Chapter 4. Stage Two: Turning the Hearing Inward

“Turn the hearing back to hear your own nature.” (Mañjuśrī’s verse)

The first stage stopped a comparison. The second stage changes where attention rests while the voice is sounding: from the work of producing it to the act of hearing it.

Two ways to be busy while singing

A singer can be busy in two quite different ways during a note.

Driving the output. Attention is on the mechanism: the throat, the breath, the jaw, the “placement”, the effort to hold the note up. The singer is steering muscles.
Receiving the input. Attention is on the result as it arrives: the pitch, the colour, the buzz of the sound in the face and head. The singer is listening.

Both are active. The difference is the direction. In the first, attention runs from the singer to the voice; in the second, it runs from the voice to the singer. The sutra’s phrase “turn the hearing back” is, in this book’s reading, a description of the second.

What the science says about direction of attention

Evidence

Chapter 2 described the brain’s two kinds of voice control: a learned command that drives the voice forward (feedforward), and an auditory and somatosensory comparison that corrects it (feedback) (Tourville and Guenther 2011). Feedback control needs sensory input to work with. The pitch-shift reflex shows that the auditory comparison runs automatically and fast, without the speaker deciding to correct anything (Burnett et al. 1998).

Evidence

Research on motor learning offers an outside parallel. Gabriele Wulf’s review of fifteen years of studies found that directing attention to the effect of a movement (an external focus) generally led to better performance and learning than directing it to the body’s movements themselves (an internal focus), across tasks from balance to golf to throwing (Wulf 2013). Most of these studies did not involve singing, and their results cannot simply be transferred to the voice.

The parallel is suggestive rather than direct. Listening to the sound of the voice is close to Wulf’s “effect” focus; managing the larynx and breath is close to her “movement” focus. The second stage asks the singer to rest attention on the effect.

Evidence

Attention also changes the ear itself. Sandrine Maison and colleagues found that focusing attention on a sound changes cochlear activity (Maison et al. 2001), through the efferent nerves that let the brain adjust the inner ear’s gain (Guinan 2006). Where attention rests is not only a mental matter; it reaches the first stage of hearing.

“Standing-wave points” in the head: what is really there

Singing teachers often speak of feeling the voice “in the mask”, at the bridge of the nose, the cheekbones or the crown, and some describe these as places where the sound resonates or forms standing waves. The second stage uses these sensations, so it is worth being exact about what they are.

Evidence

The resonances that shape the voice are the resonances of the vocal tract: the air column from the vocal folds to the lips (Chapter 6). Standing waves form in that air column. The skull is not a resonator for the voice in the same way: at the fundamental frequencies of singing, the head vibrates largely as one body, and more complex vibration of the skull begins only at higher frequencies (Stenfelt and Goode 2005). Johan Sundberg’s account of the singing voice concludes that the nasal sinuses contribute little to the sound that leaves the mouth (Sundberg 1987). What singers feel in the face and head is vibration carried through bone and tissue from the larynx and the vocal tract.

Note

So the “standing-wave points” are better called vibration points. They are real and repeatable: the same vowel on the same note produces the same pattern of buzzing, and the pattern changes when the vowel, the pitch or the effort changes. That makes them useful as an object of attention, which is all the second stage needs. They are a signal to listen to, not a place to push the sound into.

The exercises of Stage Two

Hypothesis · not established · Stage Two: turning the hearing inward (the author’s method)

The aim of the second stage is to make listening, rather than steering, the default occupation of attention during sound. The author’s hypothesis is that when attention rests on input, the brain’s own comparison of intended and heard pitch runs with less interference, and pitch becomes steadier without being forced.

1. Entering the stream (2 minutes, silent). Sit comfortably with eyes closed. Listen to whatever sounds are in the room and outside it, without naming them. After a minute, notice that there are gaps between sounds, and that listening does not stop in the gaps. Rest attention on the listening itself rather than on the next sound. This is the sutra’s “entering the stream and forgetting its objects”, used as an exercise of attention.

2. Sustain and shift (3 minutes). At the working level found in Stage One, sustain a hum or the vowel /u/ on a comfortable note for about eight seconds. For the first three seconds, attend to doing it. Then shift attention to receiving it: the pitch as it arrives, and the places in the face and head where vibration is felt. Do not change anything on purpose. Notice whether the voice changes by itself when attention shifts. Repeat on three or four notes.

3. The listening glide (2 minutes). On “ng” (as at the end of “sing”), glide slowly up a fifth and back down, like a quiet siren. Follow the glide with the ear, the way one follows a moving sound in a room, rather than steering it with the throat. If the glide catches or jumps, slow it down.

4. Making and listening (3 minutes). Sing a short phrase twice. The first time, make it: attend to the mechanics. The second time, listen to it: attend to the sound and the vibration as they arrive. Record both and compare their pitch with a tuner app or with the program below. The comparison is information, not a test to pass.

