Effect of Digital Recording Parameters on Discrimination Features of Acoustic Signals in Noise.

Abstract

An experiment was performed to determine the effects of digital signal processing sample rate and quantization code on auditory perception of sonar signals. Fifteen sonar signals were sampled and played back under nine conditions of sample rate and quantization code. In each condition all pairwise combinations of these signals in noise were presented to 35 subjects in an ABX discrimination task. The resulting matrices of discrimination errors were analyzed by multidimensional scaling. The first two scaling dimensions recovered in order of statistical significance were associated with perceptions related to (1) signal beat at low frequencies and (2) signal spectral shape in the higher frequencies. Further recovered dimensions were related to particular temporal modulation of individual signals. The importance of the first three discrimination features depended on the three sample rate conditions. Each halving of the sample rate removed one of the features from any significant contribution to the discrimination task. The quantization conditions had little influence on the significance of the discrimination features except for the mid-range sample rate (6.25 kHz).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 11, 1996
Accession Number
ADA313941

Entities

People

  • Joseph S. Russotti
  • Thomas E. Hanna
  • Thomas S. Santoro

Organizations

  • Naval Submarine Medical Research Laboratory

Tags

Communities of Interest

  • Ground and Sea Platforms

DTIC Thesaurus Topics

  • Acoustic Signals
  • Auditory Perception
  • Background Noise
  • Biomedical Research
  • Computer Programs
  • Detection
  • Digital Signal Processing
  • Electrical Engineering
  • Frequency
  • Frequency Bands
  • Frequency Modulation
  • Modulation
  • Perception
  • Processing Equipment
  • Signal Processing
  • Sonar Signals
  • Underwater Sound

Readers

  • Acoustical Oceanography.
  • Mathematics or Statistics
  • Speech Processing/Speech Recognition.