Recommended Slug: pitch-vs-timbre-vowels
Pitch is not "the highest frequency in a sound"
The human voice is not made of a single frequency alone.
For example, a voice with a fundamental frequency (F0) of 220Hz generally contains components that are integer multiples of that frequency, such as:
- 220Hz: Fundamental frequency
- 440Hz: 2nd harmonic
- 660Hz: 3rd harmonic
- 880Hz: 4th harmonic
If you play 440Hz by itself, it is a note one octave higher than 220Hz. However, when 220Hz, 440Hz, 660Hz, and 880Hz all sound simultaneously in these integer ratios, the ear typically perceives them together as a single sound based on 220Hz.
When a frequency doubles, the pitch rises by one octave. This principle is perfectly compatible with the fact that "high-frequency harmonics are contained within a low-pitched sound."
Even if a tuner shows the same note name, the pitch may not be identical
The large note names displayed on a tuner are categories based on semitones.
For example, even if the display continuously shows "C3," the following variations can exist within that range:
- A C that is 20 cents lower than the center
- A C that is almost exactly at the center
- A C that is 20 cents higher than the center
A semitone consists of 100 cents. Even if the note name does not change, there are minute deviations that determine whether you receive a "gold" score on DAM.
Therefore, when changing a vowel from "ee" to "oo," what you should check is not the letter of the note name, but whether the needle or the cent value has moved.
Timbre is "the characteristic that lets us distinguish sounds of the same pitch"
Even with the same pitch, volume, and duration, a piano, a guitar, a violin, and the human voice all sound different. That difference is timbre.
Timbre consists of several elements:
- The distribution of harmonic strength
- The attack time (how quickly the sound starts)
- How the sound decays over time
- The peaks and valleys in the spectrum
- Resonance bands (formants)
- Noise components
Terms like "bright" or "dark" are also ways to describe timbre.
The "Brightness" of Timbre is Related to the Strength of High-Frequency Components
In psychoacoustics, brightness is not just a subjective impression. It is strongly related to how much energy is present on the higher frequency side, especially the spectral centroid.
- High harmonics are prominent: bright, sharp, hard
- High harmonics are weak: dark, round, soft
However, timbre is multidimensional, and brightness alone does not determine everything.
"I" and "U" have different resonances even if the vocal fold pitch is the same
The main sound source that creates pitch is vocal fold vibration. This sound is shaped by the vocal tract, which consists of the tongue, lips, jaw, and pharynx.
The shape of the tongue and lips differs significantly between "i" and "u," so the formants, which are resonance bands, change.
Generally, the following perceptions are common.
- "I": bright, sharp, forward, high
- "U": dark, round, back, low
However, if the tuner's cent value does not change, the actual fundamental frequency F0 is almost the same. What has changed is mainly the timbre.
It is not uncommon to confuse changes in timbre with changes in pitch
Studies have shown that pitch and timbre are theoretically separate perceptual attributes, but it has been confirmed that they can interfere with each other during judgment.
In particular, changes that increase the spectral centroid and make the timbre brighter can be perceived as a rise in pitch, or vice versa. Even with musical training, this interference does not always disappear completely.
Therefore, the experience of "when I changed to 'u,' it sounded like it dropped by a semitone, but when measured, it was the same" is not a contradiction.
What the ear detected was a real change. However, there is a possibility that the type of change was misclassified as "pitch."
Vowels have a tendency for inherent F0, but it cannot be said that "u is always a semitone lower"
In speech research, it is generally reported that high vowels tend to have higher inherent F0 than low vowels. In other words, the articulation of vowels and F0 are not completely unrelated.
However, from this tendency, it cannot be said that "everyone will drop by a semitone when making a 'u' mouth."
A semitone is a large difference of 100 cents. If the measured value drops by nearly 100 cents during a vowel change, there may be other actions that are coordinated besides the lip shape.
- Pulling the root of the tongue strongly backward
- Closing the jaw
- Relaxing the area around the larynx
- Lowering the tension of the vocal folds
- Suddenly reducing the volume or air pressure
On the other hand, if the cent value does not change, the mouth shape is not the main cause of the perceived pitch drop.
The pitch that DAM sees and the "high" that humans feel do not match
The pitch evaluation of DAM mainly compares the movement of the fundamental frequency estimated from the singing voice with the scoring reference.
If the timbre is just dark, it is generally not judged as "a semitone lower."
However, if the following problems occur during vowel changes, it can indirectly affect the trajectory.
- The voice becomes too quiet
- The voice becomes breathy and the cycle becomes unclear
- The F0 drops for a moment during vowel transition
- The voice becomes unstable due to movement of the tongue or jaw
- A stable section cannot be created with short notes
The problem is not the darkness of the vowel, but the actual changes in F0 and voice stability accompanying the vowel change.
Rather than discarding the ear, divide the roles between the ear and the machine
The ear is strong in the following judgments.
- The melody is very different
- It is off by a semitone
- The timbre is bright or dark
- The voice is strained
- The expression is natural
- The impression is different from the original song
The machine is strong in the following judgments.
- Small deviations of 10-30 cents
- A drop only at the beginning of the note
- A drop only at the end of the note
- A momentary change during vowel transition
- The difference between one's internal sense and external output
Therefore, in improving the score, it is important not to "disbelieve the ear," but to "not let the ear alone be responsible for measuring tens of cents."
The conclusion is that "even with the same note name, the timbre changes, and the timbre affects the impression of pitch"
It is not wrong that it sounds different between "i" and "u."
There are three possibilities:
- The pitch may have changed
- Only the timbre may have changed
- Both may have changed
Using the cent value of the tuner or the DAM trajectory, you can distinguish these differences. The note name display alone is not sufficient for detailed judgments.
