Is it? It appears that the word “timbre” is used for several ‘things’, generating misunderstandings, confusion, and arguments.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
What is timbre and can we measure it?
- Thread starter DesertHawk
- Start date
It is. There are many colloquial usages of the term, of course, and a lot of pop descriptions of timbre that are incomplete (just like you get a lot of explanations on YouTube about our hearing works that only describe the function of the ascending auditory pathway and don't even mention the existence, never mind the function of the descending auditory pathway, leading people into misleading impressions that hearing only works from the ear to the brain not also the brain to the ear leaving the misapprehension that the ear is kind of a dumb mechanical transducer and that hearing is just a one way, mechanical process). But if you spend even a small about of time looking at the actual scientific research into timbre, whether that's McAdams, other current researchers, or the decades of prior work, you'll see that, yes, as understood by auditory science, it is a perception, not a physical attribute. Just like pitch is a perception, related to but not the same as (an not always equivalent to) periodicity of a sound stimulus, and loudness is a perception, related to but not the same as sound intensity. If you go back 140 years to classical ideas about timbre in the time of Helmholz, the idea was that timbre, or "tone color" in Helmholz's term, was an attribute of sound specifically related to the sound's harmonic content. But just like classical physics doesn't explain everything we understand today in quantum physics, the classical idea of timbre from Helmholz isn't the scientific understanding of timbre in 2025.
Now, like I said, timbre is clearly related to sound characteristics that we can measure in the physical world, especially harmonic spectrum, harmonic spectrum envelope and attack characteristics, but timbre itself is not the same thing as those things.
That's not true at all. Difference and thresholds of difference and perceptibility of difference is one of the major thing that is studied in psychoacoustics, and in timbre perception studies there's actually a pretty wide range of variation among subject -- some people can't tell the timbre of two instruments apart sometimes at all in a recording, where other subjects can, even with the same stimulus. It's not the stimulus difference that's the only element in perception of timbre, and subject listeners have differing sensitivities to different aspects of timbre. In addition, at least in part, perception of timbres that, say, allow us to identify one instrument from another, are, in part, learned and involve memory and we don't all necessarily share the same memory and learning. I highly recommend spending some time doing a little reading in psychoacoustics and timbre. The intro part of McAdams' lecture here presents not only a little overview of some of the cognitive scientific thinking on timbre but also data from research of his own showing varieties of identification responses among subject to the same stimulus. People can at do have different auditory perceptions of timbre, sufficiently different for them to be able to or not be able to identify if the particular sound is coming from one instrument or another with the identical stimulus, because people hear timbre differently and have different sensitivities to different elements the inform a perception of timbre and have different experiences and memories that shape their perceptions.
Timbre is one of the thornier and more complicated topics in auditory science. We have a working definition that kind of only defines it by what it's not, it is the thing we say allows us to distinguish one sound from another when thay have the same frequency amplitude duraion locaton etc and yet we all can successfully always distguish one sound from another on the basis of timbre. Yet we all experience it. In fact it's a fundamental part of human sonic experience. Maybe its mysterious thornines is one of the reason its been the subject of so much scientific study.
Last edited:
Zog
Active Member
- Joined
- Mar 13, 2019
- Messages
- 265
- Likes
- 296
Well, yes of course. But here we are dealing with hi-fi. That is to say hi-fi gear. So isn't it a matter of reproducing the musical quality?
- Joined
- Apr 19, 2021
- Messages
- 460
- Likes
- 659
Where is any proof that something measurably identical but with some difference in timbre can be audibly identified in a controlled listening test? I'm quite certain you won't find any. I'm perplexed and we must not be addressing similar topics. Maybe you are conflating some other issue and attributing it to timbre and transducer reproduction?
But what are those constituent components? How does one record and represent them in useful ways, such that they reveal how well listeners might be able to differentiate one timbre from another? One can almost read spectrograms, but they are more straightforward transformations (from aural to visual) than measurements. Which puts us back at the original question, perhaps better informed for the discussion, but without much of an answer unless it is "no, not really."
The above presumes that every relevant parameter is known and is being measured, and with psycho-acoustical phenomena, whether the test protocols are capable of capturing all the useful data. The former is difficult (if not impossible) to prove, the latter possibly worse.
- Joined
- Apr 19, 2021
- Messages
- 460
- Likes
- 659
It is not that complicated, just show a controlled test where the known, typical measurement are well within audible transparency, and the test subjects were able to successfully show they could identify an audible difference.
- Joined
- Apr 19, 2021
- Messages
- 460
- Likes
- 659
It seems ridiculous, right?
That's beyond de minimis. What are you measuring? It sure isn't timbre. You specify a controlled test. What is the unit of measure? What are the test protocols? You can measure the length and weight of a piece of tartan, even its colors, but how do you measure its essential "tartan-ness?" That's timbre.
Perhaps we need to rewrite the parable of the three blind men and the elephant, to the three deaf men and the saxophone.
