You just said hybrid? You might be curious to know why hearing aids need a hybrid algorithm. First, a reminder that there are always different solutions to solve the same problem. Each solution has specific advantages, and it would be optimal to provide a solution that combines them. Consider, for example, a hybrid car engine which combines two different technologies in one solution.
Hybrid solutions for hearing aid algorithms
There are also different approaches when we develop hearing aid algorithms. Each solution provides advantages but they are not always compatible, e.g., provide more audibility, more comfort, or better sound quality. The traditional approach is to choose an algorithm and tune it until it provides a good trade-off between all these aspects. This strategy can provide good results, but it will be likely that you improve one aspect at the cost of another, like better audibility of soft speech versus over-amplification of soft noise. However, this strategy is obsolete if your processor allows for the combination of different solutions on a single platform. This is the case for our hearing aids and as a result Bernafon has introduced Hybrid Technology™ with Alpha.
Hybrid Sound Processing™ to drive amplification
Bernafon Alpha hearing aids offer Hybrid Sound Processing™ which is a new amplification system based on wide dynamic range compression (WDRC). WDRC is one of the core functionalities embedded in a hearing aid. It is designed to compensate for the reduced residual dynamic range of hearing-impaired people by applying amplification as a function of the input level.
The level estimator is one of the key elements as it drives the compression. In other words, the estimated level is used to decide how much amplification will be applied in the time and frequency domains. While the amount of amplification is computed according to the selected fitting rationale, different strategies can be applied to design and implement the level estimator which is sensitive to a) dynamic signals like speech or b) noises with some specific frequencies.
The debate about compression speed and amount of compression channels
One hot topic in our industry is the design of dynamic compression. Many published articles evaluate different implementations of compression by changing the compression speed and / or the amount of compression channels. It is quite complex to understand which solution is the most appropriate as the results mainly depend on the evaluation approach, the test conditions, and the listener’s auditory capacities.
Some solutions are good to follow variations over time, such as phonemes in a speech signal, and provide a good resolution in the time domain. They will provide good audibility for the softest phoneme which is an important requirement for a hearing aid user. However, when a noise signal with some specific frequencies is added to the speech signal, then a compression with more resolution in the frequency domain might provide more comfort.
Hybrid Sound Processing™ – a unique solution
Hybrid Sound Processing™ was developed to provide the advantage of two compression strategies as a function of the input signal in a single solution:
- in the time domain, a phonemic compressor with a fast, broadband level estimator will ensure audibility for any modulated signal like speech;
- in the frequency domain, signals with lower modulation like noise are analyzed with 24 slow level estimators to provide better sound quality.
This unique solution that provides wide dynamic range compression with improved time and frequency resolution is one foundation of our Hybrid Technology™.
This unique solution that provides wide dynamic range compression with improved time and frequency resolution is one foundation of our Hybrid Technology™
You will find more information in our White Paper if you are interested to learn more about the story and the concepts beyond Hybrid Sound Processing™.