float get_msb_wmse( )
Java: float Get_msb_wmse( )
[Declared in "../coresys/common/kdu_compressed.h"]
Returns the contribution of a single bit error in the most significant magnitude bit-plane of any sample in this subband to overall image weighted MSE. The number of magnitude bit-planes is returned by get_K_max_prime. The weighted MSE contributions are expressed relative to a normalization of the original image data in which the sample values all have a nominal range from -0.5 to +0.5. This normalization framework is used for both reversibly and irreversibly processed samples, so that rate allocation can be performed correctly across images containing both reversibly and irreversibly compressed tile-components.