Here’s the problem with measuring hi-fi performance

[Blackmorec]

How many times have you read posts on forums that demand measurements to back-up claims. How many times have you seen manufacturer’s claims rubbished due to the lack of scientific evidence (measurements).

But here’s the problem. Most of what we ‘hear’ in audio is based on our psychoacoustic abilities to ‘differentiate’ the audio pressure waves reaching each ear based on differential time, phase and amplitude. Our ability to judge musical correctness is, amongst other things based on our evaluation of very specific frequency spectra and content. Does a cello sound real? Musically speaking we’re asking if we can hear the strings vibrate, with the right frequencies and harmonics, the wooden instrument body vibrate with its typical frequencies and harmonics etc. What we’re evaluating is the spectrum and decay of a host of frequencies, whilst simultaneously evaluating their time, phase and amplitude differentials.
So let’s say that we make some change to our hi-fi that increases sound stage focus and depth, adds sparkle to the treble and extra body to the bass, such that a cello now sounds more ‘lifelike’ and more focused. So the question is , what would you measure that correlates with the above and what electronic test gear would you use to do the measurement?

 

[Kal Rubinson]

So let’s say that we make some change to our hi-fi that increases sound stage focus and depth, adds sparkle to the treble and extra body to the bass, such that a cello now sounds more ‘lifelike’ and more focused. So the question is , what would you measure that correlates with the above and what electronic test gear would you use to do the measurement?

It would depend on what change you made. For example, if you used a graphic or parametric EQ, measuring FR would do.

 

[microstrip]

How many times have you read posts on forums that demand measurements to back-up claims. How many times have you seen manufacturer’s claims rubbished due to the lack of scientific evidence (measurements).

But here’s the problem. Most of what we ‘hear’ in audio is based on our psychoacoustic abilities to ‘differentiate’ the audio pressure waves reaching each ear based on differential time, phase and amplitude. Our ability to judge musical correctness is, amongst other things based on our evaluation of very specific frequency spectra and content. Does a cello sound real? Musically speaking we’re asking if we can hear the strings vibrate, with the right frequencies and harmonics, the wooden instrument body vibrate with its typical frequencies and harmonics etc. What we’re evaluating is the spectrum and decay of a host of frequencies, whilst simultaneously evaluating their time, phase and amplitude differentials.
So let’s say that we make some change to our hi-fi that increases sound stage focus and depth, adds sparkle to the treble and extra body to the bass, such that a cello now sounds more ‘lifelike’ and more focused. So the question is , what would you measure that correlates with the above and what electronic test gear would you use to do the measurement?

Proper measurements are extremely useful when assembling a system. However they must be considered as an whole, not as individual medicine for specific purposes. The first thing that I would measure for solving such problem would be the room decay times and the averaged frequency response of the system . All you would need would be a microphone such as the ECM8000, a decent soundcard and a PC running ROON or equivalent. IMHO no measurements on electronics or cables can help on the fine tuning of the system tweaks - I am assuming that acoustics is not a tweak affair!

 

[stehno]

How many times have you read posts on forums that demand measurements to back-up claims. How many times have you seen manufacturer’s claims rubbished due to the lack of scientific evidence (measurements).

But here’s the problem. Most of what we ‘hear’ in audio is based on our psychoacoustic abilities to ‘differentiate’ the audio pressure waves reaching each ear based on differential time, phase and amplitude. Our ability to judge musical correctness is, amongst other things based on our evaluation of very specific frequency spectra and content. Does a cello sound real? Musically speaking we’re asking if we can hear the strings vibrate, with the right frequencies and harmonics, the wooden instrument body vibrate with its typical frequencies and harmonics etc. What we’re evaluating is the spectrum and decay of a host of frequencies, whilst simultaneously evaluating their time, phase and amplitude differentials.
So let’s say that we make some change to our hi-fi that increases sound stage focus and depth, adds sparkle to the treble and extra body to the bass, such that a cello now sounds more ‘lifelike’ and more focused. So the question is , what would you measure that correlates with the above and what electronic test gear would you use to do the measurement?

Aside from wpc, I pay zero attention to measurements. IMO, sensitive measuring instruments potentially face some of the exact same crippling distortions our sensitive components face and as a result I suspect all sensitive instruments of every type are potentially performing far closer to their base performance levels rather than their truly optimal performance levels.

As for discerning or interpreting what we here after we make a change, well that too is potentially a bit of a pandora's box in and of itself. That's why potential changes should be methodically considered and take potentially considerable time to fully appreciate the change. That's why shoot-outs don't really tell much of anything. Especially since some components don't perform their best right out of the starting gate.

Since they've yet to invent a measuring instrument to measure sound quality, we're left to our ears. And in this audio-only industry, there's no substitution for developing or improving our ability to discern / interpret what we hear. At least there shouldn't be.