[71 dB]

There are two problems for people who suggest there is more to sound than just measurements. Firstly, measurements can't be developed to address those issues because those who make that claim have no logical/rational suggestions of what it is that needs to be measured. If they have any suggestions at all, it's typically for something which is not a property of sound but a property of human perception. Secondly, if there is some property to sound itself which we don't know about and can't measure then it's irrelevant any way! Let's say there is something and we do develop a way to measure it, we'd then need a new type of recording technology because the ones we've got can only record and reproduce amplitude and frequency. Bit of a catch 22!

G

For the most part I agree with this.

We can record and reproduce signals. Signals have properties such as amplitude, frequency, spectrum, decays time, phase etc.

Standards behind measurements are often children of their time. At some point audio technology might develop to a new level and new standards are needed.

 

[endgame]

Funny, I don't recall reading those words, nor anything even close in meaning in gergorio's post. I would add that a full measurement profile of any transducer would include amplitude, frequency and time.

Next it should be understood that measuring headphones in particular is difficult and prone to generating somewhat ambiguous data because, just looking at FR for now, position makes a huge difference, as does the specific artificial pinna and ear canal used in the measurement "head". Those things are based on averages of many humans, but that also means they are also a bit wrong for everyone. It's also really important to compare measurements of different headphones made by the same person with the same test setup. If you still have two "near identical" FR measurements, then we need to talk about what "near identical" is, and also look into the time domain and see what's going on past the initial arrival.

It's all there, it's all measured, there's nothing wrong with the data other than the difficulty of making the measurements consistent in the first place. The problem is and has always been drawing direct correlation of incomplete measurements and (unbiased) perception. Once perception bias is eliminated, and measurements are complete and accurate, the correlation is good.

This is a good explanation. Thank you.

 

[pinnahertz]

For the most part I agree with this.

We can record and reproduce signals. Signals have properties such as amplitude, frequency, spectrum, decays time, phase etc.

All aspects of any wave can be captured and crunched into the above (and much more) with a single measurement that captures enough data for the original to be reconstructed. In fact that is the basis of all modern audio measurements.

Standards behind measurements are often children of their time. At some point audio technology might develop to a new level and new standards are needed.

No, that's backwards. There are no standards behind measurements, there are standards based on measurements, but frankly when it comes to things like FR, THD and noise (you know, the audible stuff), there are no standards other than those of a specific application. When measured, they form figures of merit, typically reduced to simple single figures so they are easy to assimilate, but relatively meaningless (in terms of audibility) when they should be presented at least as 2 dimensional graphs if not 3d graphs.

THD for example is shown as a single figure. The, as an elaboration, expressed as a quantity vs frequency. Neither represents a degree of audible effect, which is a function of THD specific spectrum vs amplitude vs time in the presence of a masking signal. Nothing wrong with the measurements, though, they are dead-nuts on. Interpretation is hobbled by the need to simplify.

Please don't confuse measurement with specifications and standards, and don't confuse presentation of specs with anything that correlates to audibility. There IS correlation, but not when reduced for simplicity.

 

[bigshot]

Standards behind measurements are often children of their time. At some point audio technology might develop to a new level and new standards are needed.

If you look at the history of recorded sound as it relates to measurements, you'll find that most of the major standard measurements have been known about since the 1920s at Bell Labs. The technology of how to electronically do it has changed, but the physics of sound is very well understood. We've gotten to the point where our ability to record and play back are audibly perfect. ABX testing has proven that. We've also extended our ability to measure far beyond our ability to hear. If there's anything new to be added at this point, it's going to be in the way the technology is applied... recording techniques, signal processing, playback with multiple channels, etc... not in the way the fundamental principles of sound are measured. And as these new ways of recording, signal processing and playback are implemented, the concept of some sort of audible distortion of the sound being a bad thing will be replaced by using manipulation of the signal to improve sound.

 

[castleofargh]

For the most part I agree with this.

We can record and reproduce signals. Signals have properties such as amplitude, frequency, spectrum, decays time, phase etc.