Last edited:
Tovarich007
Active Member
- Joined
- Jan 31, 2021
- Messages
- 278
- Likes
- 385
Yes, with sophisticated equiment and a full set of measurements, because timber has objectve characteristics.
Of course, it's perceived subjectively, just like everything in life, that's the reason why, for example, people with a bad audition are able, however, to distinguish immediately a voice they know well.
Go back on page 1 of this thread and read again the very good definition of timber found on ChatGPT, it says it all.
Keith_W
Major Contributor
Um, I think you guys need to pay attention to what @chervokas is saying. Timbre is constructed in your head. There are measurements that correlate with timbre, but no actual measurement of timbre.
It's just like soundstage and imaging, it is a construction of your brain. There are measurements that correlate with soundstage, but no actual measurement of soundstage. The more I learn about measurements, the more I realise that microphones + software hear differently to ears + brains. Toole emphatically makes this point. We do not discard the baby with the bath water, we learn to interpret measurements more carefully and with knowledge of psychoacoustics.
Unfortunately, psychoacoustic science seems to be in its infancy. I think there may be many subjective phenomena that we do not have a direct measurement for, it is at best a collection of different measurements which have to be interpreted. It's not like a tone where everybody agrees that sibilance is a treble tilt, or reverberation that we can easily measure. I have been trying to understand why big speakers sound big, why big subwoofers sound big, why dipoles sound open, why horns have a certain physicality to the sound, and so on. My subjective impression of these phenomena is so strong that I am certain it's real. Either I don't know what the measurement is, or audio science has not come up with a measurement technique for it yet. I doubt if it's going to be something as simple as a static single microphone measurement, it may require a bank of microphones measuring at different points in space. Don't be limited by your lack of imagination and dogma that everything that can be measured has already been measured. I think there is much we do not understand.
It's just like soundstage and imaging, it is a construction of your brain. There are measurements that correlate with soundstage, but no actual measurement of soundstage. The more I learn about measurements, the more I realise that microphones + software hear differently to ears + brains. Toole emphatically makes this point. We do not discard the baby with the bath water, we learn to interpret measurements more carefully and with knowledge of psychoacoustics.
Unfortunately, psychoacoustic science seems to be in its infancy. I think there may be many subjective phenomena that we do not have a direct measurement for, it is at best a collection of different measurements which have to be interpreted. It's not like a tone where everybody agrees that sibilance is a treble tilt, or reverberation that we can easily measure. I have been trying to understand why big speakers sound big, why big subwoofers sound big, why dipoles sound open, why horns have a certain physicality to the sound, and so on. My subjective impression of these phenomena is so strong that I am certain it's real. Either I don't know what the measurement is, or audio science has not come up with a measurement technique for it yet. I doubt if it's going to be something as simple as a static single microphone measurement, it may require a bank of microphones measuring at different points in space. Don't be limited by your lack of imagination and dogma that everything that can be measured has already been measured. I think there is much we do not understand.
solderdude
Grand Contributor
What would be a typical excitement of a speaker other than it having to (re)produce sound by an electrical stimulus ?
A 'hit' on the speaker cabinet is not the same as an impulse applied electrically and would de the equivalent of your suggestion to hit the lid of the piano with a hammer.
Have you ever seen musicians use different speaker boxes playing them like instruments ? If so what is their timbre ?
What would be a correct comparison is to 'hit' the snares of a piano with something else than the hammer inside or when you hit the snares somewhere midway.
In this case there will still be a tone with a specific character, attack, sustain and decay.
Nice try but a speaker is not a musical instrument or voice which can have a timbre... speakers, acoustics and electronics can alter the timbre.
One could use a speaker cabinet as a cajón but it won't sound like a cajón.
solderdude
Grand Contributor
Yes it is possible to argue about the word timbre simply because there is no clear all compassing definition of it.
One speaks of timbre when something produces a certain (and usually distinguishable sound).
Exactly my point.
Speakers that measure really flat and have no deal breaking resonances will not have a 'sound' (timbre to some).
Speakers that do not measure flat and/or have substantial resonances etc. will alter the intended sound in a certain way.
When one is really known with the different tonal character of speakers you can pick them out.
That would be similar as putting a reference speaker in the enclosure of an octobass, piano or snaredrum. These will alter the sound that is emitted by the speaker in a certain way that quite likely would have a character but not the same as when that instrument is being played on.
The 'timbre' will differ.
The cabinet the speaker is placed into won't have the same timbre as the actual instrument.
So for those reasons you could say a speaker does not have a timbre but can alter a timbre. Of course everyone is free to call it a 'timbre' when they perceive it as such.
Just like everyone can say a car has a 'personality' but in fact only the driver of the car has a personality... O.K... a self driving car... will that have a personality ?
The discussion about what 'timbre' is and what can have a timbre and what does not can certainly last forever. After all... but 10 scientists in a room and try to get them to reach consensus ... personalities ...
