that we've tested (loaded with ISO 100 color-negative film and tripod-mounted) can record 75 line pairs per millimeter on the film. A 4-megapixel camera can capture about 28 lp/mm (averaging horizontal and vertical numbers, which are slightly different). Since a 4MP sensor has slightly more than twice the number of pixels along both the vertical and horizontal axis than a 1.3MP camera, we could estimate the lp/mm of a 1.3MP camera (given a similar lens and focusing system) at about 13 lp/mm. In comparing resolution, the film advantage is obvious.

But do you need all the detail captured by film? That's where the confusion starts, and the answer depends on what you plan to do with your photos. Are you going to print your digital images or film scans, and at what size? Or will you e-mail them or view them only on a computer? If all you ever want to do is make 3x5- and 4x6-inch prints, then a 1.3 or 2.1MP camera can probably handle the job. In our visual comparisons of print quality, we've found that most viewers consider prints made from a decent 6-color inkjet printer or 4-color dye-sub printer to be of photographic quality when the image files are printed at a true 200-250 pixels-per-inch resolution. (That's image file resolution and not the DPI of the printer, which is another story.) For example, a 1600x1200-pixel image from a 2.1MP camera can be sized prior to printing at 6x8 inches (1200 pixels/200 pixels=6, 1600/200=8). So this image can be described as a 6-inch by 8-inch by 200-ppi file. We often prefer resolution to be set at 250 or 300 ppi for the very best quality results. At 300 ppi, the same 1600x1200-pixel image can produce a 5.3x4 inch print.

To cut to the chase: The resolution a digital camera captures can't always be used as the single determining factor for image quality. On the color front, digital cameras vary widely in their ability to accurately capture color. But the best ones do as good a job—or better—than film-based cameras. Shadow and highlight detail is another story. Even a single-use camera loaded with 800 ISO color-negative film captures greater detail in both shadows and highlights. This detail may not be obvious on a minilab print, but can be brought out with burning or dodging in a darkroom, or with a good film scanner. At the upper end, most digital cameras use CCD and even CMOS sensors that were designed for still-image capture and printing. These capture a much wider "latitude" of exposure values, but still have a tendency to blow out highlights.

The bottom line is that many of the newer 2.1-4 megapixel cameras have been designed from the bottom up with printed image quality in mind. So even if a 1.3MP camera has enough resolution to make a sharp 3x5-inch print, it might not match the overall image quality of a similar-sized print made with a higher resolution camera.