Can speakers ever perfectly reproduce audio? - AVS Forum
Audio Theory, Setup, and Chat > Can speakers ever perfectly reproduce audio?
Tyrone Burton's Avatar Tyrone Burton 11:06 AM 06-27-2013
I've read where some people have done tests with speakers and actual bands to see if speakers have come close, and according to one I have read, he said no, not even solo piano music which he thinks even then is quite a hard a instrument for a speaker to properly reproduce.

But wouldn't the actual recording of audio also effect how "real" the speakers sounds? Such as sampling frequency, bit depth? But then again they could have used analog recordings. What do you guys think, can a speaker ever perfectly reproduce audio?

William's Avatar William 11:18 AM 06-27-2013
Easy answer, no. "Perfectly" can never be reproduce by a transducer.
Ethan Winer's Avatar Ethan Winer 12:48 PM 06-27-2013
Quote:
Originally Posted by Tyrone Burton View Post

can a speaker ever perfectly reproduce audio?

As William said, no transducers are perfect, but in a good acoustical environment you can come pretty close:

Recorded Realism

--Ethan
Tyrone Burton's Avatar Tyrone Burton 04:28 PM 06-27-2013
I see, I kind of suspected that anyway, but wasn't 100% sure. I'd imagine the same would apply to headphones too.
commsysman's Avatar commsysman 04:05 AM 06-28-2013
A piano IS incredibly hard to reproduce with any semblance of realism.

I have been upgrading my system over and over and over for more than 40 years with perfect piano reproduction as one of my standards, with limited success.

I have recently come very very close with the purchase of the Vandersteen Treo speakers.

They are simply amazing.

They are driven by a Audio Research LS-26 preamp and a Bryston 3B-SST2 amplifier, and my primary sources are a BDP-95 SACD/CD player and a Music Hall MMF-7 turntable with a Benz Ace cartridge.
William's Avatar William 04:51 AM 06-28-2013
Quote:
Originally Posted by Tyrone Burton View Post

I see, I kind of suspected that anyway, but wasn't 100% sure. I'd imagine the same would apply to headphones too.

Headphones are transducers (as are microphones) and actually present their own "special" realism problems. Like no crosstalk or reflected sound and always on axis.
Quote:
Originally Posted by Ethan Winer View Post

...but in a good acoustical environment you can come pretty close:

Recorded Realism

--Ethan

Missing from the headphone experience and can only be replaced by binaural recording or (debatable) DSP.
Tyrone Burton's Avatar Tyrone Burton 10:04 AM 06-28-2013
Kind of makes me want to find a good quality piano song and see how my speakers get along with it. My particular speakers are the Wharfedale Xarus 5000, which can be bi wired, or bi amped. Does bi amping help in any way as far as realism goes?
Ethan Winer's Avatar Ethan Winer 10:59 AM 06-28-2013
Quote:
Originally Posted by William View Post

Missing from the headphone experience and can only be replaced by binaural recording or (debatable) DSP.

Understand that more than half the problem is capturing the sound of the piano (or whatever) in the first place. That's where acoustics very much do apply. It's not so much that speakers or headphones can't reproduce accurately, because in fact they do pretty well. Good microphones also mostly work well, but where you put them is key! The piano is a large instrument that has many separate radiating source areas.

--Ethan
Gizmologist's Avatar Gizmologist 09:31 AM 06-29-2013
As different woods, glues and varnishes affect the sound of a violin to the point where one will sell for three hundred and a Stradivarius can easily go for over a million dollars, it seems rather apparent that any acoustical instrument could never be fully replicated by a paper cone and an electro- magnet.
Timothy91's Avatar Timothy91 10:11 AM 06-29-2013
When listening in a room, due to room interactions, that's a no. However, if the recording is done a very specific way, then with headphones, YES you can get perfect sound reproduction. Look into binaural recordings. Playing back these recordings in a room requires very specific setup, acoustic control conditions and even then, it's not as perfect as headphones.

So, your answer is, yes, provisionally, you can only come very close provided that the SOURCE RECORDING was done binaurally (no multi-mic/mixing consoles allowed). When you listen to a binaural recording, you will flip at the realism. It's basically 3D for your ears.
BassThatHz's Avatar BassThatHz 01:36 PM 07-05-2013
The problem is that sound doesn't travel in a straight line.
It travels as 3d spheres and point-rays, thus it is impossible to capture that information with one(or two) single-point microphone(s).

