This blog post is an extended version of a response to a question in The Modern Vocalist Forum from Mr. Steven Bradley, who wrote requesting some analysis of the scream as done by Steven Tyler in'I Don't Wanna Miss a Thing.
Discussion - Finding a Cool Video of Steven Tyler's Vocal Cords During a Scream
I was not really able to do an analysis of the particular section he had in mind of that song, but I did find a wonderful little section of a National Geographic special on YouTube, in which Steven Tyler's ENT (Dr. Steven Zeitels, of Harvard Medical School and Mass General Hospital ) was interviewed, and which included some video of stroboscopic laryngeoscopy with simultaneous audio, which makes the vocal cords look like they are moving in slow motion while the listener is able to hear the sound that is being made. Yes, that's right, a slow-mo of Steven Tyler's vocal bands during a scream that you are also able to hear at the same time.
Check out the whole story to get the sense of the article. The whole thing is very much worth watching, even if the narrator calls his scream 'falsetto'.
At about 1:10 in this video, which was made with a combination of performance and ENT's video and audio, you can see some very interesting things. Right at that time, (with the strobe on) you see the cords vibrating slowly and hear his tone quality in the ENT's office while he is singing that top Ab. There are a couple things to note:
- Location of Oscillation: Right before the narrator cuts back in, Steven sustains the Ab he is screaming for a moment, while it is being scoped. It's at that point that the close observer can see fairly clearly see what is going on, including what part of the vocal folds are vibrating. The scoop here: The vibration is at the 'top' part of the picture, which during the exam, is the posterior part of the vocal bands. In a word, this is not 'zipped up' so that vibration is only at the front section of the vocal cords as one sometimes sees, but taking place at the back close to the arytenoids.
- Phonation Motion The motion of the vocal bands is not uniform during the scream, but can be seen to have some complexity or variation . This kind of phonation results in complexity of the glottal pulse wave shape, which will be reinforced by the vocal tract if some of the frequencies are close to vocal formants. Here is what that looks like spectrographically:
The Scream, Spectrographically
To understand what this shows us, follow along from left-to-right:
The first upward peak is the sung fundamental (also called the first harmonic, or H1). Its about the Ab above the treble staff, (at 820 Hz, for those interested - I measured it). It is a very narrow peak, indicating that he is singing it with no vibrato. Look to the right now, and find the highest peak on the chart. This is at 1640 Hz, exactly twice the frequency of the fundamental, which makes it a harmonic as well, in this case, the second harmonic, H2. It is also narrow. Find the third narrow peak, and you will see the third harmonic (H3), which is at 2460 Hz. Above that, there is not much intensity of sound. NB At this point: If these three harmonics were the only sound present in the vocal tone, it would be very powerful, and sound very pure to us... clean and clear.
But, Steven's sung tone is not that simple. It has a complexity, which is a part of his 'scream' vocalism. Let's look at that.
Between the H1 and H2 peaks is another one, with a fairly wide base that ramps up to a peak almost as tall as H1. That peak is fairly wide (when compared with the harmonics we have been discussing up to this point,) and then it ramps down in a way that the harmonics do not. This display is characteristic of a multi-frequency cluster of sound energy, which is being amplified selectively by a vocal resonance, IMO probably due to the lowest vowel formant, F1. The line traces the shape of the resonance influence of the formant as it selectively amplifies the pink noise input.
Just FYI: The frequency 'center' of this sound energy is about 1218 Hz, which is not a harmonic multiple of the sung fundamental. These partials are therefore 'non-harmonic partials', components of the sound, which are not in the expected harmonic series, so our brain interprets them as noise.
The overall effect of the combination of three strong sung harmonics, with the loud noise component, gives us this particular version of the 'Steven Tyler Scream'.
Request for a '2nd Opinion'
I am by no means an expert in either the analysis of vocal screaming, or of interpreting the nuances of phonation characteristics as revealed by stroboscopic laryngeoscopy. However, there are TMV members who are very familiar with both of these territories. My request: Write comments to this blog, to affirm what I have offered, or to give other opinions about what you see or hear. I am very interested to know your reactions.
This essay was first published January 9, 2009 on The Modern Vocalist.com the Internet's #1 community for vocal professionals, voice health practitioners and pro-audio companies worldwide since November 2008.
Report Record
