Consideration about Timbre

[Nowhk]

Hi all,

new on the forum, hope to post this in the right section, with a decent English (not my native language, sorry for that).

Since some months, taking experience with music on its technical aspects and such (I start to learn DSP years ago), I've grown some puzzlement regards timbre and its perception in general. Yes, what I'll write could be applied to everythings in life, but for some reasons I'll stick with Music, which is one of my favourite hobby/passion.

I'll go with steps, since in the past made a "complete discussion" have created some troubles (yes, I've tried to undertake this kind of discussion on different community). Well, lets go.

If we should describe timbre, its composed (in a vulgar "way") as the sum of its partials and relative amplitude, perceived considering its envelope during the time. A multidimensional thing, not like pitch or loudness.

The fact is: if I see it from this point of view, every time I playback a track/song on different setups/environments, it will introduce by nature "distortion" - i.e. color - SO the timbre I'll perceive will vary every time I play the song. Speaker's frequency response/waterfall, transients response, positions, my own head, reflections of the room, and many others aspect will affect both the "harmonic fingerprint" and the whole "time" envelope (i.e. affect the spectrum).

I'm not talking about differences that will change a guitar into a piano: of course that's not my interess. And its not what happens on usual loudspeaker choice. I'm talking about changing "details" of the current instrument's timbre, such as listening a bass (i.e. made with FM synthesis, for example) to enhanced bass headphone instead of a living room's loudspeakers.

So, here we go to the "first step" of this topic: is it correct to say that, in fact, the timbreI'll perceive (of an instrument, real or produced with sound design techniques) VARY between every single listening I'll have of that piece?

Curious about your opinions. Than I'll go "deep" with my question. Thanks to all participants

 

[tomelex]

Hi, yes, every time you hear that song on any different system it will have to sound different...the electronics, the transducer, the room.

Then, every time you hear that song on your same system, if you listen for something different each time you play it, you will hear different things, this is because you are not a test equipment but a human being with a very dynamic brain. If I play a song for you once and then I play the same song again for you, the second time you will notice different things.

That is why the only possible standard in audio is that the entire system reproduce what is on the record or digital file as accurately as possible. Anything less is distortion as you say, it could be good distortion for you or not. Stereo is not in any case more than an attempt at an illusion of bringing an event to your house or ears. In my opinion, it is rather strange effect and also represents maybe, say 5% of what actually would be heard at a seat at the original event. Microphones and ears do not hear the same way, and a live event is not two point sources of sound.

I wonder what could be a deeper question....

 

[Cosmik]

I think that this is very relevant to the thread about 'room correction' above the Schroeder Frequency...

Of course, if you sample what arrives at the listener's ears it will be numerically different on each occasion unless all the conditions are absolutely identical i.e. everything in the room including the mic is rigidly fixed and not even the temperature and humidity vary. But it will not be arbitrarily different. It will differ by a consistent set of relationships that all lead back to the system's output - and your hearing is likely able to work back to that transparently. That's my theory anyway

 

[amirm]

So, here we go to the "first step" of this topic: is it correct to say that, in fact, the timbreI'll perceive (of an instrument, real or produced with sound design techniques) VARY between every single listening I'll have of that piece?

Welcome to the forum! And yes, you are very much right. It is the #1 fault in audio reproduction. Not only is the timbre different in all of our systems, it is also different in all the places it gets created!

We could fix this by capturing the timbre of the room that created the music and at least try to replicate that. Sadly there is no move toward that.

In video world this is done and indeed it is even mandated by government licensing for TV stations. But for audio, it remains a broke architecture. It is the wild west.

 

[Nowhk]

Hi all, and thanks for your replies and time
Well, this is pretty funny for me now.

Even if I totally agree with this, there are lots of people which claim that things such as Timbre is preserved (modulo the transfer function) across any listening by human perception, or that the differences won't be discriminated. Or also that the magnitude of difference you can get is "under the noise floor of interess".

Here's the whole discussion, if you are interessed (warning, its about 29 pages; I'm stubborn when I don't get the things). Unfortunately things gone weird there

Anyway, if thats true, so that our perception is able to compensate and ignore differences, whats the purpose of select one setup instead of another? You will always be able to "compensate", whatever mediums you are using. The same if you are considering only some part of the sound as foundamental.

Be aware that I'm mostly speaking about synth/studio sound, so the "live room recordings" has few meaning here (such as catch an orchestral experience as a whole).

