The low bit rate (192 kilobits per second) conversion encoder / decoder system (44.1 ms or 48 ms sampling rate) for high quality music uses a short time-domain sample block (128 sample blocks) Short enough for real-time auditory feedback to the operator. Carefully designed pairs of analysis / synthesis windows are used to achieve sufficient conversion frequency selectivity despite the use of short sample blocks. The synthesis window of the decoder has the property of producing a complex response where the product of the response and the response of the encoders analysis window is summed as one for two adjacent overlapping sample blocks. Adjacent time domain signal sample blocks are superimposed and added to remove the effects of analysis and synthesis windows. Techniques for deriving appropriate analysis / synthesis window pairs are provided. In an encoder, a discrete transform or alternatively a discrete Fourier transform with a function by alternating application of modified DCT and modified DST by time-domain aliasing cancellation techniques is used to generate a frequency domain transform coefficient. The transform coefficients are nonuniformly quantized by assigning a fixed number of bits and a variable number of bits adaptively determined based on psychoacoustic interception. Techniques for assigning fixed bits and adaptive bits are described. The transmission of the measurement beams for adaptively allocated bits is not necessary. Error codes and protected data can be distributed through the formatted frame output from the encoder to reduce sensitivity to noise bursts.