Only when something is really clear like the earth rotating around the sun there will be consensus among scientists. Just don't put a flat-earther in that mix though.
boxerfan88
Major Contributor
- Joined
- Oct 23, 2018
- Messages
- 1,028
- Likes
- 1,127
antcollinet
Grand Contributor
Really nice post - and an additional (belated) welcome to ASR from me.
Just putting those things in juxtaposition to clarify: The science of auditory perception must - by definition - exist in the domain of audible sound. Ie those sounds that exist within the bandwidth and sensitivity limitations of human hearing.
What is sometimes described as the "objectivist cudgel" (or similar) is the recognition that those limitations exist, and that especially with electronics, the differences between most of the halfway decent devices already fall outside the capability of the human auditory system to detect, for all humans (or at least so damn close to it as not to matter for music listening).
Rather than a cudgel, I prefer to describe it as a tool we can use - in an audio world of out of control marketing - to help people separate out real value in audio reproduction gear from that which is basically snake oil. It is IMO the most important purpose of this forum, in using scientific/engineering based methods to evaluate that gear.
Not my intent to apply the cudgel to your post, but I want to ensure that the absolutely correct statements along the lines of "Timbre is more complicated than you think" and "it can differ from person to person and be changed by learning and experience." (etc) are not, as often happens amongst audiophiles, misinterpreted to mean "everything matters" and "there are an infinite number of veils to be lifted" (etc).
Last edited:
antcollinet
Grand Contributor
That is not what was stated.
If there is a change to the characteristic of the sound that creates a change in the perception of timbre, then that change in the sound can be measured. It might be difficult to assign a specific measured change to a specific perception of timbre without a complex study of perception, but that doesn't alter its measurability.
It seems like timbre is like the colour brown. Brown isn't a colour. Rather, by being a colour, it doesn't really exist, because colours don't exist - they're just what humans call the effect on their eye of certain frequencies. Even with colours which are on the electromagnetic spectrum and therefore could be said to "exist" (but don't tell Isaac Newton I said that) there are optical illusions involving shadows where in the context of the "original" colour and surrounding colours something looks red but in isolation you'd never call that colour/frequency red. So you could have a "brown detector" something which detects the combination of frequencies on the electromagnetic spectrum which we tend to call brown just like you could have a given "timbre" detector (which looks for some sort of signature of frequencies, amplitude and phase which represent what humans call a violin, for example), but it's not really detecting something that actually exists.
Interestingly the Isaac Newton quote I found where he spoke about colour not existing actually uses sound as an example:
"And if at any time I speak of light and rays as coloured or endued with Colours, I would be understood to speak not philosophically and properly, but grossly, and accordingly to such conceptions as vulgar People in seeing all these Experiments would be apt to frame. For the rays to speak properly are not coloured. In them there is nothing else than a certain power and disposition to stir up a sensation of this or that Colour. For as Sound in a Bell or musical String, or other sounding Body, is nothing but a trembling Motion, and in the Air nothing but that Motion propagated from the Object, and in the Sensorium 'tis a Sense of that Motion under the form of sound; so Colours in the Object are nothing but a disposition to reflect this or that sort of rays more copiously than the rest; in the rays they are nothing but their dispositions to propagate this or that Motion into the Sensorium, and in the Sensorium they are sensations of those Motions under the forms of Colours."
Last edited:
solderdude
Grand Contributor
a definition of timbre:
It looks like 'timbre' is used for voices and instruments and describes the differences where they can be told apart.
A speaker is neither an instrument nor a voice. It is for reproduction and has its own characteristics in frequency response, radiation pattern, resonances.
The fact that speakers can be told apart by their 'sound character' could qualify them as having a timbre. Mostly because of frequency response and radiation pattern.
But when used for reproduction of sound it alters (modifies) the recorded timbre and that alteration is what makes speakers distinguishable.
It looks like 'timbre' is used for voices and instruments and describes the differences where they can be told apart.
A speaker is neither an instrument nor a voice. It is for reproduction and has its own characteristics in frequency response, radiation pattern, resonances.
The fact that speakers can be told apart by their 'sound character' could qualify them as having a timbre. Mostly because of frequency response and radiation pattern.
But when used for reproduction of sound it alters (modifies) the recorded timbre and that alteration is what makes speakers distinguishable.
Galliardist
Major Contributor
All this still strikes me as another group of researchers who have taken what I described before as an "available word" and given it yet another meaning. I can see the value in the work, but isn't it really about how we recognise sound, not a property of the source, which a whole other body of meanings of the word timbre depend on?
This version of the word sits us right in the middle of the objective/subjective audio debate, of course, and yet again, I have trouble with it in our little corner of the world. We know - and many of us have done just this - that the sound, and therefore timbre as a property of the sound source, does not have to change one jot for us to perceive it as different, if there is scope for our brains to believe that. If we relegate a term to the realm of perception, and prevent it from being measured, that means one thing - subjectivism.
Suddenly every snake oil seller gets a free pass.
We do need a science of perception, but here I suspect we need it in different terms just because we need it without opening the door to the hordes of demons, if I can put it in such a melodramatic way.
Similar threads
