File Format Overview
On this page, we outline the various file formats that are used in digital imaging and how to choose between them. After defining what a file format is, we define rendered and unrendered formats in plain English, discuss format attributes, best uses of the various formats, and finally some information about format obsolescence.
A file format is the structure of how information is stored (encoded) in a computer file. File formats are designed to store specific types of information, such as JPEG and TIFF for image or raster data, AI (Adobe Illustrator) for vector data, or PDF for document exchange. dpBestflow® is primarily concerned with digital image file formats: what they are, the differences between them, and how best to use them.
The fundamental difference between a rendered image file format and an unrendered (raw) image file format is that a rendered file format is ready for use with a bitmap graphics editor, and an unrendered file format generally requires a conversion process, known as demosaicing, before it can be used in a bitmap editor.
Raw image files are called "raw" because they contain unprocessed (or minimally processed) raw data from an image sensor. In most cases we are assuming this to be a digital camera sensor, although raw files can also come from film scanners and some digital video cameras.
Unprocessed image data cannot be used with a bitmap graphics editor like Photoshop until they are processed into rendered files. Raw files are actually grayscale, recording only a luminance value for each pixel where each photosite represents a pixel. Color is created by assembling the luminance values, which are filtered through a color filter array in place over the photosites.
Assembling and interpreting the luminance information after it passes through the color filter array is the job of the raw converter software. This process converts raw image files to rendered image files. The control settings in combination with the raw converter's mathematical models create a rendered file, which has the result of these settings being "baked in" or fixed. Once this has been done, the file has been "rendered", and any additional image edits will be pixel-based edits.
Some image file formats such as TIFF, PSD and PDF, support layers. Some file formats do not. Only file formats that support layers are really useful for serious image editing and for archiving working files, known as master files.
The reason is that layers are the way to do non-destructive pixel image editing in raster image editors like Photoshop. For this reason, we recommend that your working files should be TIFF, or possibly PSD.
Bit depth defines how many bits of tonal or color data are associated with each pixel or channel. For example, two bits per pixel only allows for black or white. Eight bits provides 256 grayscale tones or colors. When referring to an 8-bit color image, 256 is multiplied (256x256x256) by the three primary (RGB) channels to create what is commonly called 24-bit color (with a possible 16,777,266 colors).
Digital cameras, monitors, and printers increasingly support higher than 8-bit depth color. Higher bit-depth allows more distinct colors and smoother transitions between colors. Image file formats support different levels of bit-depth. For instance, GIF format supports 8 bits per pixel, standard JPEG supports 24 bits per pixel, and other file formats can support up to 64 bits per pixel.
Support for high bit depth is an important attribute for working file formats. HDR (high dynamic range) formats have 32-bit-depth to support the increased brightness range of these image files. Of course, these 32-bit files need to be compressed down to 16-bit or 8-bit depth for viewing or printing – which is all part of the HDR imaging process called tone mapping.
The most commonly used rendered file formats are TIFF and JPEG. PSD is the Photoshop native file format, and many use it for creating master files or doing other complex Photoshop work. PSB (Large Document Format) is a variant of PSD, and is designed for images that exceed the 30,000x30,000 pixel limit of PSD. PNG and GIF are most commonly used for display on the web. EPS (Encapsulated PostScript) format is as commonly used as previously. As its name implies, EPS is designed to place image data within another PostScript document.
A new JPEG variant that is gaining acceptance is JPEG XR, previously known as HD Photo or Windows Media Photo before that. Another JPEG variant is JPEG 2000. JPEG 2000 has not been widely adopted by photographers since raw and DNG are considered to be better formats for their needs.
Raw file formats
Unrendered, or raw, file formats are the best formats for digital capture. In fact, JPEG capture is the result of raw capture processed to JPEG by the camera. Raw formats come in two varieties: so-called proprietary raw, and standardized raw, of which there is currently only one, the DNG (digital negative) format.
Read more in Raw File Formats
Knowing which file format to use for each stage of digital imaging workflow is key. The chart below gives you this information at a glance. For more complete information, use the following links:
|Figure 1 This chart summarises the best file formats to use for each stage of digital imaging workflow.|
A major challenge with regard to the preservation of digital image files is the long term readability of file formats. This is especially true if they are proprietary, which describes most camera maker’s raw formats.
Read more in Archive File Formats section