Non-Linear Frequency Compression (NFC), or SoundRecover by Phonak, is a form of frequency lowering applied in hearing instruments. Developed in response to an age-old perfect storm that exists in hearing instrument fittings, NFC is intended to improved access to high-frequency information. The bandwidth of modern hearing aids may be insufficient for a number of reasons: hearing loss tends to be the worst in the high-frequency region; ear-level hearing instruments have technological limits when it comes to high-frequency output; feedback can be a limiting factor with increasing high-frequency gain; and high-frequency consonants, critical for speech understanding, are also some of the softest parts of speech. SoundRecover has been successful in addressing these compounded challenges in high-frequency amplification, and has been applied in over a million hearing instrument fittings. Many studies have been conducted to further enhance our understanding of the benefits and limitations of this technology. This paper aims to summarize the findings of these studies and provide an evidence basis and clinical guidelines on the applications and benefits of SoundRecover.
Does NFC Work for Adults?
Yes, for many individuals NFC will improve the detection of high-frequency speech and environmental sounds without creating unpleasant sound quality and consonant confusions. Often, benefit has been documented in studies when the clinician first met broadband targets for audibility and then fine-tuned NFC parameters to ensure audibility goals were met. These adjustments tended to involve a strengthening (decreasing the cut-off frequency) of NFC.
Simpson (2005) applied a prototype of NFC on adults with moderate to severe sloping hearing losses. At the group level, NFC significantly improved detection of affricate and fricative consonants. In a follow-up study with steeply sloping loses (Simpson, 2006), benefit was not seen. However, this could have been due to the fact that prototypes were used, subjects did not have previous hearing aid experience, or the settings were not optimized to provide benefit for these losses.
At the University of Western Ontario, Danielle Glista and colleagues (2009) evaluated NFC with adults with moderate to profound sloping hearing losses. Five out of 13 subjects obtained significant benefit from NFC.
Rachel Ellis, as part of her doctoral thesis at University of Manchester, (2012) found that 9 out of 11 adults with bilateral sensorineural hearing loss (SNHL) demonstrated a significant improvement on speech in quiet and noise measures with NFC.
Andrew John and colleagues (2013) demonstrated a significant effect of NFC with 28 adult subjects with asymmetric hearing loss. Gains were seen on high-frequency speech tests and self-reported gains in sound quality and speech understanding in quiet. Benefit was not seen in noise, however in a previous study with children (Wolfe, et al., 2011) Dr. Wolfe found that improvement in noise only became significant after six months of use.
Should I Apply NFC for Mild Hearing Losses?
Due to bandwidth limitations that exist for all commercially available ear-level hearing instruments, there is evidence that NFC can aid in the detection of very high-frequency environmental sounds and phonemes, in particular the /s/ phoneme for female talkers which peaks around 9kHz. As the third most common letter and one of the most important linguistic markers in the English language denoting plurality, possession and sometimes tense, audibility for /s/ is required for optimal understanding. In the case of mild hearing loss, the NFC parameters are set very lightly with no compression applied below about 5kHz. This is intended to provide the additional high-frequency access with minimal impact on signal quality.
Wolfe et al. (submitted, JAAA) studied the benefit of NFC for mild hearing loss in a group of 11 children. Data showed significant improvement in the detection of an 8kHz warble tone and the female production of /s/ when compared to performance with the same device with NFC off. It is important to recognize that most clinical tools, including speech material and electro-acoustic measurements, will not be sensitive to changes in audibility at 9kHz. In this study, the Phonak Phoneme Perception Test (PPT) and aided threshold testing using warble tones were the measures used that were sensitive to the benefits obtained with NFC.
Does NFC Improve Listening in Noise?
Some people benefit in noise with NFC; others do not. To better understand what those patient variables are, we need further research. Several studies have reported that, on average, subjects’ ability to understand in noise improved with SoundRecover compared to conventional amplification.
Bohnert (2010) demonstrated that 7 of 11 subjects with severe to profound SNHL performed better in noise with SoundRecover than with conventional amplification.
Wolfe et al. (2011) showed that children with up to moderately-severe SNHL improved on the Bamford Kowal Bench speech-in-noise (BKB-SIN) test after six months of NFC use.
How long will it take for a Patient to Receive Maximum Benefit from NFC?
