The re-encoding of data to make it smaller. Most image file formats use compression because image files tend to be large and consume large amounts of disk space and transmission time over networks.

The process of reducing the amount of bandwidth or data rate for video.

The broadcast standards used today – PAL, NTSC and SECAM – are analogue video compression systems. For digital systems, pictures are analysed looking for redundancy and repetition and so discard unnecessary data. The techniques were primarily developed for digital transmission but have been adopted as a means of handling digital video in computers and reducing the storage demands for digital VTRs.

Compression (amplitude)

1. (Data transmission) A process in which the effective gain applied to a signal is varied as a function of the signal magnitude, the effective gain being greater for small rather than for large signals.

2. (Video) The reduction in amplitude gain at one level of a picture signal with respect to the gain at another level of the same signal. Note: The gain referred to in the definition is for a signal amplitude small in comparison with the total peak-to-peak picture signal involved. A quantitative evaluation of this effect can be obtained by a measurement of differential gain.

3. (Production) A transfer function [as in gamma correction] or other nonlinear adjustment imposed upon signal amplitude values. cf. black compression; differential gain; image compression; transfer function; white compression.

Compression (bit-rate)

1. (Multimedia) Compression is used in the digital environment to describe (a) initial digital quantization employing transforms and algorithms encoding data into a representation that requires fewer bits or lower data rates or…

2. Processing of an existing digital bit stream to convey the intended information in fewer bits or lower data rates. Compression (bit rate) may be reversible compression, lossless or it may be irreversible compression, lossy.

Compression, lossless

1. Lossless compression requires that the reproduced reconstituted bit stream be an exact replica of the original bit stream. The useful algorithms recognize redundancy and inefficiencies in the encoding and are most effective when designed for the statistical properties of the bit stream.

2. (Video) Lossless compression of image signals requires that the decoded images match the source images exactly. Because of differences in the statistical distributions in the bit streams, different techniques have thus been found effective for lossless compression of either arbitrary computer data, pictures, or sound.

Compression, lossy

1. (Video and multimedia) Bit-rate reduction of an image signal by powerful algorithms that compress beyond what is achievable in lossless compression, or quasi-lossless compression. It accepts loss of information and introduction of artifacts which can be ignored as unimportant when viewed in direct comparison with the original. Advantage is taken of the subtended viewing angle for the intended display, the perceptual characteristics of human vision, the statistics of image populations, and the objectives of the display.

2. The lost information cannot be regenerated from the compressed bit-stream.

Compression, quasi-lossless (Video and multimedia)

1. Bit-rate reduction of an image signal, by an algorithm recognizing the high degree of correlation ascertainable in specific images. The reproduced image does not replicate the original when viewed in direct comparison, but the losses are inobvious or unrecognizable under the intended display conditions. The algorithm may apply transform coding, predictive techniques, and other modeling of the image signal, plus some form of entropy encoding.

2. While the image appears unaltered to normal human vision, it may show losses and artifacts when analyzed in other systems (i.e., chroma key, computerized image analysis, etc.)

3. The lost information cannot be regenerated from the compressed bit-stream. cf. bit-rate reduction; compression; compression, lossless; compression, lossy; image compression.

Compression ratio

The ratio of the data in the uncompressed digital video signal to the compressed version. Modern compression techniques start with the CCIR 601 component digital television signal so the amount of data of the uncompressed video is well defined – 75 Gbytes/hour for 625/50 and 76 Gbytes/hour for the 525/60 standard.

The compression ratio should not be used as the only method to assess the quality of a compressed signal. For a given technique, greater compression can be expected to result in worse quality; but different techniques give widely differing quality of results for the same compression ratio. At the same time results will vary depending on picture content. The only sure method of judgment is to make a very close inspection of the resulting pictures.