A friend recently asked if you could surgically make someone sound better. My explanation was it’s not a testable thing (you can’t really test a voice to check if it’s good before you consider the operation complete). So it’d be tough to test unlike visible surgeries like a nose job. I was wondering if there’s a scientific way to check if someone sounds good when they sing rather than testing by hearing them.
In: Biology
It’s definitely testable, just not objectively measurable. The only remotelt objective measure of whether something is good is how popular it is.
Singing on key is objectively measurable. Notes are a mathematical distance from each other in regard to frequency. Now, there is more to singing than that, but if you can sing on pitch, everything else can be taught.
There are a couple issues with these assumptions. Let’s start with the actual physical act of singing (I’ll touch on the concept of if someone “sings well” afterward). Singing uses a whole complex set of interrelated systems: your vocal chords in your larynx, the shape of your throat, how you shape your mouth, your nasal cavity, etc. Small changes (ie the shape of your lips, position of your tongue) in one part can have large effect on the sound that exits your mouth. Much of singing is about technique: muscular mastery of these systems to get the exact pitch(es) and timbre the singer is aiming for. It takes years and years to gain the command to have an actual “good voice”.
Because this system is so complex, your voice is constantly changing: hour by hour, from day to day, and even over the course of your entire adult life. Think about how much you voice changes when you get just the sniffles, not even a full-blown cold. Test results at one point in time would not match results from another point in time.
Lastly, what does it mean to “sing well”? That is a very subjective notion. Different cultures, and even different people within a culture, will define such a trait differently. Let’s pick American singing tradition. Listen to any top 40 song today and then go listen to a pop song from 100 years ago. [For example, try this one right here.](https://www.youtube.com/watch?v=-Ao63TvxvPk) The singing techniques are wildly different and any scientific audio analysis would give different results. Edith Day is going for a different vocal sound than modern singers. Let’s pick Adele. Can you definitively say from this that Edith Day here sings better than Adele or vice versa? And we are just dealing with Western style music here. Other cultures don’t even use the same pitches and scales that America does.
From my knowledge, we humans can’t modify vocal cords like that. I’ve been made aware that just about any form of contact with them damages them in very special ways.
For the other part: What can be tested is how someone can correctly produce a certain sound. But that’s not hinged on the cords, it’s on the person’s usage of them.
Singing “good” is like, 90% breathe control and relaxing a bunch of muscles in your vocal tract. Most if not everyone could sing “good”, we just tense up with expectation and imitation. What we will be able to do, probably fairly shortly, is transform someone’s singing voice to sound like whomever they would like, even live, with computers. But that isn’t objectively better.
Without hearing them? No. Singing is an exercise in anatomical coordination as well as precision in that coordination. Thats all muscle memory, it comes from the brain.
You could check that someone sings properly on tune (basically how much auto-tune would need to correct it). That’s not telling you if it’s a pleasant voice per say, but it’s already a huge part of singing well. And it’s not just a binary thing where someone is either off-key, or right on. Every-one is going to be somewhere between perfection and horrible. Some people are just close enough to perfect that regular folks can’t tell the difference. In some cases, slight “imperfections” can also be a style effect.
But surgery is not going to give you that. It has more to do with your ears, training, and technique, than about your vocal chords.
Regarding if a voice itself sounds pleasant, regardless of if it’s on key, that’s definitely subjective. But subjective does not mean untestable. You can still have a large panel of judge give subjective ratings to a lot of voices. And then machine learning algorithm can use that to extract the key features that makes a voice generally liked. It’s going to vary a bit with cultures, and you will always have some people who loves someone’s voice, while other hates it. But that doesn’t mean there can be clear general trends. It’s just like for physical beauty. It’s a matter of taste, but there are still people who are wildly considered beautiful, and others who aren’t.
You can objectively measure specific characteristics of someone’s singing ability, like pitch, tempo, vibrato, timbre, range, phrasing, breathing, articulation.
But “sings well” is too vague for a scientific study. You first have to precisely define what “well” means, because I think you’ll find it’s more related to things like popularity, appearance, lyrics, backing music, cultural norms etc than any specific collection of objective data points. And probably also what “singing” is, given how much variety there is in music, such as the existence of skat and spoken-word songs.
Latest Answers