Most subjects show some initial benefit at the time of fitting, but there is evidence with children that performance can continue to improve over at least six months’ time after initial NFC fittings.
Data published by Wolfe and colleagues (2011) showed that a group of 15 children with moderate to moderately-severe hearing loss continued to improve on recognition of high-frequency phonemes and listening in noise at 6 months of NFC use compared to data collected after 6 weeks of use. This study shows there is some acclimatization effect with NFC.
Ellis (2012) also showed limited evidence of some acclimatization period with adults using NFC.
What is the Best Way to Assess NFC?
After you have met targets across as broad a bandwidth as possible, turn NFC on. The weakest settings that result in benefit should be applied. Excessive compression should be avoided. There are behavioral and electroacoustic tests available to optimize and validate the strength of the NFC settings. Best practices would dictate that clinicians optimize and verify SoundRecover electroacoustically using speech stimuli available in newer real-ear test systems. Then, if possible, behavioral measures should be used to validate the settings with improved performance.
Behavioral Speech Tests. Since SoundRecover is designed to only impact the highest-frequency speech sounds, standard speech tests are generally not sensitive to its effects. The UWO Plurals and Ling-6 sound test package was developed by Susan Scollie and Danielle Glista at the University of Western Ontario (Scollie, Glista, Tenhaaf, Dunn, Malandrino, et al., 2012; Glista & Scollie, 2012). The Phoneme Perception Test (PPT) was developed by researchers at Phonak. The UWO Ling-6 test, Plurals test and the PPT have been shown in multiple published peer-reviewed studies to be sensitive to SoundRecover. These tests are available from Phonak and can be used to optimize SoundRecover parameters and document that amplification goals have been met.
The Plurals test is only appropriate for English speakers and can be used with children down to four years of age. With these speech-based tests, the clinician seeks to improve audibility of high-frequency sounds like /s/ without creating confusions, for example, between /s/ and /sh/. The PPT is language independent and allows assessment of high-frequency detection, distinction and recognition. In the closed-set recognition test, the patient’s high-frequency consonant recognition is assessed. The distinction test allows clinicians to assess confusions between sounds like /s/ and /sh/.
Measuring aided thresholds in the booth can also be useful to assess a patient’s access to high frequencies. When doing such functional gain testing with modern hearing instruments, it is recommended to use warbled tones and an ascending method of establishing threshold. The PPT is available in Phonak Target Software 3.2
The UWO Plurals test and calibrated Ling-6 Sound test package is available from Phonak. Please reference part number 041-1418 when ordering.
Electroacoustic Tests. The speech band stimuli available in the AudioScan Verifit and the calibrated Ling-6 stimuli available in the GN Otometrics Aurical equipment can be very useful in determining the optimal SoundRecover settings. The live speech mode can also help assess audibility and overlap of the /s/ and /sh/ sounds. Settings should to be adjusted to maximize audibility for /s/ without causing excessive overlap (and hence confusion) with /sh/.
For a complete guide to verification of NFC with the AudioScan Verifit, download the following guidelines developed by Glista and Scollie (2009) at UWO: http://www.phonakpro.com/com/b2b/en/evidence/topics/soundrecover.html
Should I Apply NFC for Musicians and Patients who Listen to Music?
There is more to learn about the relationships between hearing loss, NFC, and music appreciation. Anecdotally, there are some musicians and music listeners who report benefit. The best approach for music aficionados may be to offer them programs with and without SoundRecover for evaluation with their own musical instrument or musical style.
Parsa et al. (2013) studied the impact of NFL on sound quality ratings of speech and music. No impact was found, indicating that a range of NFC parameters exists that will not adversely affect sound quality for hearing-impaired listeners.
In a blinded study of 40 hearing instrument users conducted by Miranda Uys (2012) in South Africa, greater appreciation and enjoyment of music was reported with SoundRecover than with conventional signal processing. Participants were adults with moderate to severe SNHL. The several musical qualities were compared between conditions and are shown in Figure 1.
Figure 1. Musical qualities compared with NFC and traditional processing.
How does Target Software Apply SoundRecover?