Calculation

Measuring steadiness from a recording. A simple way to estimate the pitch of a sustained note is autocorrelation: a periodic signal resembles itself shifted by one period, so the lag at which a frame best matches itself gives the period. The program below makes a two-second test tone at 220 Hz that slowly drifts 20 cents flat, tracks its pitch in 50 ms steps, and reports the median error and its spread in cents. To use it on your own voice, replace the test tone with a mono WAV file read by scipy.io.wavfile.read.

import numpy as np

sr = 16000
t = np.arange(0, 2.0, 1 / sr)
# a test "voice": 220 Hz, slowly drifting 20 cents flat, with 3 harmonics
drift = 220 * 2 ** (-20 / 1200 * t / t[-1])
phase = 2 * np.pi * np.cumsum(drift) / sr
x = sum(np.sin(k * phase) / k for k in (1, 2, 3))

def f0_autocorr(frame, sr, fmin=80, fmax=1000):
    frame = frame - frame.mean()
    ac = np.correlate(frame, frame, mode="full")[len(frame) - 1:]
    lo, hi = int(sr / fmax), int(sr / fmin)
    lag = lo + np.argmax(ac[lo:hi])
    # parabolic interpolation around the peak for sub-sample accuracy
    a, b, c = ac[lag - 1], ac[lag], ac[lag + 1]
    lag += 0.5 * (a - c) / (a - 2 * b + c)
    return sr / lag

hop, win = 800, 1600                  # 50 ms steps, 100 ms windows
f0 = np.array([f0_autocorr(x[i:i + win], sr) for i in range(0, len(x) - win, hop)])
cents = 1200 * np.log2(f0 / 220)
print(f"frames: {len(f0)}, median {np.median(cents):+.1f} cents, "
      f"spread (SD) {cents.std():.1f} cents")
# frames: 38, median -9.5 cents, spread (SD) 5.4 cents

A drift from 0 to $-20$ cents has a median near $-10$ and a standard deviation near $20/\sqrt{12} \approx 5.8$ cents, which is what the tracker finds. Real voices have vibrato and onsets; for those, analyse only the steady middle of each note, as Chapter 8 does.

What the second stage does not claim

It does not claim that singers should never think about technique. Technique is learned by attending to the mechanism, often slowly and deliberately. The claim is about performance: that once a skill is learned, attention spent steering it during the note is more likely to disturb it than to help it, and that listening is a better resting place for attention than effort.

Nor does it claim that the vibration felt in the head is the voice the audience hears. It is one more signal from inside. The difference from Stage One is that the singer no longer judges it against an outer model; they simply receive it.

Summary

During a note, attention can drive the output (steering muscles) or receive the input (listening to sound and vibration). Stage Two moves it from the first to the second.
The brain corrects pitch automatically when it has sensory input to work with; in motor learning generally, attention on the effect of a movement tends to work better than attention on the movement.
What singers feel as resonance in the head is vibration conducted from the larynx and vocal tract; the standing waves that shape the voice are in the vocal tract. The sensations are still a useful anchor for attention.
The exercises shift attention from making to listening, and the program shows how to measure whether pitch becomes steadier.
That this shift steadies pitch is a hypothesis.

References

Burnett, Theresa A., Marcia B. Freedland, Charles R. Larson, and Timothy C. Hain. 1998. “Voice F0 Responses to Manipulations in Pitch Feedback.” Journal of the Acoustical Society of America 103 (6): 3153–61.

Guinan, Jr., John J. 2006. “Olivocochlear Efferents: Anatomy, Physiology, Function, and the Measurement of Efferent Effects in Humans.” Ear and Hearing 27 (6): 589–607.

Maison, Sandrine, Christophe Micheyl, and Lionel Collet. 2001. “Influence of Focused Auditory Attention on Cochlear Activity in Humans.” Psychophysiology 38 (1): 35–40.

Stenfelt, Stefan, and Richard L. Goode. 2005. “Bone-Conducted Sound: Physiological and Clinical Aspects.” Otology & Neurotology 26 (6): 1245–61.

Sundberg, Johan. 1987. The Science of the Singing Voice. Northern Illinois University Press.

Tourville, Jason A., and Frank H. Guenther. 2011. “The DIVA Model: A Neural Theory of Speech Acquisition and Production.” Language and Cognitive Processes 26 (7): 952–81.

Wulf, Gabriele. 2013. “Attentional Focus and Motor Learning: A Review of 15 Years.” International Review of Sport and Exercise Psychology 6 (1): 77–104.

Fan, Y. (2026). Turning the Hearing Inward: A Bioacoustic Study of the Ergen Yuantong Listening Method [反闻闻自性] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23666557 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).