What is timbre and can we measure it?

[Newman]

The premise of this discussion escapes me. It seemed to be brought up as "Timbre throws a curveball that challenges our understanding of Shannon-Nyquist" which is silly.

I got more the impression that the OP did not realise that a musical note was anything other than a single frequency: "sound of different instruments playing the same note" ... "passing a set frequency (a note)" ... " if notes have the same sound wave, how can they sound different?" ... "a bugle horn playing e flat and an electric guitar playing e flat are objectively different sounding (at least I think you can say that's objective). Yet, both should be around 311hz. So, if the frequency is the same, what's causing them to sound "different"?"

Not realising that when any musical instrument other than a sine wave generator plays a given note, a huge number of frequencies are in play.

 

[antcollinet]

Let's throw that question open. Someone here must have a word? I know a lot of you just want to use the word "timbre" for that anyway. I don't see how timbre for modifying the sound works.

I think it depends on the specific change.

In the example given (moving instrument from concert hall to gymnasium), the difference is obviously reflections, including reverb. The sound heard is the result of direct sound and (multiple) reflections within the space the instrument is played in.

Interesting to think about is that the sound emitted by the instrument has not changed. The timbre of the instrument is therefore not changed, and what we are hearing is that timbre impacted by the transfer function of the listening space.

So in this case, the word is “reflections."


If the sound were put through a low-pass filter, the word (phrase) would be "band limiting".

IMO, the word used needs to describe the specific changes.

 

[Geert]

Not realising that when any musical instrument other than a sine wave generator plays a given note, a huge number of frequencies are in play.

And that note has an 'envelope', in which the volume and frequency spectrum changes over time. That's probably all the OP was asking for.

From https://www.audiolabs-erlangen.de/resources/MIR/FMP/C1/C1S3_Timbre.html

 

[MaxwellsEq]

If we think us guys on the internet can come along with so.e presumptive thought experiment based on our semi informed POV relative to their actual expertise and show how they're wrong, it's really not that different than that guy on YouTube saying they debunked the sampling theorem

You can't attack my argument, so you attack my capability to design a thought experiment, in a roundabout way. You must know that when you stoop into ad hominem attacks, you are probably losing the argument.

I assume you are well aware that almost all psychological research is struggling with a replication crisis?

 

[audiofooled]

This presentation might be of interest:

May I ask then, what is it that we are trying to define (and hopefully come up with a method of how to measure) here?

EDIT: link to Dan Russell's web page https://www.acs.psu.edu/drussell/demos.html

 

[solderdude]

You can only analyze a recorded signal.
It would be really complicated to 'grab' the essentials of that recorded signal in still plots or numbers.
Animated plots (over time) showing the spectrum from start to end might give some insights of that recording.
That is a lot of hassle to say which instrument one heard...

Would that matter in any way as well ?

 

[chervokas]

You can't attack my argument, so you attack my capability to design a thought experiment, in a roundabout way. You must know that when you stoop into ad hominem attacks, you are probably losing the argument.

I assume you are well aware that almost all psychological research is struggling with a replication crisis?

Have you read any of the research into timbre and perception? Psychological, neurological, any of it?

Your thought experiment seems interesting, but like all such things its just an imaginary hypothetical that may not may not have anything like the result you expect. We have no idea about it until it is a real experiment with results we can analyze. Also it's not clear to me it would even help us solve the most fundamental.questions in the science of timbre like why as humans to we recognize a clarinet as a clarinet even when two sounds coming from it differe in harmonic spectrum, spectrum envelope and attack?. Your ideas of what timbre is seem to be coming from a place that's not informed by specific knowledge of the science of timbre perception and cognition and all the work that's been done on it. Maybe I'm wrong, maybe you've studied it, rejected it and come up with your own idea of what timbre is. But my impressions was that you came up with the ideas.in the absence of specific knowledge of any prior work.

 

[ahofer]

Reread my post more carefully. Timbre is INDEPENDENT of pitch and intensity.

This cannot be, given that pitch and intensity are closely related to frequency and amplitude and how they vary over time, the only ingredients of sound. Even if there is some new ingredient of sound you are proposing other than frequency and amplitude, timbre would still be strongly related to the former.

 

[MaxwellsEq]

Your ideas of what timbre is seem to be coming from a place that's not informed by specific knowledge of the science of timbre perception and cognition and all the work that's been done on it

Back to ad hominem.

 

[chervokas]

This cannot be, given that pitch and intensity are closely related to frequency and amplitude and how they vary over time, the only ingredients of sound. Even if there is some new ingredient of sound you are proposing other than frequency and amplitude, timbre would still be strongly related to the former.