SoundRecover is enabled by default for most fittings and can be manually enabled or disabled in any program or fitting. For pediatric fittings using the DSLv5 formula, SoundRecover is activated according to the recommendations of researchers at University of Western Ontario (UWO). By default, it is activated for all hearing losses with a high frequency pure-tone average (2000, 3000 and 4000 Hz) of greater than 20dB.
For all other fittings, the SoundRecover rationale was developed with input from Professor Hugh McDermott who developed the technology. It is activated by default when the hearing loss at 8000 Hz is over 25dB and the difference between the hearing loss at 3000 Hz and 8000 Hz is more than 30dB.
Does Cognitive Ability Impact Benefit of NFC?
The answer to this is maybe not. It is believed that cognitive function, specifically, working memory span, is strongly related to overall hearing aid benefit. Ellis and Munro (2013) investigated the role of cognitive ability in the processing frequency compressed speech. Their results showed that there was no correlation between cognitive ability and speech recognition ability in any of several NFC conditions. In a different study, Arehart and colleagues (2013) showed that speech understanding in noise for older listeners with hearing loss and poor working memory was more impacted by distortions (including another form of frequency compression, not Phonak SoundRecover).
Why Apply SoundRecover for Children?
In the last ten years, a growing body of research has become prominent in the literature documenting that frequencies beyond those available through traditional hearing aids were valuable for children developing speech and language (Stelmachowicz, 2004: Pittman, 2008). The acknowledgment of the importance of frequencies above 4000 Hz, led Phonak to develop and implement SoundRecover. Multiple studies have documented the benefit of SoundRecover for children. Compared to adults, children typically show more benefit (Glista, et al., 2009) and adapt quickly.
Wolfe (2009; 2010; 2011; submitted, JAAA) and Glista (2009) documented speech perception benefit with SoundRecover for children. Additionally, in his evidenced-based systematic review of frequency lowering for school-aged children with hearing loss, McCreery (2012) determined that there was moderate evidence to support SoundRecover for sloping losses.
How Should SoundRecover be Applied when there are Expected Dead Regions?
The benefit of NFC has been shown in patients with and without dead regions. A dead region is certainly not a criterion for exclusion and may be a predictor of benefit. More work is needed in this area. Ellis (2012) showed that the presence of a dead region did not predict benefit (or lack thereof) for NFC.
How Should I fit NFC when there is an Asymmetric Hearing Loss?
The best method of fitting NFC in the case of asymmetric hearing loss is unknown. At present, the parameters of NFC in both ears is dictated by the hearing loss in the better ear, and hence are applied symmetrically. It is plausible that additional benefit would be gained by ear-specific optimization of SoundRecover.
Wolfe et al. (2013) investigated the benefit of NFC with 28 adults who had sloping moderately-severe to severe SNHL that was at least 15dB different between ears across at least 3 high-frequency test frequencies (2000 -6000 Hz). In this study, patients performed better at detecting /f/ and 6000 Hz /s/ and using the PPT. Scores on the Speech, Spatial, and Qualities of Hearing scale (SSQ) demonstrated significant improvement on the speech and qualities subscales. Overall, SSQ scores were better with NFC. No significant differences were seen on the spatial subscale. There was a trend toward better understanding on the BKB-SIN with NFC enabled, but the effect was not significant.
In this experiment, there was no difference in performance between the default settings (NFC settings based on the better ear) and a condition where NFC parameters were applied for each ear independently. However, it is notable that NFC settings were not optimized using real-ear measurements. Particularly with the steeply sloping nature of hearing loses in this experiment, small alterations in NFC settings can result in large improvements in audibility. Ear-specific real-ear measures to customize NFC settings could have led to different outcomes.
What are the Potential Disadvantages of NFC?
Consonant confusions may result from aggressive NFC settings. The Target software default kneepoint for applying NFC will never be set below 1900 Hz. It can be manually lowered down to 1500 Hz. These settings should be used with extreme caution because they are likely to have a detrimental impact on sound quality. At these aggressive lowering settings, not only will consonants be highly compressed, but high-frequency information will be placed in the spectrum of the second formant of most vowels. This can lead to significant distortion for many speech sounds.
Who is not an Appropriate Candidate for NFC?
NFC should be used with caution for patients who have a ski slope and corner audiograms requiring extremely aggressive NFC settings, defined as compression thresholds below 2000 Hz. These settings could introduce distortion and confusions for some users. NFC may also not be appropriate for reverse sloping audiograms where hearing thresholds are normal in the high frequencies.
