Sonic Enchant SPiN Management Technology
Noise reduction systems in hearing aids encompasses both directional microphones and noise reduction algorithms. Although they go hand in hand, they’ve been two separate uncoordinated strategies with lots of limitations.
Directionality by way of microphone control is well known. It can be grouped in two types, adaptive and fixed. Adaptive directional mode has polar patterns that continuously track and reduce dominant sounds from the sides and the back. Fixed mode can incorporate one of several types of polar plots, like cardioid, hypercardioid, supercardioid, etc. , and unlike adaptive mode, fixed mode does not change the pattern according to variations in the listening environment. Both types can process signals in separate frequency bands to further cancel noise sources simultaneously with different spectral characteristics from different directions.
But there are limitations to the success of directional microphones. The signal has to be close enough to the listener (6 ft or less), the noise has to be on the sides or the back, and the room has to have minimal reverb. Another limitation is that directional modes tend to cause a low frequency roll-off that would be compensated by the system through amplification at the low ends, resulting in higher internal microphone static noise levels. On top of that, it gets more complicated when you throw in the human factor. Some people prefer omnidirectional patterns even in noise for different reasons. Users may still want to hear subtle elements of their surrounding environment and feel robbed of it due to the nulls (angle of maximum attenuation) in the polar plots. Sometimes, the lower output level due to directionality may reduce the volume of all sounds too much, included the desired sound in front. Sometimes, the user can simply forget to switch to the right listening program that has directionality at the right time.
Many of these issues can be resolved with automatic directionality and a multiband design. Automatic directionality monitors the overall level and the temporal and spectral characteristics of the environment to transition in and out of directionality based on the information it receives. A multiband design can help preserve low frequencies with omnidirectionality in the lower bands (to solve the low frequency roll off issue mentioned earlier) while activating directionality in higher frequencies.
Whereas directional microphones provide measurable improvements in speech recognition in noise, the effect of noise reduction algorithms is more subtle. While it’s widely accepted that noise reduction can provide listening comfort, there’s not much evidence that shows it can improve speech recognition. But it’s still helpful and widely used because it can help reduce listening effort, improve SNR, improve acceptance of noise, improve speech recall, etc.
Sonic believes that if they coordinate the operation of directionality and noise reduction (NR), which has traditionally been uncoordinated, they can help improve the speech-in-noise listening experience much better. They also devised a 3-prong strategy that apply to both directionality and NR: 1) make the activation automatic, 2) make the system adaptive, and 3) apply a multiband design to the system.
-
Automatic activation
a. For directionality: when background noise starts, SPiN Directionality engages the directional mic, and returns to omni mode when the background noise stops.
b. For NR: for noise that has not been attenuated by SPiN Directionality, SPiN NR comes in and remove them. SPiN NR is a fast-acting modulation based digital NR algorithm that detects temporal characteristics of the sound, its modulation rate and depth. Signals with a high modulation rate are desirable (speech) where as signals with a low modulation rate are undesirable (noise). SPiN NR is automatically activated to preserve speech and reduce noise.
-
Adaptivity:
a. For directionality: SPiN Directionality constantly alter the internal time delay between the front and rear mics to adaptively change the polar patterns in response to spatially dynamic listening environments (i.e. noise sources that are moving relative to the listener). Based on the level and location of the noise source, SPiN Directionality uses null-steering to select the polar plot with the best SNR in each frequency band.
b. For NR: SPiN NR uses extremely fast time constants to efficiently respond to rapid fluctuations in noise and attenuate noise even between the smallest speech pauses and preserve speech down to the phonemic level.
-
Multiband design:
a. For directionality: SPiN has independent directional systems in 16 different frequency bands (compared to 4 bands in the past). This provides enough resolution for isolating and suppressing noise from different directions across the frequency spectrum with extreme accuracy. This multiband design also allows the low-frequency input to be preserved by staying in omni mode for low frequencies while permitting directional responses in the mid and high frequencies.
b. For NR: SPiN NR works in the same 16 frequency bands as SPiN Directionality to synergistically identify noise of varying spectral content and reduce gain in the narrow bands where noise is detected.
To decide when to activate SPiN Directionality and SPiN NR, Sonic added a third component to their SPiN Management system called SPiN Engage. It’s simply a mechanism to allow user control of when the onset activation of SPiN Directionality and SPiN NR will occur, based on the SNR level set forth by the user. It offers 3 help settings: High help will cause an immediate onset of SPiN Directionality and SPiN NR to rapidly put an emphasis on the speech signal. Medium help is associated with a higher SNR threshold onset to delay the activation of these 2 systems. Low help will raise the SNR threshold onset further to retain maximum auditory awareness of environmental noise.
The SPiN NR gives you 4 values: Off, Low (3dB to 4.5dB), Medium (4.5dB to 7.5dB) and High (6dB to 15dB) for (min NR to max NR respectively). But note that attenuation is only as much as needed. It is not always at the max attenuation.
SPiN Directionality freely adapts the microphones anywhere from omni to fixed directionality in all 16 frequency bands. It has 2 settings. High Performance setting allows the system to adjust to the narrowest directionality pattern available. Medium Performance limits the system to a wider directionality pattern.