Also, speakers don't radiate a 360 degree sound field (of the same pattern) as musical instruments (or worse a collection of instruments).
[There is one hi-end speaker I know of that radiates in a 360 sound field, but can only do so: as a mono-symmetrical pattern.]

If you could solve those two problems AND could play that back with the same SPL and bandwidth with no added distortion, then in theory it should be 100%.

But because of room acoustics it would only be valid for playback in the same auditorium it was recorded in and not your home; but it would otherwise be very very close, far closer than current tech.

nano-goo air wink.gif
Timothy91's Avatar Timothy91 02:19 PM 07-06-2013
Quote:
Originally Posted by BassThatHz View Post

The problem is that sound doesn't travel in a straight line.
It travels as 3d spheres and point-rays, thus it is impossible to capture that information with one(or two) single-point microphone(s).)

This is incorrect. Sound is a series of spatially delayed and frequency altered energy. Two microphones placed within a biologically accurate pair of ears (with pinna allowing the sound to reflect and be captured by the microphone in each ear) is all you need to capture a HIGHLY ACCURATE rendition of how humans hear sound. You only have two ears, you only need two microphones, properly placed within a human hearing model-head to capture TRULY life-like sound.

Here, read up, it's fascinating:
http://www.kallbinauralaudio.com/a-brief-history-of-binaural-audio/
arnyk's Avatar arnyk 05:10 AM 07-07-2013
Quote:
Originally Posted by Tyrone Burton View Post

Kind of makes me want to find a good quality piano song and see how my speakers get along with it. My particular speakers are the Wharfedale Xarus 5000, which can be bi wired, or bi amped. Does bi amping help in any way as far as realism goes?

Biwiring and passive biamping have no audible effects.

If you want to change a regular speaker over to proper biamping, you have to re-engineer it.
arnyk's Avatar arnyk 05:13 AM 07-07-2013
Quote:
Originally Posted by Ethan Winer View Post

Quote:
Originally Posted by William View Post

Missing from the headphone experience and can only be replaced by binaural recording or (debatable) DSP.

Understand that more than half the problem is capturing the sound of the piano (or whatever) in the first place. That's where acoustics very much do apply. It's not so much that speakers or headphones can't reproduce accurately, because in fact they do pretty well. Good microphones also mostly work well, but where you put them is key! The piano is a large instrument that has many separate radiating source areas.

This is a key point. It isn't like recordings are perfect and then the speakers ruin them. It isn't like microphone feeds are perfect and then audio production ruins them. I've done recordings using some of the most idealized mics there are, and there is still no comparison with the mic feed and the live sound.
arnyk's Avatar arnyk 05:19 AM 07-07-2013
Quote:
Originally Posted by BassThatHz View Post

The problem is that sound doesn't travel in a straight line.

Sound does travel in a straight line unless it is refracted by passing between materials with different acoustic properties. Air does not have a perfectly consistent temperature and so there are slight amounts of refraction in the real world, but these are secondary effects.

It is definitely possible to make sound radiators where the sound does not radiate in a spherical pattern. In fact, building a perfectly spherical radiator can be challenging.
humbug2's Avatar humbug2 12:38 PM 07-07-2013
In addition to the above posts If you are listening to free standing speakers there is air between you ear and the speaker. Sound transmission will be affected by the density and the humidity. Depending upon the material your speaker cones are made of they may be affected by humidity.

There are many imperfect components between the imperfect microphones and your speakers. The answer to your question is absolutely not.
Timothy91's Avatar Timothy91 02:55 PM 07-24-2013
Quote:
Originally Posted by humbug2 View Post

In addition to the above posts If you are listening to free standing speakers there is air between you ear and the speaker. Sound transmission will be affected by the density and the humidity. Depending upon the material your speaker cones are made of they may be affected by humidity.

There are many imperfect components between the imperfect microphones and your speakers. The answer to your question is absolutely not.