 

[fas42]

Hi all, and thanks for your replies and time
Well, this is pretty funny for me now.

Even if I totally agree with this, there are lots of people which claim that things such as Timbre is preserved (modulo the transfer function) across any listening by human perception, or that the differences won't be discriminated. Or also that the magnitude of difference you can get is "under the noise floor of interess".

Here's the whole discussion, if you are interessed (warning, its about 29 pages; I'm stubborn when I don't get the things). Unfortunately things gone weird there

Anyway, if thats true, so that our perception is able to compensate and ignore differences, whats the purpose of select one setup instead of another? You will always be able to "compensate", whatever mediums you are using. The same if you are considering only some part of the sound as foundamental.

Be aware that I'm mostly speaking about synth/studio sound, so the "live room recordings" has few meaning here (such as catch an orchestral experience as a whole).

Very interesting first post, Nowhk! I haven't done so as yet, but will read that 29 pages worth, soon. In the meantime, will state that my take is that a particular recording should always sound the same, as a subjective experience; and, that all recordings will sound quite distinct from each other, at every level of impression - an analogy would be the people you know; that they are dressed differently, or are ill, makes zero difference in your ability to register all their distinct qualities - and they are are all very, very different in major, and extremely minor ways.

 

[amirm]

Here's the whole discussion, if you are interessed (warning, its about 29 pages; I'm stubborn when I don't get the things). Unfortunately things gone weird there

It is remarkable how we all focused on "macro level" issue of over fidelity and tonality and they zoomed right into "micro level" arguing over definitions, what happens at note component level, etc.

I guess we are big picture people and they, small.

 

[fas42]

Part of the puzzle is that the mind is a powerful assistant to putting together the subjective experience - if the clues and cues are of sufficient "strength" then our brains lock into a particular belief about some aspect of the auditory environment - we hear what appears to be a piano somewhere distant; if nothing contradicts that first impression, we are quite sure that it is a real piano, even though we are hearing only a tiny, tiny bit of what that instrument sounds like, if right next to us. The mind adjusts what is heard, and so, "it is a piano" - we decide to go and see the instrument in the flesh. As we get closer, all the auditory information keeps confirming it a piano - OR, discordant aspects start to rear their ugly head, contradictory sounds start signaling that it's fake, "it's only a hifi playing". The difference is that the brain now has enough information to switch 'models' of what's happening - we now are sure that it was nothing to get excited about, after all ...

The "filling the gaps" mechanism of the brain is extremely capable - fed the right info, it never decides to give up the impression that it is hearing some particular instrument or sound ... high quality playback can exploit that beautifully, and, it always "sounds the same".

 

[Thomas savage]

Hi all, and thanks for your replies and time
Well, this is pretty funny for me now.

Even if I totally agree with this, there are lots of people which claim that things such as Timbre is preserved (modulo the transfer function) across any listening by human perception, or that the differences won't be discriminated. Or also that the magnitude of difference you can get is "under the noise floor of interess".

Here's the whole discussion, if you are interessed (warning, its about 29 pages; I'm stubborn when I don't get the things). Unfortunately things gone weird there

Anyway, if thats true, so that our perception is able to compensate and ignore differences, whats the purpose of select one setup instead of another? You will always be able to "compensate", whatever mediums you are using. The same if you are considering only some part of the sound as foundamental.

Be aware that I'm mostly speaking about synth/studio sound, so the "live room recordings" has few meaning here (such as catch an orchestral experience as a whole).

As soon as one engages introspection to analyse consciously the workings of extremely complex sub-conscious mechanisms ( in this case our auditory goings on) we will often come a cropper.

The result is the origin of #FakeNews ..

 

[Nowhk]

The mind adjusts what is heard, and so, "it is a piano" - we decide to go and see the instrument in the flesh. As we get closer, all the auditory information keeps confirming it a piano - OR, discordant aspects start to rear their ugly head, contradictory sounds start signaling that it's fake, "it's only a hifi playing". The difference is that the brain now has enough information to switch 'models' of what's happening - we now are sure that it was nothing to get excited about, after all ...

Important to note (as I stated in the first post): I'm not talking about confuse a piano to a guitar, that's never is going to happens (except extreme cases, which is not in this discussion) I'm talking about details that be/won't be considerated on different listening, thus getting different perception of that intrument's timbre.

I know we focus each time on different details, but when we focus on detail X, I should perceive the same, every time.