Standards behind measurements are often children of their time. At some point audio technology might develop to a new level and new standards are needed.

new measurements or standards can always helps for specific things we wish to define, as in finding a new way to describe a specific variable, or struggling to make a fine crossover between physical sound and somebody's interpretation of it. mostly because we don't have a mind reader or the complete 3D scan of the listener. those issues have nothing to do with sound as a signal or as a physical wave.
but that's missing the point anyway. we can't be fine describing the entirety of a song in a simple 2 axis graph, 1 amplitude changing over time (X2 for stereo), and then pretend like that very signal is something super complicated we don't know how to characterize or measure. this is paradoxical.

 

[71 dB]

All aspects of any wave can be captured and crunched into the above (and much more) with a single measurement that captures enough data for the original to be reconstructed. In fact that is the basis of all modern audio measurements.

Impulse response describes fully a linear system, so yes (assuming the system is linear enough).

 

[TYATYA]

Interesting statement which if true means that you personally have a preference for lower fidelity, as you think it "sounds far better".

Yes, you said my mean ! There are difference in low range. Solid impact, quick cut off (less decay), detail. Those are a bit melting down on wav multichanel

 

[gregorio]

we can't be fine describing the entirety of a song in a simple 2 axis graph, 1 amplitude changing over time (X2 for stereo), and then pretend like that very signal is something super complicated we don't know how to characterize or measure. this is paradoxical.

The difficulty for many audiophiles so often comes down to an inability to understand or accept human perception. In English and most modern languages we have words which equate to sound, music and noise. The last two however are words used to describe what is a purely human perception/preference and are not things which actually exist. Therefore any discussion or description of music must be based on human perception. This is hardly a new development or point of only scientific interest, it's been a topic of considerable debate amongst those seriously interested in music for a couple of centuries, a debate which spills into the much wider general public occasionally. So how can some/many audiophiles not be aware of this? The only logical answer is that they have little interest in or understanding of music, despite statements to the contrary. Their interest and knowledge is in their perception/preference of audio reproduction equipment, not the audio itself (or it's "fidelity") and therefore these audiophiles are not, by definition, audiophiles. Paradoxical indeed!

Yes, you said my mean ! There are difference in low range. Solid impact, quick cut off (less decay), detail. Those are a bit melting down on wav multichanel

Sorry, I've no idea what you are trying to say.

G

 

[bigshot]

Music is organized sound. Noise is disorganized sound.

 

[gregorio]

Music is organized sound. Noise is disorganized sound.

That's a quite common definition of music which originates from Edgard Varese.but, while it works as a good explanation of his compositional goals it doesn't work as a definition of music. If it were true, then normal speech would be music, so would the sound of most car engines but an entire musical movement would not be music (Indeterminacy/Aleatoric) and neither would many/most of the compositions of a number of the C20th greatest composers (Cage, et al). Of all the attempts at a simple definition of music, I like Berio's the best: “Music is everything that one listens to with the intention of listening to music”.

G.

 

[71 dB]

Music is organized sound. Noise is disorganized sound.

A lot of people would say a lot of my favorite music is noise. I guess the concept of something being organized is subjective...

 

[RRod]

Where did you find DTS 5.1 and identically mastered 5.1 wav to compare?
Same question for DTS and AAC.

Perhaps he meant DTS core vs. the DTS-HD MA version, converted to WAV? I've never seen DTS and a true-blue 5.1 WAV track on the same disc.

Also, when comparing DTS to AC-3 one has to make sure that he has the default dynamic range compression off on the AC-3.

 

[bigshot]

Normal speech can be music if it's Shakespeare sonnets. But most people don't make much effort to organize their words.

 

[Strangelove424]

Shakespeare's a good example, as he usually wrote in iambic pentameter. I think some car engines are pretty musical too. I roll my windows down when certain cars pass by. And manufacturers tune engine and exhaust notes to create identifiable, often brand specific tones.

 

[pinnahertz]

Perhaps he meant DTS core vs. the DTS-HD MA version, converted to WAV? I've never seen DTS and a true-blue 5.1 WAV track on the same disc.

Also, when comparing DTS to AC-3 one has to make sure that he has the default dynamic range compression off on the AC-3.

Doesn't matter, the comparison is invalid and impossible because of potential changes in mastering.