When listening to well recorded binaural recordings (with dummy head and good mics), the playback in good earbuds is a highly convincing experience and the closest thing to reproducing 'perfect' sound the way the human ear perceives it. No other recording/playback method can do better unless the technology/engineer behind the dummy-head, microphones or headphones are improved further. Because it's a field of audio that is not garnering much interest from the audio community, then the advances aren't likely to continue in this genre. It happens to be the most promising audio reproduction model conceived of so far by science. Nothing else is as realistic or comes as close to "real" sound.
Michal - Poland's Avatar Michal - Poland 05:59 AM 08-02-2013
That is main problem what one calls the prefect reproduction of audio. I think we are living in a real world which is of course not perfect . Why should be repruduction perfect? I would advise to use therms 1.nonrecognizable compairing to reality instead of pefect. 2. due to lack of comparaison to reality I would suggest presenting to author impression of reality. In my praktice I experience impression of reality but I must note that I am never shure it is original reality...uufffff.... eek.gif But if all recordings are sounding real it seems to be possibile. Good test for my confidence is sound of rain which is common to hear and not so much variety of rains as for other nature.

I try to contribute to Your discussion by presenting my test of speaker reality based on my abilities to judge. First I do recording of any material transmitted by speaker with certified flat microphone placed in about 1-1,5 meter in listening vertical and horizontal axis (very important) . Then with the same recording stuff I do normal recording with cable of the same material from headphones output when speaker disconnected. This two recordings I compare listening them - usually through headphones. The point is it doesn't matter what changes are induced by this headphones ,what are deficiences of recording stuff, and what are sounds of material - in ideal speaker reproduction sound recorded through speaker should not vary from the onecoming by cable (considering that cable does not change much and mic is flat) except speaker sound will be reverbated by room. As this reverbation should be taken in account as no change. Method is quite usefull. Another simplified method is to switch listening with headphones from amplifier heaphones out and mic amplifier out. without recording.

If anyone try this I am sure he will be shocked as me with differences and will not deliberate about sound of cables. . Next shock is awaiting when it seems - oh just change this and this. No way!biggrin.gif

Ah there is one more - even with ideal reproduction full reality may be experienced only out of space where speakers are transmitting. This I observe always.
Michal - Poland's Avatar Michal - Poland 06:23 AM 08-02-2013
Quote:
Originally Posted by Ethan Winer View Post

Understand that more than half the problem is capturing the sound of the piano (or whatever) in the first place. That's where acoustics very much do apply. It's not so much that speakers or headphones can't reproduce accurately, because in fact they do pretty well. Good microphones also mostly work well, but where you put them is key! The piano is a large instrument that has many separate radiating source areas.

--Ethan
But what about listening directly to piano? Of course sitting behind or at sidewall of instrument makes difference but should we say it is no reality of piano? Reality is something belonging to each position of mic according to instrument and to me it is not considering full vibration. Ok - we can listen to sound of wooden side wall of piano and it will be ideal wooden sidewall reproduction.. The problem shall only araise when someone listening to speaker will not accept it as piano complaining speakers.. But that is the art of understanding what we should hear and what no and when we can complain..
Michal - Poland's Avatar Michal - Poland 06:38 AM 08-02-2013
Quote:
Originally Posted by Gizmologist View Post

As different woods, glues and varnishes affect the sound of a violin to the point where one will sell for three hundred and a Stradivarius can easily go for over a million dollars, it seems rather apparent that any acoustical instrument could never be fully replicated by a paper cone and an electro- magnet.

Excuse me but I must mention that I had many years before widerange speaker only 5 Watts nominal assembled in labirynth driven with ultralinear tube amplifier and I may state that once listening to radio programm I recognized Stradivarius violin by beauty of tembre - I never had experience with violins but listening then to many violins before I got confidence : It has to be Stradivari. And speaker confirmed. I rememmber almost this sound till now as extremly deep, warm and energetic. So Your note is only these.
simp1yamazn's Avatar simp1yamazn 07:51 AM 08-02-2013
Quote:
Originally Posted by Timothy91 View Post

When listening in a room, due to room interactions, that's a no. However, if the recording is done a very specific way, then with headphones, YES you can get perfect sound reproduction. Look into binaural recordings. Playing back these recordings in a room requires very specific setup, acoustic control conditions and even then, it's not as perfect as headphones.

So, your answer is, yes, provisionally, you can only come very close provided that the SOURCE RECORDING was done binaurally (no multi-mic/mixing consoles allowed). When you listen to a binaural recording, you will flip at the realism. It's basically 3D for your ears.

