Compression

Compression is a method of encoding a file to smaller file size. JPEG (Joint Photographic Expert Group) file format enables lossy and non-lossy way of compression. By non-lossy compression, each pixel is saved as unique 24-bit information. In this case, the file size is compressed into smaller size with standard compression algorithm without loosing of any color information. The compression ratio in this case is low. When the file is restored all of color information are restored without loosing any color information.

In lossy compression, the image is recalculated into 8x8 pixels blocks. This compression algorithm use averaging encoding and is created on basis of human eye perception in that way that the picture is still acceptable for visual applications, but some data are discarded entirely. When the image is restored, some information are lost, but this is not so critical as it looks. Lossy compression can reduce file size to 10:1 ratio without visual difference. The difference is visible when image is magnified till each pixel is visible.

The effect of compression appears in two ways. On section 'A' on this picture, the consequence of compression (Photopaint 50 steps compression ratio) is visible when magnified.

JPEG uncompressed, 6x magnified
JPEG 50 steps compression. The picture is still acceptable for screen applications.
JPEG 100 steps compression. The picture is too low quality even for screen viewing.

If the file is resaved more times with the same compression ratio, the quality of image decreases each time when file is saved. When updating the images which will be compressed for final use, the processing should be done with original image.

Example: decreasing of image quality by resaving with same compression ratio: original
5x resaved picture by 10 steps compression ratio.

 Last change 04 januar 2002