Should NFC be used as a Feedback Management System?
NFC should be applied to increase audibility and intelligibility of the signal. There are suggested clinical protocols mentioned above (see What is the Best way to Assess NFC) for how a clinician can confirm that these goals have been met. It is convenient that, in this case, increasing high-frequency audibility can reduce the risk of feedback. However, it is not recommended to apply NFC purely as a feedback manager, as this could introduce distortion without benefit. Evaluation of the physical fit or use of the feedback manager is the recommended course for addressing feedback.
The feedback test in Target software allows the system to act in the most precise way to address feedback when and where it is produced, limiting any negative impact. Phonak WhistleBlock is an active system that will only apply tagging as needed on a continuously measured feedback pathway, thus maximizing signal quality.
Does applying NFC for Children Disrupt the Development of Normal Tonotopic Frequency Mapping in the Auditory Cortex?
By shifting the highest frequency sounds down, NFC provides improved access to high-frequency information where there was a documented shortcoming in modern hearing instruments across all manufacturers. It is recommended that clinicians avoid overly-aggressive NFC settings and select the mildest setting possible that delivers benefit.
Stelmachowicz el al. (2004) demonstrated that the bandwidth of hearing aids inadequately represented the high-frequency sounds of speech, particularly for female talkers. As with cochlear implants, some technologies change the place of stimulation to meet the goals of achieving better auditory access, discrimination, audibility and overall speech and language outcomes. NFC does result in the encoding of higher-frequency information at lower frequencies. However, the documented benefits include improved access to high-frequency environmental and speech sounds, particularly important for female and child talkers who make up a majority of speech stimuli for small children.
Would it be better to use a Hearing Instrument with Extended Bandwidth rather than NFC?
The bandwidth of a hearing instrument may be reported two ways. One is the ANSI standard bandwidth measurement. The other is the sampling rate of the analog-to-digital (A/D) converter. In digital systems, the highest frequency that can be accurately coded is one-half of the sampling rate, so a device with a 20,000 Hz sampling rate would have an upper frequency limit of 10,000 Hz. This is referred to as the Nyquist frequency. Both the ANSI bandwidth and the Nyquist frequency have been reported by hearing instrument manufacturers to promote the high-frequency capabilities of their instruments. Neither of these values, however, represents the perceptual bandwidth of the hearing instrument, or the highest frequency a patient may hear with the device.
It is not uncommon, given two devices with different reported bandwidths, for a patient to obtain more audible high-frequency amplification from the device reporting a narrower bandwidth. Careful assessment using electroacoustic measures or sensitive speech materials is needed to identify which device is more beneficial for the patient. For a more comprehensive overview of these measurement tools, refer to the section, What is the Best Way to Assess SoundRecover?
Wolfe et al. (submitted, JAAA) compared performance with a reported “extended bandwidth pediatric hearing instrument” with the Phonak Nios S with and without NFC. Patients obtained more functional high-frequency amplification with the Nios with NFC off and on than with the reported “extended bandwidth” device. The most benefit was obtained with NFC on.
For more information about acoustic versus perceptual bandwidth, please see the detailed overview by Prof. Hugh McDermott: http://www.phonakpro.com/com/b2b/en/evidence/topics/soundrecover.html
Arehart, K. H., Souza, P., Baca, R., & Kates, J.M. (2013.) Working memory, age, and hearing loss: susceptibility to hearing aid distortion. Ear and Hearing, 34(3), 251-260. doi: 10.1097/AUD.0b013e318271aa5e.
Bohnert, A., Nyffeler, M., & Keilmann, A. (2010). Advantages of a non-linear frequency compression algorithm in noise. European Archives of Otorhinolaryngology, 267(7), 1045-1053. doi: 10.1007/s00405-009-1170-x.
Ellis, R. J. (2012). Benefit and predictors of outcome from frequency compression hearing aid use. A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy. Faculty of Medical and Human Sciences.
Ellis, R. J. & Munro, K. (2013). Does cognitive function predict frequency compressed speech recognition in listeners with normal hearing and normal cognition? International Journal of Audiology, 52(1), 14-22.