The "filling the gaps" mechanism of the brain is extremely capable - fed the right info, it never decides to give up the impression that it is hearing some particular instrument or sound ... high quality playback can exploit that beautifully, and, it always "sounds the same".

If it always "sounds the same", why the need of choice one (pro) setup instead of another (pro) one? That's always the question.

If the content you extrapolate its not exactly the scomposition of physics elements, but reside on a high level of interpretation, choose 10 different PRO loudspeakers of listening should give to me the same. As well using headphone instead of monitor/loudspeaker. The content should stay above the sound (which become a medium).

Instead, it seems I'm really experiences different things listening a bass on a "smiley" eq system instead of a dump-higher one. I'wouldn't call a speaker choice a "preference": because the way it distort the signal actually shapes the whole timbre.

Maybe what confuse me is a side-by-side comparison instead of a long-term perception? Not sure... so I've tried an experiment: I record the same FM synth bass+kick with a different-final EQ curve (as emulating different speakers frequency response; not a different master), and play it randomly across some days, without knowing which is one version (a sort of AB test). I always catch which version was, because of the way it shape the partials, thus the timbre. And the perception change every time. I reckon is the same bass+kick "class", yes, but the final result (i.e. the overall sound, thus the perception) its different, one is "heavier", triggering different emotions to me.

Really not able to fit to the concept of "sounds the same"

As soon as one engages introspection to analyse consciously the workings of extremely complex sub-conscious mechanisms ( in this case our auditory goings on) we will often come a cropper.

I see your point. But if you try to define how things works, realising that "we got the same", the question above stay the same, without analyse yourself: why should I buy setup A instead of B? In the end, both should give "the same"...

If the task is create a product that will always create the same "illusion" (i.e. a fixed audio subjective experience), why switch and choose across BILLIONS of different setups? I don't believe its just "marketing" hehe

 

[Cosmik]

I know we focus each time on different details, but when we focus on detail X, I should perceive the same, every time.

This assumes that human hearing, mood, consciousness is deterministic and repeatable. I don't believe that, myself, at many levels. It's like saying that beans should always taste the same whether or not you have just eaten hot chillies beforehand. Or that food doesn't taste better when you're hungry, etc. It's one reason why you shouldn't continuously modify your recipes based on moment-by-moment feedback from your own perception of the taste!

 

[Nowhk]

This assumes that human hearing, mood, consciousness is deterministic and repeatable. I don't believe that, myself, at many levels. It's like saying that beans should always taste the same whether or not you have just eaten hot chillies beforehand. Or that food doesn't taste better when you're hungry, etc. It's one reason why you shouldn't continuously modify your recipes based on moment-by-moment feedback from your own perception of the taste!

I totally agree with this. Which basically put out the "it sounds the same" concept.

This way of thing of course hook to my first part of claim, which is "things change constantly" during the time (due to setup, but also mood, prev experience, and such). I.e. timbre differs each time.

But before go further on my real "question", I'd like that most of them agree with this hehehe, so we can be all aligned

 

[Cosmik]

I think one angle on this is that audio people are fixated on the frequency domain view of sound. Because of this, they begin to imagine that hearing is literally the same as a Fourier transform. They gamely attempt to accommodate transients in this view too.

But if our appreciation of music is at an 'object' level, then the Fourier aspect may just be 'the front end', a mathematical simplification and a useful engineering tool. If we look at a statue, we appreciate it at 'object' level. We appreciate its form in sunlight, fog, snow, dusk. If it gets a bit chipped or someone sprays it a different colour, the 'object' survives at some level. We can even understand it if its scale is changed.

 

[Nowhk]

I think one angle on this is that audio people are fixated on the frequency domain view of sound. Because of this, they begin to imagine that hearing is literally the same as a Fourier transform. They gamely attempt to accommodate transients in this view too.

But if our appreciation of music is at an 'object' level, then the Fourier aspect may just be 'the front end', a mathematical simplification and a useful engineering tool. If we look at a statue, we appreciate it at 'object' level. We appreciate its form in sunlight, fog, snow, dusk. If it gets a bit chipped or someone sprays it a different colour, the 'object' survives at some level. We can even understand it if its scale is changed.

True, you are right! I can also agree with this.

But, again, what's the reason of prefer one setup instead of another, if not for the color its add to it? Its like to say "I prefer listen to X1 via Y1 because of the Z1 it adss", but in the same time saying "Z1 is not important for the underlaying content that X1 will transmit". So my question persist: why do you choose Y1 instead of Y2 if the color it adds is irrelevant?