You can certainly get great audio quality from a good set of headphones and I think everyone is aware that the quality and methods of the recording are of the utmost importance. That being said, we don't hear with our ears alone. Certain frequencies are felt as much as they are heard. Headphones simply can't reproduce the effect of a LF wave traveling through you.

I think "perfect" can be hard to define in this case. Are we talking perfect in the sense that we get exactly what came out of the instrument? Exactly what came out of the mains at a concert? Say at a live concert the lead guitarist is playing guitar X, through pre-amp y, and through amp z (can you tell I know nothing about guitars??). What the guitar player hears is different from what the engineer hears which is different from what the people in the front row hear, which is different from what people in the back row hear.

So how do you define perfection?
arnyk's Avatar arnyk 12:48 PM 08-02-2013
Quote:
Originally Posted by Timothy91 View Post

Quote:
Originally Posted by humbug2 View Post

In addition to the above posts If you are listening to free standing speakers there is air between you ear and the speaker. Sound transmission will be affected by the density and the humidity. Depending upon the material your speaker cones are made of they may be affected by humidity.

There are many imperfect components between the imperfect microphones and your speakers. The answer to your question is absolutely not.

When listening to well recorded binaural recordings (with dummy head and good mics), the playback in good earbuds is a highly convincing experience and the closest thing to reproducing 'perfect' sound the way the human ear perceives it. No other recording/playback method can do better unless the technology/engineer behind the dummy-head, microphones or headphones are improved further. Because it's a field of audio that is not garnering much interest from the audio community, then the advances aren't likely to continue in this genre. It happens to be the most promising audio reproduction model conceived of so far by science. Nothing else is as realistic or comes as close to "real" sound.

IME the possibility of perfect reproduction from speakers is already lost at the output of the microphone. I agree that a pair of good headphones driven by a high quality binaural head is probably about as good as it gets, but it still isn't perfect.
Ivan Beaver's Avatar Ivan Beaver 06:07 PM 08-04-2013
Quote:
Originally Posted by arnyk View Post

IME the possibility of perfect reproduction from speakers is already lost at the output of the microphone. I agree that a pair of good headphones driven by a high quality binaural head is probably about as good as it gets, but it still isn't perfect.
I have not read the whole thread-so forgive me if this has been already covered.

I guess it comes down to what one considers "perfect".

If "perfect" means that you can't tell the difference between a speaker reproducing a signal and the original signal-consider this.

Let's assume the rest of the chain is "perfect"-mic-recording medium-playback electronics and you have a loudspeaker that has a single source as the reproducer and a totally flat amplitude and phase response.

Let's take a couple of different instruments and have them play the same freq/note-a trumpet- a flute- a violin and a electric guitar.

The problem starts in that each of these instruments produces the same freq-but in different ways. so the polar pattern of each instrument is different. Yet the loudspeaker will have a single polar pattern-that probably is not anything like any of the instruments.

So the "sound" in the room (as it bounces off of different surfaces-due to the different polar patterns) will be different. So it simply cannot be "perfect" or a totally accurate reproducer-even if everything is "perfect"-it is not the same.

I agree that good quality headphones are the best bet-at least to reproduce what the mic originally heard-if for no other reason that they remove the listeners room from the equation,
mraub's Avatar mraub 07:23 PM 08-04-2013
The poster who claimed he could identify a Stradivarius violin through speakers, should take a look at this study, in which professionals could not identify them in live performances:

http://blogs.discovermagazine.com/notrocketscience/2012/01/02/violinists-cant-tell-the-difference-between-stradivarius-violins-and-new-ones/
Michal - Poland's Avatar Michal - Poland 03:26 PM 08-06-2013
Quote:
Originally Posted by mraub View Post

The poster who claimed he could identify a Stradivarius violin through speakers, should take a look at this study, in which professionals could not identify them in live performances:

http://blogs.discovermagazine.com/notrocketscience/2012/01/02/violinists-cant-tell-the-difference-between-stradivarius-violins-and-new-ones/

I do not want to argue with You that in mentioned there conditions Stradivari had no primacy. I just accept stated there facts But if few people called proffesionals by their way of earning money cannot or don't want observe something, then it is impossiblie to see for all others in concert conditions? I have a friend who doubts if Bentley/Rolls Royce is any better than his Skoda - Octavia. May be he cannot see difference same as many other Skoda owners. And if I see. then who is wrong? Was I unfair to recognize this Stradivari? Shall it be unfair or lie if I tell you that I cuould hear mechanical watch on speaker hands, steps of solists in opera hall?

