An encoded-data converting apparatus and method are provided. First encoded data generated by using first encoding to encode image data is input. Header information included in the first encoded data is extracted. Header information based on second encoding is generated from the header information included in the first encoded data. Variable-length-code conversion is set up based on a predetermined parameter for use in the second encoding and an encoding parameter used in the first encoding. Encoded data in one frame is extracted from the first encoded data. A variable length code in the encoded data in the frame is converted in accordance with the variable-length-code conversion. DC components in the encoded data in the frame are decoded into quantized values and the quantized values are used to perform predictive encoding based on the second encoding. The obtained outputs are shaped and output as data encoded by the second encoding.