File Format Pros and Cons


All the technique in the world won't help you if you capture the image in the wrong file format. With a world of format options available, the challenge is to determine the proper file type and compression scheme for the type of work you're doing. This is a challenge because you hear so many differing opinions. Some people will tell you that JPEGs are fine, while others shake their heads in scorn if you capture in anything other than uncompressed 16-bit TIFFs. Who's right, and how do you go about making a decision?

What Are My Choices?

The primary variables in this process are file format, bit depth, and compression schema. Before you decide on a strategy for format and compression, you first need to understand what these variables mean.

File Format

File format (also called file type) refers to the type of file generated by the camera as it saves the image. The file type determines the amount of information in the file, how it's structured, and how it can be accessed and edited by programs such as Photoshop. For most users, the choice is between TIFF files and JPEG files.

The TIFF file is the standard format in the imaging world, making it the obvious choice for many digital photographers. The problem with TIFF files is file size; they tend to get pretty big. You can compress TIFF files using a lossless compression mode called LZW, which makes the files a bit more manageable and easier to store. Lossless compression means that the image quality is not degraded in the process of compression, and for many it's the perfect compromise between compression and optimum quality. The downside to LZW files is that they are still much larger than their JPEG counterparts, and they take a bit more time to open and close in Photoshop because Photoshop has to run the compression and decompression algorithms in each instance.

Many digital camera users love to save files in JPEG format because the files are extremely small, and you can fit hundreds of files on a single media card. The downside of JPEG files, of course, is that you do loose a bit of detail. Is it enough for you to notice, and does it really matter? I'll explore these issues more in the next section, so stay tuned.

 Window
Nikon D2H,
1/60 sec, f/8
EV +1
Focal length 14mm

 Pienza Texture
Nikon D2H,
1/60 sec, f/4
Focal length 28mm

8-bit Versus 16-bit

More advanced cameras allow you to capture images in 16-bit mode, as opposed to the standard 8-bit mode used by conventional digital cameras. The number of bits in an image (referred to as the bit depth) determines the number of gray tones it can support. 8-bit images support 256 shades of gray, as calculated by two to the eighth power. As the bit depth increases, the gray scale expands exponentially. In 12-bit images, the number increases to 4,096 shades of gray, and in 16-bit images there are 65,536 shades of gray.

 Palio Procession
Nikon D2H,
1/60 sec, f/5.3
EV +1
Focal length 300mm

 Palio Banner (crop1)
Nikon D2H,
1/125 sec, f/5.3
ISO 640
Focal length 78mm

But wait a minute; you might be asking what a grayscale range has to do with a color image. In an RGB image, each color channel is rendered in a grayscale model that represents the distribution of the channel color (red, green, or blue) throughout the image. In Photoshop, select Window, Channels to open the Channels palette and click the red channel; you'll see a grayscale image representing the red distribution. In an 8-bit image, that red channel is composed of 256 shades of red; in a 16-bit file, there are 65,536 shades of red. More shades mean smoother transitions and gradations, less banding, and a much smoother image.

The bit depth is critical for image editing because each time you edit in Photoshop, you erode image quality, literally throwing away pixel information. Stretch a curve or Levels slider, and you create gaps in the tonal range as the histogram expands and shifts (a phenomenon known as combing because of the comb shape it creates in the histogram). This process diminishes quality in both 8-bit and 16-bit images; however, the 16-bit images look much better because of the additional image data resident in the file.

If the quality loss is so significant, why don't more people notice it? One reason is that the overlap of image data across color channels tends to mask banding within each individual channel, creating a quality loss that's often felt rather than pinpointed directly. Also, the more editing you do, the more chronic the problem becomes.

Another reason you might not notice the quality degradation is that many people using 8-bit files don't even try to do the things a 16-bit file can do. Try increasing the exposure by two or three stops on an 8-bit image, and things get ugly in a hurry. The image can pixelate in the shadows, take on color cast, or create flat, solarized areas that are void of information. People train themselves not to try these kind of corrections, even with 16-bit images. I find the extra latitude provided by 16-bit files to be indispensable when I work in Photoshop, especially when I'm exploring expressive exposure options within an image.