Point is that what descreens best instruments from other is beautifull responding to musician intension - it supposed to be a dialogue or duo even. In bad hands nothing is amazing . For simple technical testing power of Stradivari can be hidden. Why not? - It is only wooden chamber with any strings and shines only in best hands in concert halls.. What can I say - You just stated that I shouild rethink. I will not because it took place about 40 years ago - for me too long time had passed, no other evidence happened What to do? . .

By the way: Is Stainway&Sons the also the same as others? Was somwhere a blind test to confirm? Of course no problem with Stainway - thay can still fight such opinions by law ..
Rgds Michal
arnyk's Avatar arnyk 07:01 AM 08-07-2013
Quote:
Originally Posted by mraub View Post

The poster who claimed he could identify a Stradivarius violin through speakers, should take a look at this study, in which professionals could not identify them in live performances:

http://blogs.discovermagazine.com/notrocketscience/2012/01/02/violinists-cant-tell-the-difference-between-stradivarius-violins-and-new-ones/

One big difference is that these were DBTs - true blind tests.

"The test was a true “double-blind” one, as neither the players nor the people who gave them the violins had any way of knowing which instrument was which."
8mile13's Avatar 8mile13 03:27 PM 08-08-2013
There is a double blind taste test for coca cola vs. pepsi. Eventhough, in my opinion, everybody should be able to tell the difference, nobody had a 100% score. Such a test tells me that certain senses (or sub-senses, whatever) are blocked because, in my opinion, were people should be able to tell the differences, they could not tell the difference (between coca cola and pepsi). It reminds me of blindfolding a person and turning them around several times which results in disorientation. I am pretty sure that there is some form/degree of disorientation and insecurity, since people are cut of from their usual way of experiencing things, going on in every blind test/double blind test. I am not saying here that blind tests or double blind tests is nonsense, what i am saying here is that things are not as simple as some folks here present them wink.gif

http://www.doublefine.com/news/comments/coke_vs._pepsi_vs._humans_at_double_fine/
JHAz's Avatar JHAz 03:43 PM 08-08-2013
Quote:
Originally Posted by 8mile13 View Post

There is a double blind taste test for coca cola vs. pepsi. Eventhough, in my opinion, everybody should be able to tell the difference, nobody had a 100% score. Such a test tells me that certain senses (or sub-senses, whatever) are blocked because, in my opinion, were people should be able to tell the differences, they could not tell the difference (between coca cola and pepsi). It reminds me of blindfolding a person and turning them around several times which results in disorientation. I am pretty sure that there is some form/degree of disorientation and insecurity, since people are cut of from their usual way of experiencing things, going on in every blind test/double blind test. I am not saying here that blind tests or double blind tests is nonsense, what i am saying here is that things are not as simple as some folks here present them wink.gif

http://www.doublefine.com/news/comments/coke_vs._pepsi_vs._humans_at_double_fine/

The idea that the experience is so different as to make the test invalid would, I suppose, potentially be relevant to the pepsi v coke test if the testees had never had a sip of coke or pepsi from a cup that didn't have a brand name emblazoned upon it. Otherwise, there's just not that big a difference. Nobody is blindfolded in a blind test. They just can't see which is which.

Skepticism being the bedrock of science, being skeptical about the results is legitimate. But when the science doesn't align with our "feelings" about how it should have come out, it might be time to assess whether our feelings are accurate as we might think they are.
8mile13's Avatar 8mile13 09:56 AM 08-09-2013
Quote:
Originally Posted by JHAz 

Skepticism being the bedrock of science, being skeptical about the results is legitimate. But when the science doesn't align with our "feelings" about how it should have come out, it might be time to assess whether our feelings are accurate as we might think they are.

Seems to me that is pretty clear that basically everybody can tell that pepsi cola and coca cola do not taste the same. In fact all participants of this double blind-taste test ''expressed confidence in their ability to distinguish brands.'' Nevertheless none of them got a 100% score and only one of the nine participants got four out of five.

There are all kinds of pepsi cola vs. coca cola studies with all kinds of results rolleyes.gif
Coke vs. Pepsi: The Pioneers
http://www.neatorama.com/2011/03/15/coke-vs-pepsi-the-pioneers/
Up
Mobile  Desktop