IMPAX performs lossless JPEG compression on almost all incoming files, maintaining image quality for primary diagnosis. Images that are already compressed in a supported format by the transmitting station, such as MPEG, are not lossless JPEG compressed.
In some circumstances, to reduce the transmission time for non-diagnostic purposes, you may want to lossy compress an image using JPEG compression or wavelet compression before transmitting it. To use lossy compression when transmitting images, enable compression for the station you are transmitting to. When using wavelet compression, files are compressed using a proprietary compression algorithm and can only be uncompressed by systems that support that proprietary algorithm.
To configure how lossy JPEG and wavelet compression algorithms are used based on the station and modality type, use the Compression panel in Network Management (Image tab, 
). For lossy JPEG compression, specify the Q Factor to use; for wavelet compression,
specify the compression ratio.
 | CAUTION! Compression ratio set in Network Management do not apply only to transmitted images. If lossy JPEG compression is enabled for image caches or archives (in Source Manager, Compressor tab), the compression ratios also apply to those images. Therefore, use extreme caution when selecting higher compression ratios for transmission, as these values may also affect compression used in the image cache and archive. |
When using lossy JPEG compression, a compromise between image size and quality is accomplished by specifying the Q Factor (quality factor). For example, a higher-quality image creates a larger image size. The Q Factor is applied only to stations that have compression enabled (whether in Network Management or in Source Manager).
Configure the Q Factor for specific stations or modalities as appropriate. If a Q Factor is configured for both a modality type and a particular station, the station configuration takes priority over the modality configuration. For example, multiple 12-bit CRs can be set by modality and have a lower Q Factor; however, if an 8-bit CR station also exists, it should be configured explicitly by station name with a higher Q Factor. A conservative default value should also exist to ensure that image quality is sufficient for stations or modalities that are not explicitly specified.