The current state of loudspeaker design puts a lot of emphasis on flat frequency response and other level measurements for sound reproduction. This methodology addresses the issue of precision since it allows for a measured comparison of different products and designs.
However, these measurements alone do NOT address the issue of accuracy. While it may appear that way, a speaker that measures flatter on the frequency response may not be more accurate than one that is not when used in the context that they are meant for: listening by people.
Why is this? When measured by a machine, the data is interpreted without prejudice and therefore can be precise and accurate at the same time.
When the data (sound from the speaker) is passed to the human and interpreted, physical prejudice is unavoidable. This is not speaking of preference for sound or music style but rather the physical differences between the hearing of individuals.
As an example let's take person A and person B. Person A has a theoretical "perfect" hearing or a level volume of hearing throughout each and every frequency. Person B has a realistic hearing with volume loss below 30htz and above 19khtz and various minor in scale volume losses in between. When the same exact frequency sweep is heard by these 2 people, the result is different. They interpreted data has the same precision but different accuracy. The same data is outputted but different data is received.
The relevance of this information is extremely useful to determine the future state of audio reproduction. With the increase ease and availability of active/digital crossovers and DSP, it isn't far until the day when each user of an audio system has their own "hearing profile" embedded into the design of the speaker. In this type of system, each individual listener will be able to hear the data (music, movies, etc) reproduced precisely AND accurately. Add this type of implementation to the recording and mastering phase and you will be able to experience the "art" of the artist as they intended. That isn't to say this type of system doesn't have challenges of it's own: aka when 2 or more people listen at the same time. Perhaps algorithms to balance the hearing of the 2 individuals?
I am actually working on such a system with my significant other. We have taken detailed hearing measurements of our hearing and my next DIY project will involve a digital crossover to account for these hearing differences. Once it's done, I'll post a full write up of the process and our interpreted results
However, these measurements alone do NOT address the issue of accuracy. While it may appear that way, a speaker that measures flatter on the frequency response may not be more accurate than one that is not when used in the context that they are meant for: listening by people.
Why is this? When measured by a machine, the data is interpreted without prejudice and therefore can be precise and accurate at the same time.
When the data (sound from the speaker) is passed to the human and interpreted, physical prejudice is unavoidable. This is not speaking of preference for sound or music style but rather the physical differences between the hearing of individuals.
As an example let's take person A and person B. Person A has a theoretical "perfect" hearing or a level volume of hearing throughout each and every frequency. Person B has a realistic hearing with volume loss below 30htz and above 19khtz and various minor in scale volume losses in between. When the same exact frequency sweep is heard by these 2 people, the result is different. They interpreted data has the same precision but different accuracy. The same data is outputted but different data is received.
The relevance of this information is extremely useful to determine the future state of audio reproduction. With the increase ease and availability of active/digital crossovers and DSP, it isn't far until the day when each user of an audio system has their own "hearing profile" embedded into the design of the speaker. In this type of system, each individual listener will be able to hear the data (music, movies, etc) reproduced precisely AND accurately. Add this type of implementation to the recording and mastering phase and you will be able to experience the "art" of the artist as they intended. That isn't to say this type of system doesn't have challenges of it's own: aka when 2 or more people listen at the same time. Perhaps algorithms to balance the hearing of the 2 individuals?
I am actually working on such a system with my significant other. We have taken detailed hearing measurements of our hearing and my next DIY project will involve a digital crossover to account for these hearing differences. Once it's done, I'll post a full write up of the process and our interpreted results
Sounds like a bad idea to me. We hear things everyday with our ears. The source doesn't change because we hear it.😉
dangwei,
What makes you think that perception of a sound event is different just because there are individual differences in low level hearing processes? I'm not saying this isn't the case (hearing aids do work) but how do you plan to test for perceptual differences?
What makes you think that perception of a sound event is different just because there are individual differences in low level hearing processes? I'm not saying this isn't the case (hearing aids do work) but how do you plan to test for perceptual differences?