Actually timbre is 100% unrelated to pitch and loudness. It's the whole definition of timbre -- that attribute of auditory sensation that allows us to distinguish between two sounds similarly presented and having the same loudness and pitch. It's the whole point and definition, and to some degree the source of the auditory mystery of timbre. But by definition, timbre IS independent of pitch and loudness. Of course pitch and loudness are also both perceptions, related to but not the same as the acoustic correlates in frequency and intensity, so, for example we have an idea that timbre seems to have some, though not wholly understood or clear, relationship to the harmonic spectrum and harmonic spectrum envelope of the sound stimulus. But by definition the perceptual auditory attribute we call timbre is independent of the perceptual attributes of pitch and loudness.

 

[chervokas]

Back to ad hominem.

It's not ad hominem. I'm not attacking you personally. I'm questioning whether your ideas are informed by specific knowledge of the science of timbre that you are responding to, or whether they are generalized ideas of your own unrelated to particular knowledge of the decades of research into the subject. It's not about you personally, it's about understanding the the basis of the ideas you're advancing -- are they or are they not based in specific knowledge of the relevant science.

 

[Theta]

My understanding is timbre is describing the differences in sound of different instruments playing the same note.

My understanding is also that this site effectively evaluates musical equipment by passing a set frequency (a note) and evaluating how faithfully the equipment reproduced that sound.

So here are my questions. First, what is timbre? Or maybe I should say what causes timbre? I have an extremely rudimentary understanding of how ears work but if ears just pick up vibrations in sound waves and if notes have the same sound wave, how can they sound different? What's the physical phenomenon that allows this?

Secondly, is this picked up on our "normal set" of measurements? Is there something in a FFT that would represent the "timbre" of the sound? And if not, is this a potential area that measurements are lacking?

Thanks.

Timbre is essentially important in the voice region. It refers to the way you identify the voice of your wife in comparison to the voice of your sister. In Hi Fi it is what enable you to identify a Boisendorfer piano compared to a Yamaha piano or a fender telecaster to a Gibson Les Paul.

 

[Keith_W]

A good demonstration of the subjective / individual nature of timbre can be heard with Yanny/Laurel:

Some people hear "Yanny", some people hear "Laurel". It is exactly the same sound. But why do we hear one or the other? It is because Yanny/Laurel have both sounds mixed together. If you are more attuned to higher pitch, or if you do not have high frequency hearing loss, you will hear "Yanny". But if you don't hear higher pitch, you will hear "Laurel". What you hear also depends on your playback device, but even then two people sitting next to each other might hear differently. On some devices, I can hear both depending on how my brain was primed (e.g. if the word YANNY flashes on the screen, I hear "Yanny").

 

[ahofer]

Actually timbre is 100% unrelated to pitch and loudness. It's the whole definition of timbre -- that attribute of auditory sensation that allows us to distinguish between two sounds similarly presented and having the same loudness and pitch. It's the whole point and definition, and to some degree the source of the auditory mystery of timbre. But by definition, timbre IS independent of pitch and loudness. Of course pitch and loudness are also both perceptions, related to but not the same as the acoustic correlates in frequency and intensity, so, for example we have an idea that timbre seems to have some, though not wholly understood or clear, relationship to the harmonic spectrum and harmonic spectrum envelope of the sound stimulus. But by definition the perceptual auditory attribute we call timbre is independent of the perceptual attributes of pitch and loudness.

I see the field-specific (but tautological) definition. But in the context of audio engineering this seems to stretch the normal definition of “independent” beyond recognition. Our perception of pitch is highly dependent on frequency, and so is timbre. But I’m not sure what these semantics have to do with this discussion.

This is an audio site. The accurate or inaccurate reproduction of pitch and timbre of instruments in audio reproduction is a creature of frequency and amplitude as recreated by components in a space and interpreted by the listener’s brain. There has to be a way to discuss these things “independent of” the latter (which is where your definitions are more important) without getting corner-cased to death.

EDIT: trying to be concise and falsifiable: In audio reproduction, if “timbre” or “pitch” are perceived as inaccurate it is either a feature of the signal at the ear or some processing in the listener’s ear/brain. The signal is made entirely of frequency and amplitude as experienced over time, and those attributes dominate audible timbre. Wow and flutter, for instance, perceived as pitch variability, are measurable in the signal. Psychoacoustics are obviously critical to musical experience, but apart from status and appearance, it’s difficult to assess the psychoacoustic effects of amps, speakers, DACs and turntables in any generalizable way.

 

[Titurel]

This cannot be, given that pitch and intensity are closely related to frequency and amplitude and how they vary over time, the only ingredients of sound. Even if there is some new ingredient of sound you are proposing other than frequency and amplitude, timbre would still be strongly related to the former.

