Home
Our Services
DSP Applications
Audio
Optimization of MPEG1/2 (Layer 1, 2 & 3) decoder
Optimization of MPEG1/2 (Layer 3) encoder
MPEG1 Layer 3 is prominently called as 'MP3'. MP3 is an audio coding technology which compresses the audio data (pcm samples) based on the human hearing capabilities. MP3 technology uses the psychoacoustic model to assess the audio signal perceivable by the human ear.
Psychoacoustic model is basically used to analyze the audio signal in the frequency domain. As the conclusion of the analysis, the inaudible frequency lines are allocated with less number of bits than the frequency lines which are audible to human. The bit allocated to the audio frame is processed iteratively using quantization technique. Finally, it is coded using Huffman coding technique.
ARM946E-S with 8KB instruction cache, 8KB data cache, no TCM (Tightly Coupled Memory).
| ROM (Bytes) | RAM (Bytes) | ||
|---|---|---|---|
| Code | Table | Stack | Data |
| 25676 | 22801 | 42802 | 860 |
| Layer | MIPS |
|---|---|
| Mode 0 | 105.96 |
| Mode 1 | 92.70 |
| Mode 2 | 60.83 |
| Mode 3 | 47.8 |