Abstract
Dynamic range compression (DRC) is a widely-used compensation strategy in hearing aids. However, the choice of the compression parameters, such as time constants, is still the subject of an ongoing debate. This contribution evaluates the efficacy of fast-acting DRC as a hearing-loss compensation strategy in a range of experimental conditions. First, fast-acting DRC was investigated considering temporal masking of narrowband stimuli. The results of a model-driven evaluation showed that the measures of temporal resolution can be improved with fast-acting compression with a very short release time (10 ms). Second, the effects of compression on speech audibility and noise-induced forward masking were evaluated in a highly-controlled scenario. The application of very short compression time constants was shown to improve HI listeners' consonant recognition performance. Finally, despite the benefits of fast-acting compression apparent in controlled conditions, it may introduce distortion in realistic scenarios, such as a reduction in the signal-to-noise ratio (SNR). A novel signal-to-noise-ratio-aware compensation strategy is discussed, which switches between fast- and slow-acting compression depending on the presence of the target signal and therefore preserves the natural relationship between the target and the background. An objective evaluation of the algorithm is presented and its potential applications are discussed.
Original language | English |
---|---|
Title of host publication | Proceedings of 2018 Joint Conference - Acoustics |
Number of pages | 9 |
Publisher | IEEE |
Publication date | 2018 |
Pages | 1-9 |
ISBN (Print) | 9781538671139 |
DOIs | |
Publication status | Published - 2018 |
Event | 2018 Joint Conference - Acoustics - Grand Lubicz Hotel, Ustka, Poland Duration: 11 Sep 2018 → 14 Sep 2018 |
Conference
Conference | 2018 Joint Conference - Acoustics |
---|---|
Location | Grand Lubicz Hotel |
Country | Poland |
City | Ustka |
Period | 11/09/2018 → 14/09/2018 |
Keywords
- Hearing aids
- Compression
- Speech intelligibility
- Psychoacoustics
- Audio signal processing