Page History

...

The low-resolution coefficients are encoded using a modified form of Simple Packing. Their precision is controlled by bitsPerValueEach packed integer value uses bitsPerValue bits. The decimal scale factor and binary scale factor are chosen based on bitsPerValue.

The packed integer value is calculated approximately as follows: 

packed_value = round(((x * 10^D * (n(n+1))^p) - R) * 2^(-E))

...