JPEG Quality and Compression

Uses of the JPEG (Joint Photographic Experts Group) file format are far and wide. Initially, the file format was intended for use in Web applications, but has found a home in the high-end print production markets, as well. The JPEG file format can be your best friend if used properly, or your worst enemy if implemented incorrectly.

A JPEG file is encoded by using an adjustable lossy compression approach. This means that to achieve smaller file sizes, image data is actually thrown away. In small doses, the JPEG compression approach can be very effective and efficient. However, in larger amounts, the resulting file will contain "noise" and undesired artifacts in the image.

As you apply greater amounts of JPEG compression, and image gets smaller in size. However at this reduction of file size comes the deterioration of image quality.

Take a look at these images. They all have been compressed at different levels of JPEG. Since these artifacts are seen most easily on images with hard edges, the examples are shown using the monochrome sample image.

	1. The control image. This is an uncompressed image. Notice how this image has crisp and clean edges. This image will reproduce well on press.
	2. Image with a large amount of JPEG compression applied.   Notice that this image is much softer on the edges, and they are "fuzzy". If this image were to go to press, there would be even more "noise" and artifacts produced in the finished product. Generally you can not see this noise on a computer monitor.
	NOTE: This image have been simulated to reflect the appearance when they will be printed.

Since the lossy compression approach actually removes image data, once an image has been compressed (or over compressed), the damage done is permanent. If you try to open the over-compressed image in Photoshop and convert it to a TIFF image to repair the problem, you are only masking the true problem. Unfortunately there is no repair for an over compressed JPEG image.

   See also  "The Facts About JPEG Compression"