Quality scalable coding with mapping different ranges of bit depths

A method for decoding a scalable data stream into which a picture is encoded, the scalable data stream comprising base layer data representing the picture with a first sample bit depth and enhancement layer data representing a prediction residual associated with a second sample bit depth that is higher than the first sample bit depth, the method comprising:

• decoding the base layer data into a lower bit-depth reconstructed picture;
• decoding the enhancement layer data into the prediction residual;
• generating a prediction of the picture based on the lower bit-depth reconstructed picture, the prediction associated with the second sample bit depth, wherein the generating includes performing, on the lower bit-depth reconstructed picture, a mapping of samples of the picture based on the first sample bit depth, the second sample bit depth, a bit-shift operation, and an offset value that depends at least on the first sample bit depth to obtain the prediction of the picture;
• and
• reconstructing the picture with the second sample bit depth based on the prediction and the prediction residual.

An encoder for encoding a picture into a scalable data stream, comprising:

• a base encoder for encoding the picture to obtain base layer data corresponding to a representation of the picture with a first sample bit depth;
• a predictor for providing a prediction of the picture based on the representation of the picture with the first sample bit depth, the prediction associated with the second sample bit depth that is higher than the first sample bit depth,
• the predictor including a mapper, which comprises hardware or a program executed using a computer, for performing, on the representation of the picture with the first sample bit depth, a mapping of samples of the picture based on the first sample bit depth, the second sample bit depth, a bit-shift operation, and an offset value that depends at least on the first sample bit depth to obtain the prediction of the picture;
• a residual encoder for encoding a prediction residual of the prediction to obtain enhancement layer data;
• and
• a combiner for forming the scalable data stream based at least on the base layer data, and the enhancement layer data.

A non-transitory computer-readable medium for storing data associated with a video, comprising:

• a data stream stored in the non-transitory computer-readable medium, the data stream comprising a scalable data stream related to an encoded picture of a video, wherein the scalable data stream is formed for storing in the non-transitory computer-readable medium by operations including:
• encoding the picture to obtain base layer data corresponding to a representation of the picture with a first sample bit depth;
• generating a prediction of the picture based on the representation of the picture with the first sample bit depth, the prediction associated with the second sample bit depth that is higher than the first sample bit depth, the generating including mapping of samples of the representation of the picture based on the first sample bit depth, the second sample bit depth, a bit-shift operation, and an offset value that depends at least on the first sample bit depth to obtain the prediction of the picture;
• encoding a prediction residual of the prediction to obtain enhancement layer data;
• and forming the scalable data stream based on a combination of the base layer data and the enhancement layer data.