mp3 audio format, the most popular
With the rapid development of file compression technology, MP3 has become the most popular music format today.
The encoder transforms the original sound into the frequency domain through a hybrid filter bank. Using a psychoacoustic model, it is estimated that it may be sufficient to be The perceived noise level is then quantized and converted to Huffman coding to form an MP3 bitstream. The decoder is much simpler and its task is to extract the sound signal from the encoded spectral line components through inverse quantization and inverse transformation.
2.4 Modified Discrete Cosine Transform Modified Discrete Cosine Transform (MDCT) refers to converting a set of time-domain data to frequency-domain data for time-domain variation. MDCT is an enhancement of the DCT algorithm. The first fast algorithm is the Fast Fourier Transform (FFT), but FFT has operations on complex numbers and MDCT are all operations on real numbers, which is convenient for programming. When compressing audio data, first divide the original audio data into fixed blocks, and then perform forward MDCT (Forward MDCT) to convert the value of each block into MDCT 512 coefficients. When decompressing, the reverse MDCT (Reverse MDCT) The 512 coefficients are restored to the original sound data, and the original sound data before and after are inconsistent, because redundant and irrelevant data are removed during the compression process. The FMDCT transformation formula is: k=0, 1,…, N/2-1 where N is the length of the transformation window, that is, the number of sample points per block, N=8, 16 ,… ., 1024, 2048. n0=(N/2+1)/2, X(n) is the value in the time domain, X(k) is the value in the frequency domain. If N takes 1024 points, it will become 512 frequency domain values. The IMDCT transformation formula is: 4 Modified Discrete Cosine Transform Modified Discrete Cosine Transform (MDCT) refers to converting a set of time-domain data to frequency-domain data to learn the changes in the domain. weather. MDCT is an enhancement of the DCT algorithm. The first fast algorithm is the Fast Fourier Transform (FFT), but FFT has operations on complex numbers and MDCT are all operations on real numbers, which is convenient for programming. When compressing audio data, first divide the original audio data into fixed blocks, and then perform forward MDCT (Forward MDCT) to convert the value of each block into MDCT 512 coefficients. When decompressing, the reverse MDCT (Reverse MDCT) The 512 coefficients are restored to the original sound data, and the original sound data before and after are inconsistent, because redundant and irrelevant data are removed during the compression process. The FMDCT transformation formula is: k=0, 1,…, N/2-1 where N is the length of the transformation window, that is, the number of sample points per block, N=8, 16 ,… ., 1024, 2048. n0=(N/2+1)/2, X(n) is the value in the time domain, X(k) is the value in the frequency domain. If N takes 1024 points, it will become 512 frequency domain values. The IMDCT transformation formula is: 4 Modified Discrete Cosine Transform Modified Discrete Cosine Transform (MDCT) refers to converting a set of time-domain data to frequency-domain data to learn the changes in the domain. weather. MDCT is an enhancement of the DCT algorithm. The first fast algorithm is the Fast Fourier Transform (FFT), but FFT has operations on complex numbers and MDCT are all operations on real numbers, which is convenient for programming. When compressing audio data, first divide the original audio data into fixed blocks, and then perform forward MDCT (Forward MDCT) to convert the value of each block into MDCT 512 coefficients. When decompressing, the reverse MDCT (Reverse MDCT) The 512 coefficients are restored to the original sound data, and the original sound data before and after are inconsistent, because redundant and irrelevant data are removed during the compression process. The FMDCT transformation formula is: k=0, 1,…, N/2-1 where N is the length of the transformation window, that is, the number of sample points per block, N=8, 16 ,… ., 1024, 2048. n0=(N/2+1)/2, X(n) is the value in the time domain, X(k) is the value in the frequency domain.
