Audio compression reduces file weight through psychoacoustic modeling and sample rate reduction:

• Psychoacoustic Masking (MPEG Layer III): The human auditory system cannot perceive quiet sounds occurring immediately before or after louder frequencies (temporal masking), or frequencies drowned out by adjacent louder tones (simultaneous masking). The encoder discards imperceptible frequency data, reducing file weight by 70% to 85%.
• Bitrate Selection & File Size Formula ($S = \frac{B \times T}{8}$):

$$\text{File Size (KB)} = \frac{\text{Bitrate (kbps)} \times \text{Duration (seconds)}}{8}$$

• 64 kbps: Ultra-compact speech for podcasts and voice notes ($0.48\text{ MB/minute}$).
• 128 kbps: Standard web streaming and mobile distribution ($0.96\text{ MB/minute}$).
• 192 kbps: Balanced high-fidelity music ($1.44\text{ MB/minute}$).
• 320 kbps: Pristine master broadcast quality ($2.40\text{ MB/minute}$).
• Stereo to Mono Channel Downmixing: Voice recordings rarely require stereo separation. Converting stereo to mono halves the channel data, reducing file size by 50% before compression without degrading vocal clarity.
• Sample Rate Optimization: Downsampling from $48\text{ kHz}$ to $44.1\text{ kHz}$ or $32\text{ kHz}$ strips ultrasonic frequencies beyond the human hearing threshold ($20\text{ kHz}$).