Issues in Compression Method Selection in a Codec
 

• Lossless or lossy.  Application/coding efficiency - dependent.
• Coding Efficiency.  Use optimal coding efficiency within given constraint instead of the most efficient.
• Consistency in coding efficiency.  Large variations may not be acceptable among different data sets; e.g. synchronization problem.
• Resilience to transmission errors.  Robustness in transmission across noisy channel may be desired.
• Complexity trade-offs.  Low overall encoder-decoder complexity, or just low decoder complexity?
• Nature of degradation in decoder output.  Degrees of acceptance of artifacts introduced by lossy compression; which also may interact with errors in transmission across a noisy channel.
• Data representation.  Multiphase decoding such as adding detail information onto a rough image may imply the requirement of a hierarchical representation of the data.  This type of scheme is termed scalable compression methods.
• Multiple usage of the encoding-decoding tandem.  In lossy compression, resiliency of data may be essential when encoded and decoded in multiple stages.
• Interplay with other data modalities, such as audio and video.    Commonality in compression across data modalities is useful.  For example, a audio frame should be consistence with its corresponding compressed video frame in order to reduce timing errors.
• Interworking with other systems.  Transcoding from one compression method to another may be needed in an environment with multiple data modalities.  For example compressed data obtained using frame-by-frame editing (does not exploit temporal redundancy) from one codec may be passed to a different codec for broadcasting which does exploit temporal redundancy.