In a recent article here on GraphicDesign.com, Understanding File Formats in Photoshop author James George explained some of the most common file formats used for graphic design in Photoshop. For photographic imaging, some different file formats are best used for specific purposes. Following are the most common formats along with basic descriptions of their use.
A raw capture records all the image data from a camera’s sensor with only a very minimal amount of processing in the camera required to save the file to the memory card. A raw capture contains the most possible data from the camera and thus represents the best possible quality.
Raw files come from your camera with a format specified by the camera manufacturer; common examples are Canon’s .CR2 and Nikon’s .NEF formats. Native raw files are encoded and cannot be directly modified. And though they may have the same file extension, raw files from different cameras are often programmatically different from one another: a .CR2 file from a 5D Mark II is different than one from a 50D. Raw files in your camera model’s native format must be recognized by your photo software for you to work with those files. Current imaging software supports nearly all digital cameras available on the market, and support for new models is continually updated. However, when a new camera model is released, there may be a period of time during which your files cannot be read by the software until an update is released.
DNG stands for Digital Negative, an open-source image file format developed and standardized by Adobe. Especially if you capture raw, you should seriously consider integrating DNG in your workflow. Though some photographers use DNG just for archival copies, I believe this is missing the real benefits. I use DNG for my master working files—I usually capture in raw and then convert the raw files to DNG as one of the first steps in my workflow.
With raw files from most common cameras, converted DNGs preserve all the image data from the original raw file. (The raw data in a DNG is most often demosaiced, but otherwise unmodified, so the results of processing in Lightroom are the same as with native camera raw files.) A DNG file also provides the ability to to save XMP metadata directly into the file, eliminating the need for the sidecar files required to save edits for native raw files.
Adobe has consistently demonstrated a commitment to improving and advancing the format and has recently announced a new specification 1.4. The new DNG spec provides functionality for much faster processing, smaller files and enormous flexibility. With the new DNG spec you can also resize DNGs as they are created and apply optional lossy compression.
(Note: I’m a big believer in DNG and it’s great for my workflow. You should learn more about it to decide if it’s right for you; there’s lots of information about DNG on the Web.)
TIP: Convert JPEG to DNG
Using the latest versions of Lightroom or Adobe Camera Raw, you can convert JPEGs to DNG to edit those photos non-destructively and preserve the most possible quality.
Can be used for Web, email and print. Standard JPEG files always use 8-bit data and can use any color space. Exporting as JPEG requires you to specify a Quality level, using numeric values from 0 to 100. The Quality setting determines the amount of compression to be applied to exported JPEG files. JPEG compression is always lossy—data is discarded as the file is saved in order to reduce file size. The higher the Quality setting, the larger the file size will be. A higher Quality setting will apply less compression, providing better-looking (and larger) files, while with lower Quality settings more compression is applied, potentially resulting in degraded image quality. With JPEG compression, some data is always discarded, even at the maximum Quality setting. If you’re exporting JPEG files to send to a lab for printing, you should usually set the Quality at 100, unless instructed otherwise. For Web use, I usually limit the JPEG Quality to between 70 and 80, unless I need to adhere to provided specifications.
TIP: JPEG quality levels in Lightroom and Photoshop
Lightroom’s JPEG Quality slider values are different than Photoshop’s Save As options for JPEG, where you set the quality level using an integer from 1 to 12. In Lightroom, Quality 100 is equivalent to Photoshop’s level 12 quality. Lightroom’s Quality setting of 84 is similar to Photoshop’s level 10.
PSD (Photoshop Document)
You can use PSD to work on photos in Photoshop or Elements. Exporting as PSD allows you to set the bit depth (see below). A Photoshop-format file can include layers, masks, transparency, vector elements and type, as well as many specialized elements used in graphic design such as 3D objects. Photoshop files use a proprietary format, so you can only open these files using supported software, unlike the open-format TIFF described below.
TIFF (Tagged Image File Format)
TIFF can be used for virtually any purpose except viewing in a Web browser. TIFF is often used as the format for working Master image files and derivative files for printing. When exporting TIFFs you can specify bit depth and apply compression. Compression on TIFF files is lossless—no data is discarded. Compressed TIFF files can be significantly smaller than uncompressed; however, large compressed files can take longer to open and save than uncompressed files. (I usually prefer to work with uncompressed TIFF files.) TIFF files can contain nearly everything a PSD can contain; although layers are preserved they are usually not editable outside Photoshop or Elements, though common file attributes such as alpha channels can often read by other software.
TIP: for photographs, use TIFF instead of PSD
I generally recommend you don’t use PSD files for your Photoshop work—use TIFF instead. TIFF provides all the same capabilities as PSD (layers, type, alpha channels, etc.), with a distinct advantage: TIFF is a non-proprietary, open file format specification. This offers much greater viability for long-term archival than does PSD. Also, you can easily import TIFFs into Adobe Lightroom, whereas with PSDs, you have to save them with Maximize Compatibility enabled.
Note: When compared with a PSD, the file size of a TIFF will often be larger; this is due to the fact that a TIFF file always contains a flattened preview by default.
Bit Depth (TIFF and PSD only)
When saving TIFF or PSD files you have a choice of 8 bits/component or 16 bits/component. This sets the amount of data used to render the file. Higher bit-depth means more data is being used to describe the pixels in the file. An 8-bit file uses 256 levels of brightness per color channel; a 16-bit file uses 65,536 levels.
Most often, more data equates to higher quality, especially if you’re doing lots of editing on the file. If you are saving layered files to be used as Masters for image editing, I recommend you use 16-bit whenever possible. A 16-bit image provides much more processing flexibility than does an 8-bit image—16-bit data can be manipulated further before the appearance of the image starts to degrade. 16-bit allows smoother transitions between colors and reduces the appearance of posterization (where areas of color become solid and transitions become hard-edged). One example of image degradation problems is the appearance of visible bands in what should be a smooth gradient, such as a blue sky. Adjustments to a smooth blue sky in 8-bit will start to break down the image data and reveal banding or posterization more than when working in 16-bit.
If you’re going straight from capture to print, without editing, 16-bit won’t usually offer a benefit. An example would be someone shooting digital JPEG and outputting directly to a printer, which is common for wedding and portrait photographers. (Also, the most common JPEG standard does not support 16-bit color depth.)
So if you feel that some of your images won’t need much adjusting in Photoshop, or you have images where you’re not concerned with quality, it’s usually OK to work in 8-bit. However, a professional workflow requires capturing the maximum amount of data possible from the beginning, and working with as much data as possible throughout the image editing workflow. This is especially true with fine art nature and landscape photography.
When you’re exporting final files for printing, 8-bit is usually appropriate. (However, some printers can take advantage of 16-bit; ask your vendor how they want the files saved.)
Video formats: digital movies from most popular cameras and smart phones are recorded using a variety of formats, including MOV, MPG, AVI and AVCHD. Single frames from videos can be exported as one of the other still image formats.
In a typical digital photography workflow you may use several — if not all — of these common image formats. Generally speaking, you want to start with a raw capture, convert to DNG at some point in the workflow, edit the raw or DNG files in your parametric editor (Lightroom, ACR, Aperture, etc.), use TIFF for working on layered Master files in Photoshop, etc., and finally export finished TIFF and/or JPEG files for printing and sharing.