Taking Sound Recordings For Environmental Cue Model Development (#59)
Generating sound models for training devices is as much an art as engineering science. With proper data, however, the sound engineer has a much greater chance of producing the highest fidelity aural cueing environment possible for the training device. This document provides general guide-lines and recommendations for obtaining sound recordings for use in environmental/aural cue model development.
The application note covers:
- Recording Plan and Recommendations
- State Data Correlation and Recording
- Recording Guidelines and Tips
- Recording Equipment
- Microphone Types and Tips
The purpose of obtaining sound recordings from the reference platform (aircraft, helicopter, vehicle, etc.) is to provide the sound simulation engineer sufficient and accurate sound and engineering data for production of the trainer sound simulation model.
The recordings themselves are used in two ways. First, the audio data is used as the reference baseline for the sound model development. Through the course of development, the sound engineer will tune and compare the modeled sounds against the base-line recordings in an effort to replicate the recorded data as close as possible. Second, for greater fidelity and accuracy, some portions of the recordings will be used as part of the final sound model. This emphasizes the need for a quality recording.
For level D commercial simulators, use of the recordings is taken one step further whereby the recordings become part of the test and validation process. In these cases, recordings are also taken in the training device at pre-identified simulator test states (e.g. 300knots, 10,000ft, Engine Cruise). Sound production in the trainer is analyzed and compared to the original platform reference recordings for accuracy. Again this emphasizes the need for high quality recordings.
In this respect the phrase GIGO (garbage in, garbage out) applies. If recordings are poor and insufficient information is provided, then the environmental cue modeling will be at best an approximation and at worst a poor representation (read hokey) that bears no resemblance to the actual reference platform sound environment. Hence, every effort should be made to obtain the highest quality recording possible.