Optimum music reproduction can be achieved in only one way
There is no other way.
DSP and similar are of no help, indeed they are evil
Lately I've been talking about suspending the speakers with rubber bands...
Also OB-no box might be a good place to start
Indeed we're talking about the system, where the human being is the terminal...
There is no other way.
DSP and similar are of no help, indeed they are evil
Lately I've been talking about suspending the speakers with rubber bands...
Also OB-no box might be a good place to start
Indeed we're talking about the system, where the human being is the terminal...
The bandwidth starts from very wide at childhood to restricted in adult life. Typically it progressively gets worse as we age. Every 5 years there would be some significant change. Above the age of 50 the hearing could have dropped off a lot. Some can hear up to 16 Khz but many can't hear over 12Khz or even less. Very small kids ( under 3 I guess!) supposedly can hear up to 30Khz ! But this rapidly drops as they age. But no two people ( or even the two ears in one person ) loose the bandwidth the same way. Some can hear HF better than others in old age.
So making a system tuned to each persons ear doesn't seem workable or even worth while. Many with restricted bandwidth enjoy the music today just as well as they did in their younger days. Maybe we should just focus on making reproducing systems which make 'musical' sounds and of course spend more time making the 'music' itself !😉
So making a system tuned to each persons ear doesn't seem workable or even worth while. Many with restricted bandwidth enjoy the music today just as well as they did in their younger days. Maybe we should just focus on making reproducing systems which make 'musical' sounds and of course spend more time making the 'music' itself !😉
Yes, it's all about band-width ...also time and space...we see and fell only a fraction of a fraction. That's why the system should be conceived as a full
pass-band device 😀
The masking of the sound made by the sound itself ...it's the most fatiguing issue to fight. As music cannot live without the sound, and viceversa, the sound is ancillary to the music.
speaking about music reproduction, well, that creates a real black hole in time and space ! It's a breaktrough in man's mind...it makes happen something happened in another timer, another place .
pass-band device 😀
The masking of the sound made by the sound itself ...it's the most fatiguing issue to fight. As music cannot live without the sound, and viceversa, the sound is ancillary to the music.
speaking about music reproduction, well, that creates a real black hole in time and space ! It's a breaktrough in man's mind...it makes happen something happened in another timer, another place .
Last edited:
Hi Dangwei, an interesting but seriously flawed proposition.
I like classical music and I prefer it live - however I live in a rural area far from the nearest orchestra etc so I play my Vinyls and CDs. I want my music reproduction to be as close as possible to what I experience at the opera house.
After a lifetime of motor cycle helmet wind noise my hearing is seriously degraded. Because of this, what I experience at the opera house is unique to me, but it is what I regard as the truth. And it is more or less what I hear through my stereo.
I like classical music and I prefer it live - however I live in a rural area far from the nearest orchestra etc so I play my Vinyls and CDs. I want my music reproduction to be as close as possible to what I experience at the opera house.
After a lifetime of motor cycle helmet wind noise my hearing is seriously degraded. Because of this, what I experience at the opera house is unique to me, but it is what I regard as the truth. And it is more or less what I hear through my stereo.
Loudspeakers are not flat (or linear) because peoples hearing is flat. Loudspeakers are flat because they should reproduce sound exactly as it has been recorded.
Our hearing is everything else but linear. But that doesn't matter because our brain compensates for that - on an individual basis. If there is severe hearing loss, it should be compensated at the individual ear (by hearing aid or earphone), but not in a loudspeaker, when other people listen to it too.
Rudolf
Our hearing is everything else but linear. But that doesn't matter because our brain compensates for that - on an individual basis. If there is severe hearing loss, it should be compensated at the individual ear (by hearing aid or earphone), but not in a loudspeaker, when other people listen to it too.