Glista, D., & Scollie, S. (2009b). Pediatric verification considerations for instruments with SoundRecover (non-linear frequency compression) using the latest Audioscan Verifit® tests. National Centre for Audiology, University of Western Ontario, Canada. Retrieved from http://www.phonakpro.com/com/b2b/en/evidence/topics/soundrecover.html
Glista, D., & Scollie S. (2012). Development and Evaluation of an English Language Measure of Detection of Word-Final Plurality Markers: The University of Western Ontario Plural Test. American Journal of Audiology, 21(1), 76-81. doi: 10.1044/1059-0889(2012/11-0036).
Glista, D., Scollie, S., Bagatto, M., Seewald, R., Parsa, V., & Johnson, A. (2009). Evaluation of nonlinear frequency compression: Clinical outcomes. International Journal of Audiology, 48(1), 632–644. doi: 10.1080/14992020902971349.
John, A., Wolfe, J., Schafer, E., Hudson, M., Fox, K., Wheeler, J., et al. (2013). In asymmetric high frequency hearing loss, NLFC helps. The Hearing Journal, 66(9), 26-29.
McCreery R., Venediktov, R., Coleman, J., & Leech, H. (2012). An evidence-based systematic review of frequency lowering in hearing aids for school-age children with hearing loss. American Journal of Audiology, 21(2), 313-328. doi: 10.1044/1059-0889(2012/12-0015).
McDermott, H. (2010). SoundRecover – The importance of wide perceptual bandwidth. Background Story. Phonak AG: 2010. Retrieved from http://www.phonakpro.com/com/b2b/en/evidence/topics/soundrecover.html
Parsa, V., Scollie, S., Glista, D., & Seelisch, A. (2013). Nonlinear frequency compression: effects on sound quality ratings of speech and music. Trends in Amplification, 17(1), 54-68. doi: 10.1177/1084713813480856.
Pittman, A. L. (2008). Short term word-learning rate in children with normal hearing and children with hearing loss in limited and extended high frequency bandwidths. Journal of Speech Language and Hearing Research, 51(3), 785-797. doi: 10.1044/1092-4388(2008/056).
Scollie, S., Glista, D., Tenhaaf, J., Dunn, A., Malandrino, A., Keene, K., et al. (2012). Stimuli and normative data for detection of Ling-6 sounds in hearing level. American Journal of Audiology, 21(2), 232-241. doi:10.1044/1059-0889(2012/12-0020).
Simpson, A., Hersbach, A., & McDermott H. (2005). Improvements in speech perception with an experimental non-linear frequency compression hearing device. International Journal of Audiology, 44(5), 281-292.
Simpson, A., Hersbach, A., & McDermott, H. (2006). Frequency-compression outcomes in listeners with steeply sloping audiograms. International Journal of Audiology, 45(11), 619-629.
Stelmachowicz, P. G., Pittman, A. L., Hoover, B. M., Lewis, D. E., & Moeller, M. P. (2004). The importance of high frequency audibility in the development of speech and language for children with hearing loss. Archives of Otolaryngology- Head and Neck Surgery, 30(5), 556-562.
Uys, M., Pottas, L., Vinck, B., & van Dijk, C. (2012). Iinfluence of non-linear frequency compression on the perception of music by adults with a moderate to severe hearing loss: subjective impressions. The South African Journal of Communication Disorders, 59, 53-67.
Wolfe, J., John, A., Schafer, E., Nyffeler, M., Boretzki, M., & Caraway, T. (2010). Evaluation of nonlinear frequency compression for school-age children with moderate to moderately severe hearing loss. Journal of the American Academy of Audiology, 21(10), 618-628. doi: 10.3766/jaaa.21.10.2.
Wolfe, J., John, A., Schafer, E., Nyffeler, M., Boretzki, M., & Caraway, T., et al. (2011). Long-term effects of non-linear frequency compression for children with moderate hearing loss. International Journal of Audiology, 50(6), 396–404. doi: 10.3109/14992027.2010.551788.
Wolfe, J., et al. (Submitted). Journal of the American Academy of Audiology.
Cite this content as:
Jones, C. (2014, March). Sound bytes on SoundRecover. AudiologyOnline, Article 12552. Retrieved from: http://www.audiologyonline.com