 

[RayDunzl]

I think one angle on this is that audio people are fixated on the frequency domain view of sound.

Which yardsticks* (if any) would you propose for us amateur Measurers and Adjusters to use to investigate our audio environments?

*meterstick may be substituted in some locales

 

[Frank Dernie]

I have a simplistic view of timbre.
I see the timbre of a musical instrument as the combination of overtones. The combination of overtones have fairly strongly differing relative amplitudes between instruments, and still subtle differences between different examples of the same instrument - enough one has to imagine to justify at least part of the difference in monetary value between a Stradivarius and well made modern instruments.

In my mind hifi equipment does influence the timbre of instruments since it adds harmonic distortion, so the timbre experienced due to the addition of second or third harmonics etc. to the fundamental will certainly be changed by any addition due to the hifi being used.
At what level is this audible? Don't know.
At what level is this obvious/ unpleasant? I don't know either, but I have always been satisfied in my own mind that it is the timbre of instruments which is changed by a hifi, and if the person listening is not familiar with the instrument being played it is entirely possible for them to believe significant distortion is revealing a more pleasant timbre when in fact it is creating a timbre they prefer over the real one.
My view is this is the reason so many amplifiers with high distortion levels are still popular in reviews.

 

[RayDunzl]

But, again, what's the reason of prefer one setup instead of another, if not for the color its add to it?

Having colored my electrostats and cones and domes to emit comparable spectra, I prefer, when critically listening, the imaginary image that the electrostats project, as opposed to the imaginary image the cones and domes spray, into my untreated sidewall and ceiling room space.

I think I measure about a 10dB direct-to-reflected sound energy advantage with the dipole panels.

 

[RayDunzl]

At what level is this audible?

Around 1% for pure tones (maybe). That's a -40dB level (relative to the fundamental) for a single harmonic.

---

Here's an experiment at cancelling harmonic distortion - https://audiosciencereview.com/forum/index.php?threads/distortion-cancellation-experiment.1843/

Pure tones make harmonic distortion rather easily heard, it is hidden (from direct observation) with music, but (to me) is noticeable as (searches for words) something like noise obscuring clarity.

---

Harmonic distortion comparison of ML and JBL: https://audiosciencereview.com/forum/index.php?threads/jbl-lsr-308-in-the-house.1066/#post-28025

 

[Frank Dernie]

Around 1% for pure tones (maybe). That's a -40dB level (relative to the fundamental) for a single harmonic.

---

Here's an experiment at cancelling harmonic distortion - https://audiosciencereview.com/forum/index.php?threads/distortion-cancellation-experiment.1843/

Pure tones make harmonic distortion rather easily heard, it is hidden (from direct observation) with music, but (to me) is noticeable as (searches for words) something like noise obscuring clarity.

---

Harmonic distortion comparison of ML and JBL: https://audiosciencereview.com/forum/index.php?threads/jbl-lsr-308-in-the-house.1066/#post-28025

I have heard before that it is easier to hear on pure tones but I wonder on a single instrument, like a 'cello playing a Bach suite, whether one may hear a change in actual timbre but not be aware it is being created by the system.

Edit, and thanks for the links - can't believe I missed them when new!
second edit - I've sussed it, I was on holiday away from internet connection when the first thread took place.

 

[Cosmik]

True, you are right! I can also agree with this.

But, again, what's the reason of prefer one setup instead of another, if not for the color its add to it? Its like to say "I prefer listen to X1 via Y1 because of the Z1 it adss", but in the same time saying "Z1 is not important for the underlaying content that X1 will transmit". So my question persist: why do you choose Y1 instead of Y2 if the color it adds is irrelevant?

Maybe overdoing the analogies and metaphors, what I get from audio systems are, in order of merit:

1. A 3D scene and clear separation between the 'objects' - it's a thrilling illusion. Dynamics can be visceral.
2. A beautiful 2D video sequence on a large hi-res screen - I can appreciate the content, but it's not as stunning as (1)
3. A workmanlike 2D video representation on a 22" screen - if I really love the 'art' I can get something from it
4. An understandable representation of the art but it's small and rough and just tells me that I might like to investigate it in better quality

Preferring (1) over (2) may be the curse of the audiophile. Maybe (1) doesn't really give much more enjoyment than (2) but it's a great trick to pull off!