Rudolf
Quite right Rudolf but compensation can be done at the speaker level. It just doesn't make a lot of sense. Anyway, speaker designer, recording/mixing/mastering engineers do just that, compensate for hearing loss they might have. Adds to Audio's Circle of Confusion
🙂 Thank you for the reply! The issue is: People's hearing is not flat...nor linear. Nor does our brain compensate for this. You cannot hear what you cannot hear. For example take a hearing test across the "common" 32 band EQ frequencies. You will discover that your audible perception between this range may differ up to 5 dbs from one end to the other. In the middle there may be differences of up to or more than 1db. For the individual frequency, an individual's hearing may be able to detect differences in volume of as little as .1 db.
A hearing aid is nothing but a loudspeaker placed closer to the ear. 😉
As for Marku's reply:
Many sound engineers do already compensate for their own hearing variances, at least the competent ones. However, they cannot account for variances in our hearing. Take myself for example. I have rolling off hearing loss after 18Khz and perhaps due to too much listening of jazz vocals, hearing differences of more than 2db between 400 and 800hz. However, I want to listen to the music that is intended by the recording. As of now, there is no hearing aid or medical apparatus that can serve this function.
This idea is really not a viable for commercial or mass offering since the speaker maker cannot account for the variations of each buyer. However, in the area of DIY, this is where with the help of digital crossover, quite simple to accomplish. I dare not say I know how the brain works and hence the only layout the proposition. What I can however say for sure is that when speaking of accuracy and precision, our ears do add inconsistency and distortion to the chain
Last edited:
The frequency is expressed in Hertz, Hz ( capital, it's a name )
and Kilo ( X 1000 ) is capital, too .
The last proposition is simply absurd 🙄
and Kilo ( X 1000 ) is capital, too .
The last proposition is simply absurd 🙄
I really like this reply. It shows that there is no one answer when it comes to music! In your viewpoint, you are are looking for the reproduction of your music and what you heard. In my viewpoint, I want to reproduce the music the artist created and hears/intends for us to hear. Of course this isn't an issue of correct and incorrect.
In relation to the original post, to you my view is precise as the points relate to the same topic but to you, inaccurate as they all hit very far from where your intentions lie. Hence the difference between precision and accuracy.
Thanks! This is great reading! For those that do not know or lazy to read: The El Greco Fallacy summarized is that:
The subjects of painter El Greco's paintings are elongated. It is then suggested that since El Greco suffers from astigmatism, his perception is visually narrow and hence the reason for the elongated subjects in his paintings. The flaw or fallacy in this case is that if El Greco suffers from astigmatism, his view of his canvas and what he paints will also be narrower and hence cancel out his visual perception.
I don't know if this applies to music "reproduction" since I am neither the artist, recorder nor mixing engineer. As a DIY speaker maker, my goal is to recreate the sound as true as possible to what the artist hears (the recorder and mixing engineer as well), and if it means my hearing is vastly different from his, to correct the playback system to adjust for my personal hearing variance.
If the variance in hearing between individual listeners is so large, then there is no point in trying to hear through the artists ears. The currently accepted means to preserve accuracy or precision is the best that we are going to get. Van Gogh was another good example, i do not want to see through his eyes, indeed i do not have to, i already see his perception of the world in his style. Altering my view to see the same view as he did is altering my perception by distorting the original medium.
A bit like going to a live concert, sitting in your preferred row, then recording the same concert from a different row and doctoring the recording so that it sounds like the seat you sat in during the concert. A bit pointless, in my opinion.
A bit like going to a live concert, sitting in your preferred row, then recording the same concert from a different row and doctoring the recording so that it sounds like the seat you sat in during the concert. A bit pointless, in my opinion.
Last edited:
Are you sure you're talking about a proper hearing test and not equal-loudness contours?
Do you see the white triangle?
Although you know that there really is no white triangle, you still see it. Your conscious mind can't overrule subconscious processes.
Attachments
- Status
- Not open for further replies.