Niels Henrik Pontoppidan

Research Area Manager/ Research Engineer

> Learn more

Hearing-impaired people have reduced sensitivity to temporal fine structure. In complex listening situations, normal-hearing listeners benefit from TFS; however, hearing-impaired listeners not only suffer from an inability to hear the weak parts of speech signals, they also fail to benefit from TFS. Research should lead to a better understanding of the TFS dimension of hearing loss.

Temporal fine structure in speech signals

A sine wave is equivalently characterised by its frequency F (measured in Hz) or periodicity T (measured in seconds). Key assumptions in the TFS hypothesis are:

a. Accurate (relative compared to auditory bandwidths) decoding of a sine wave in the auditory system is based on sensing the periodicity in the signal – the TFS


b. Voiced parts of speech signals can be described as a sum of TFS signals with slowly varying temporal envelopes.

The role of TFS in speech recognition

Experiments show that TFS is not important in easy listening situations such as a single talker in quiet1. But it does become important when the difficulty increases with a competing talker2, where TFS is believed to contribute to talker identification and source segregation. Moreover, the benefit that hearing-impaired listeners obtain from TFS in difficult situations is significantly smaller than for normal-hearing listeners2.

TFS also plays a role in spatial hearing. We have found correlations between binaural TFS measurements and spatial unmasking, and a significant benefit from TFS in spatial unmasking. That TFS only seems important in difficult situations underpins our belief that TFS contributes to the auditory system’s ability to resolve ambiguities and utilise the redundancy of speech. Thus the robustness of human speech recognition also lies in the fine details.

Is fine structure temporal or spectral?

The role of TFS in speech recognition has been debated. The increased auditory filter bandwidths observed with hearing impairment could be responsible for the reduced ability of hearing-impaired listeners' to utilise fine structure cues. This debate is similar to the long-lasting ‘Place versus timing pitch’ debate3.

TFS in our current research

Hearing-impaired listeners report that a boost in audibility works reasonably well for single speakers where the listening situation is easy, and the problem is merely a matter of limited audibility. However, this does not restore people’s ability to recognise speech in difficult situations, with competing talkers, for instance.

We believe that future audiology should compensate for audibility and TFS deficits. In pursuit of this goal, we are continuing our research on quantifying the TFS benefits in increasingly realistic situations. We are also expanding our knowledge of how hearing-impairment affects sensitivity to TFS.

Further reading

1Shannon R, Zeng FG, Kamath V, Wygonski J, Ekelid M (1995). Speech recognition with primarily temporal cues. Science, New Series 270(5234), pp. 303-304.

2Hopkins K, Moore BCJ, Stone MA (2008) Effects of moderate cochlear hearing loss on the ability to benefit from temporal fine structure information in speech. Journal of the Acoustical Society of America, 123(2), p. 1140-1153.

3Santurette S (2011). Neural coding and perception of pitch in the normal and impaired auditory system. PhD dissertation, Department of Electrical Engineering, Technical University of Denmark.

Lunner T, Hietkamp RK, Andersen MR, Hopkins K, Moore BCJ (2012). Effect of Speech Material on the Benefit of Temporal Fine Structure Information in Speech for Normal-Hearing and Hearing-Impaired Participants. Ear and Hearing, 33(3), p. 377-388.

Learn more

  • learnmore-poster


    Sensitivity to Low-Frequency Temporal Fine Structure

  • learnmore-poster


    Effect of speech material on the benefit of temporal fine structure