Waveforms and vowels
In electronics, 4 waveforms are generally used: Sine, square, triangle, and sawtooth. This is because, when measured with an oscilloscope, they display those forms. They seem something specific of this field, but in the true those waveforms are recurrent in nature, except the square one.
Let us analyze the human voice, in order to determine the waveforms of vowels. In order to obtain a clean form, we need to sing them with a low voice. Otherwise, there will be other resonances and oscillations that will modify the original form.
The A vowel is the most open one. It displays a form similar to a reverse sawtooth.
The A vowel
The sawtooth waveform
A sawtooth wave's sound is harsh and clear and its spectrum contains both even and odd integer harmonics of the fundamental frequency.
The spectrum of the A vowel
The content in harmonics depends also in the sound pressure we exert while singing. As we were singing with a low voice, the harmonics do not emerge in all their power in the image above.
From a sawtooth wave, we can infer that the A vowel contains few low pitched harmonics, but a lot of high pitched ones. Indeed, it tends to go out and stay neither in the chest nor in the head.
Moreover, the sawtooth wave happens particularly in bowed string instruments, such as violins and cellos, which are driven by a sawtooth-like motion.
The E vowel is in the middle between a sine wave and a sawtooth wave. Therefore, it tends to be rich in higher harmonics and partially lacking in the lower ones. As today everybody seems to distort recording by enhancing the higher tones under the pretext of making the music more "airy", a singing style supported over the E vowel is a perfect match. It also makes singing easier, as less air is needed and sound emission can be better controlled.
The E vowel
The spectrum of the E vowel
The I vowel is almost a perfect sine wave and lacks all the higher harmonics. It sounds like a flute, which produces an even more perfect sine.
The I vowel
The sine waveform
The spectrum of the E vowel
As it lacks harmonics, the I vowel is not very fit as a support.
The O vowel is similar to a triangle waveform. A triangle wave contains only odd harmonics. However, the higher harmonics roll off much faster than in a square wave (proportional to the inverse square of the harmonic number as opposed to just the inverse). Considering this peculiarity, the O vowel is not the best choice for supporting the other sounds.
The O vowel
The triangle waveform
The spectrum of the O vowel
The U vowel is the closest one. It is sung lowering the larynx and extending the vocal tube the furthest it can. Therefore, it reounds well in the chest and has a dark aspect.
The U vowel
Its shape resembles a reverse sawtooth, which implies that the lower harmonics are enhanced, but the higher ones tend to disappear.
The reverse sawtooth waveform
It is this characteristic that confers a metallic, strong, and shiny character to the voice, when the sounds are supported on this vowel. This is the case of the lunge technique, one of the most prominent representative of which was Mario Del Monaco.
The waveform that is left out is the square wave, which has only odd harmonics and a very harsh character. It usually happens when we record music and microphones are overdriven. Therefore, the signal is distorted and clipped in its top, forming a square wave.
The square waveform
When the signal is clipped, depending on the technology you are using in the amplification stages, you have a different harmonic content. Valves produce only consonant harmonics. Therefore, even when you do a bad recording, it sounds nice. All solid state technology (transistors, FETs, etc.), instead, tends to produce only dissonant harmonics, increasing the distortion.