If you understand that timbre and vowel are very similar phenomena and, if you accept that vowel identity is not bound to pitch and intensity (or gender), it's not a difficult jump to accept that the same is true for timbre. This is why I suggested that Gringoaudio1 look into formants. For you, I will go a step further and suggest strongly that vowel formants and timbre are very closely linked concepts.

addendum: Understand that in the context of the standard definition of timbre, the term pitch is synonymous with "fundamental pitch" or simply "fundamental," not frequency. Similarly, intensity refers to the intensity of the entire sound. I highly recommend folks following this thread to look at spectrograms of vowel formants. I haven't read the entire entry, but the first couple of paragraphs in article linked below cover the basics pretty well.

Phonetics - Vowel Formants, Acoustics, Articulation | Britannica

Phonetics - Vowel Formants, Acoustics, Articulation: The resonant frequencies of the vocal tract are known as the formants. The frequencies of the first three formants of the vowels in the words heed, hid, head, had, hod, hawed, hood, and who’d are shown in Figure 3. Comparison with Figure 2...

www.britannica.com

 

[Cbdb2]

You guys are arguing about semantics. in this definition "...that attribute of auditory sensation which enables a listener to judge that two nonidentical sounds, similarly presented and having the same loudness and pitch, are dissimilar."

Loudness is not the same as amplitude of the individual fourier components its how loud the sound seems, yes related but not the same. Pitch is not the frequency of the fourier components its the frequency thats percieved, if at all, again similar but not the same, think about formants or non harmonic sounds. Whats the pitch of a spoken word or a cymbal? No pitch but lots of fourier components so there not the same.

From that definition I get that Timbre is how a sound is percieved, related to the signal but not the same. Like I said earlier, rather vague, and probably why its hardly ever used.

 

[Titurel]

You guys are arguing about semantics.

Not entirely. Any semantic argument in this thread principally derives from people assuming a word's meaning is independent of the domain in which it is used. This is a difficult sale to absolutists, of which there are more than a few in this forum.

As to the original 2 conjectures, I believe the answers are "It's complicated*" and "Not to any great purpose until someone develops a standard unit of measure and/or representation readily applicable to any (audio) task other than head scratching.

* alternatively, It's open to debate.

 

[Gringoaudio1]

Well not exactly.

Any signal that is not a sine wave can be transformed in the frequency domain into a set of individual sine waves.

But harmonics of a tone are always at multiples of that tone. So a signal consisting of 50Hz, 105Hz, and 233Hz does not contain any harmonics.

Very good point. I’d better find a more accurate term for those non harmonic frequencies. Sloppy of me. What do you call them? I’ve seen ‘inharmonic overtones’ used to describe them just now while searching.

 

[Guermantes]

Very good point. I’d better find a more accurate term for those non harmonic frequencies. Sloppy of me. What do you call them? I’ve seen ‘inharmonic overtones’ used to describe them just now while searching.

I was taught "partials" as a general term -- they can be harmonic or inharmonic and includes the fundamental.

 

[Gringoaudio1]

As a Fourier analysis of a sound would show many sinusoidal components that all combine andcontribute to the character of the sound (I’m afraid to use the word ‘timbre’ now as it may be reserved to describe the character of musical instruments or voices) I asked MetaAi to summarize what causes a sound to be perceived as having a pitch. I know the answer I just wanted it written by other than me.
MetaAi
“For a sound to have a perceived pitch, the Fourier analysis would reveal a concentration of energy around a specific frequency, often referred to as the fundamental frequency. This concentration of energy is typically due to the presence of a strong fundamental frequency component, along with its integer-multiple harmonics (e.g., 2x, 3x, 4x, etc.).

When enough of these sinusoidal components dominate at or around this fundamental frequency, our auditory system interprets this pattern as a specific pitch. This is known as the "pitch perception" or "tone perception" mechanism.

In essence, the brain analyzes the distribution of energy across different frequencies and identifies the fundamental frequency as the perceived pitch. This process is often referred to as "pitch extraction" or "pitch detection".

It's fascinating to explore how our brains process complex sound patterns to create the sensation of pitch!”

And:

“Non-harmonic overtones, also known as inharmonic partials, can indeed contribute to the overall timbre or "tone color" of a sound, but they don't typically contribute to the perception of pitch.

In fact, the presence of non-harmonic overtones can sometimes make a sound seem more "noisy" or "complex," rather than affecting its perceived pitch.

However, there is an interesting exception: when non-harmonic overtones are present in a sound, they can sometimes create a phenomenon called "pitch ambiguity" or "pitch uncertainty." This is where the brain has difficulty pinpointing a single, clear pitch, due to the presence of these non-harmonic frequency components.

So, while non-harmonic overtones don't directly contribute to pitch perception, they can influence the overall sound quality and, in some cases, create pitch ambiguity.”