The downside is that 16-bit files are much larger than their 8-bit counterparts. For example, a 16-bit file that's 23MB reduces to just 11MB in 8-bit mode. Not only that, but each time you add another layer, the file size bumps up even more. That same 16-bit file with three total layers weighs in at a whopping 84MB; and that's just three layers. With 10 layers, the file balloons to 299MB. Change the file to 8-bit, and the total size drops back down to 23MB, including all 10 layers.

Thus, on the one hand, 16-bit files show better quality, especially if you're doing post editing in Photoshop. On the other hand, 16-bit files are huge. Many people choose to shoot and edit in 16-bit mode and then convert to 8-bit in Photoshop by selecting Image, Mode, 8 Bits/Channel, to save file space. This is a viable option as long as you realize that downsizing the file results in a loss of file info that can never be regained.

 Palio Banner (crop2)
Nikon D2H,
1/125 sec, f/5.3
ISO 640
Focal length 78mm

 Siena Boy (crop)
Nikon D2H,
1/60 sec, f/6
Focal length 80mm

Compression Options

Compression can be a good thing. It makes files small and compact, easy to store, and easy to attach to emails. Compression can also be a bad thing. It can erode image quality, making things looks fuzzy and grainy. There are several compression schemes available today, but for the digital photographer, the only real options are JPEG files and LZW compression for TIFF files.

TIFF image files optionally use LZW lossless compression, meaning that there is no quality loss caused by the compression, and the file is kept in its original state.

 Siena Boy
Nikon D2H,
1/60 sec, f/6
Focal length 80mm

Compression works by recognizing repeated pixel values in an image, replacing the many instances in shorthand fashion with just one instance. As a result, LZW is most effective when compressing images with large flat areas of color. Featureless areas compress better than detailed areas. In addition, LZW is more effective for grayscale images than color, and it might not be effective at all for 16-bit images.

The JPEG file is named for the Joint Photographic Expressions Group, which documented and defined the file format. JPEG files allow for a high rate of compression, resulting in files that are much smaller than TIFF files, even if those TIFF files use LZW compression.

However, JPEG files sacrifice a small amount of quality as a result of the compression process. But this small degree of quality is not the primary drawback. The dangerous thing about JPEGs is that the degradation in quality occurs each time an image is opened and saved. Open and close a file once, and you don't see much impact on overall quality. Open and save it 10 times, and it will look exponentially worse than the first time you open it. Thus, like the proverbial frog sitting in the simmering pan of water, the quality erodes over time in such a way that you don't even notice itbut by the time you do, it's irrevocable.

How Do I Decide?

All the file formats and compression options described thus far are good choices, depending on the circumstance in which they're being used. Obviously, you don't want to post a 16-bit TIFF file on the Internet, and you wouldn't want to repeatedly save the only copy of your wedding pictures as JPEG files. The best approach is to keep a series of files, each in a different format, for the various stages of work you're doing.

Master Files

The first step is to create and maintain pristine copies of all the images that are important to you. Think of them as master files you can copy and manipulate as needed. These mint-condition originals should be captured and maintained using the highest quality resolution and the broadest exposure range, being careful not to clip the highlights or flatten the shadow areas. Because I use Nikon equipment, I save all my files as .nef files, which are pretty small while still maintaining the 16-bit depth and all the original file data. The important thing is to save your files in the 16-bit RAW format supported by your system.

Edit Files

Whenever you want to modify an image, duplicate the master file and begin editing from there. You might feel strongly that you know just what you want to do with an image and that it's safe to edit the master file, but I advise against it. I've made the same mistake many times; I edit the original and end up wishing I could go back and crop a different way, or set a different exposure emphasis. These digital files are going to be around a long, long time, and hard drive storage is cheap. Duplicate the file and save the original master; doing so keeps your options wide open and is always the safest way to go.

Special-Purpose Files

After you've edited the file the way you want it, you might find yourself saving multiple copies of the same image for special purposes. Perhaps it's to post on the web, attach to an email, or reduce to a smaller size. This is the time to turn to lossless compression such as JPEG. You already have the original master as well as an edited copy, so it's less critical if you over-compress these iterations. In Photoshop, select File, Save As or File, Save For Web to open the appropriate dialog boxes.

 Galgano Window
Nikon D2H,
1/160 sec, f/9.5
Focal length 24mm




The Art of Photoshop for Digital Photographers
The Art of Photoshop for Digital Photographers
ISBN: 0672327139
EAN: 2147483647
Year: 2006
Pages: 141

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net