レコードをメール: Maximizing Audibility and Speech Recognition with Non-Linear Frequency Compression by Estimating Audible Bandwidth