‘Sound less nasal’ is not a complete voice instruction

That sentence contains enough ambiguity to keep a voice coach busy for an afternoon.

What does “nasal” mean here? Normal nasal resonance on nasal sounds? A stylistic timbre? A dialect feature? A squeezed voice that the listener happens to describe as nasal? Actual hypernasality? A microphone position that exaggerates part of the spectrum?

Those are not the same problem.

The useful non-clinical skill is narrower: explore an easier, clearer production without treating nasal resonance as a defect that must be removed.

Your nose is supposed to participate in speech

Speech resonance comes from sound produced at the vocal folds travelling through the vocal tract. The shape of the pharyngeal, oral and nasal cavities changes the resulting spectrum.

ASHA’s Resonance Disorders guidance states that normal speech uses an oral/nasal balance appropriate to the intended sound, and that resonance varies across sounds, languages and dialects. It also distinguishes resonance disorders from voice disorders and from ordinary differences in speech. ASHA: Resonance Disorders.

That immediately kills the bad global target:

nasal = bad

Nasal consonants need nasal resonance. Languages and dialects differ. A consumer speaking app should not treat one spectral pattern as the ideal human voice.

Resonance, phonation and effort get mixed together in everyday language

People use “nasal,” “throaty,” “blocked,” “thin,” “tight,” “chesty” and “forward” as perceptual descriptions. They are useful clues, not diagnoses.

A speaker may describe a voice as nasal when the real issue they care about is effort. Another may care about clarity. Another may simply dislike the timbre on a recording.

So before changing anything, ask:

What listener or speaker problem are we trying to solve?

Possible functional targets include:

• reduce a squeezed/effortful feeling;
• keep speech clear while using less effort;
• explore a different timbre for a specific speaking context;
• distinguish ordinary resonance from a microphone/capture change.

Those can be tested without pretending to diagnose anatomy.

The safe contrast is ease, not ‘denasalization’

Use a short phrase:

We need to review the final version today.

Record Take A normally.

For Take B, do not try to block nasal vibration or force the sound into a body location. Instead, make one modest change toward easier phonation: reduce unnecessary squeeze, keep the jaw/face comfortable, and let the words remain clear.

Now ask:

• Did effort decrease?
• Did clarity stay intact?
• Did the voice still sound like me?
• Did I chase a sensation instead of a communication result?

If Take B merely sounds different but feels no easier and communicates no better, “different resonance” is not automatically an improvement.

What about humming and straw exercises?

Semi-occluded vocal tract exercises—humming, lip trills, straw phonation and related tasks—are common in voice work. They can change acoustic and aerodynamic conditions during phonation.

The evidence is real, but the internet version is often too simple.

A 2026 scoping review of straw-based SOVT exercises included 107 studies and emphasized that exercise parameters such as straw diameter and water depth materially change the mechanical load; the review also notes distinct mechanisms for straw phonation in air versus water-resistance therapy. Comparing Straw-Based Semi-Occluded Vocal Tract Exercises.

Research in healthy speakers is less complete than the popularity of these exercises might suggest. A scoping review of vocal training in healthy individuals found many studies focused on immediate acoustic/aerodynamic effects, while medium- and long-term effects and dosage were much less often reported. Vocal Training in Healthy Individuals: A Scoping Review.

And a 2026 study highlighted in Ptichi’s research backlog found that increasing water resistance/load could reduce perceived ease in the tested conditions. The important product lesson is not “never use a straw.” It is dose and setup are part of the exercise.

More sensation is not more progress

Voice exercises are especially vulnerable to the “I can feel it, therefore it works” error.

A learner discovers facial vibration during humming. The sensation becomes the target. Then they try to maximize it. Soon the exercise is no longer about easier speech; it is about reproducing a sensation.

Sensation can be a useful cue. It is not proof of correct physiology.

Ptichi should treat somatosensory cues as optional anchors tied to a functional outcome:

Did this cue help me produce easier, clear speech on new words?

not:

Did I feel enough vibration in the mask?

Research suggests

Current voice literature supports a few bounded points:

• resonance reflects how the vocal tract filters sound;
• normal oral/nasal balance depends on the intended sounds and varies across languages/dialects;
• SOVT exercises can change phonatory conditions and may have useful immediate effects in some populations;
• exercise configuration and dosage matter;
• evidence from clinical populations does not automatically validate a healthy-speaker consumer program.

The evidence does not support a generic “nasality score” that tells every speaker to become less nasal.

Our read

Resonance is where product design can easily become pseudoscientific because the metaphors are so vivid: chest voice, mask, forward placement, open throat, nasal voice.

Metaphors can help a person explore a sound. The product should never confuse the metaphor with a measured anatomical state.

Ptichi’s safe model is:

perceptual contrast → effort/naturalness → listener consequence → optional supported acoustic description

That keeps the training useful even when the microphone cannot tell us why a voice sounds the way it does.

When ‘nasal’ may be a health/clinical question

ASHA distinguishes hypernasality, hyponasality and other resonance disorders, which can involve structural or functional causes and require differential assessment. A generic consumer recording is not enough for that job. ASHA: Resonance Disorders.

Persistent abnormal resonance, pain, worsening hoarseness, voice loss, swallowing/breathing difficulty or other concerning changes should not be treated as “train harder.”

This long read is not a diagnostic guide and not a therapy protocol.

A better self-listening experiment

Choose three short phrases containing both oral and nasal sounds:

We may need another minute.

The monitoring is normal now.

Marta made the final change.

Record them normally.

Then record a second set with one goal only: comfortable, clear production with less unnecessary squeeze.

Do not try to remove all nasal sensation. Do not chase maximum vibration.

Listen later and rate:

• clarity;
• effort;
• naturalness;
• whether the difference is actually useful.

If the answer is “it just sounds different,” that is a valid result.

Transfer: leave the humming behind

If you use an optional humming or straw exercise as a cue, the transfer test is connected speech without the prop.

After the exercise, say a new 20-second explanation. If ease disappears as soon as the straw leaves your mouth, the drill has not yet transferred to the task you care about.

Continue with projection without pushing when your real goal is audibility, or clear articulation when the listener is missing words rather than disliking the timbre. The voice-controls overview keeps these problems separate on purpose.