Impossible audio measurements

What kinds of audio measurements are impossible?

[odizcvw]

Great question and great answer. Sometimes soundstage is holographic in the most intriguing way. Sometimes you feel the emotion of the artist involuntarily. You don't know why, but you literally tear up when certain tracks play. Sometimes your toe taps involuntarily. What are the measured aspects that dictate these responses? I'd guess it has something to do with ability of low level information to come through with near perfect phase. Lot's of things can destroy it. But it does not seem to be called out by SOTA frequency response, distortion, slew rate or square wave perfection.

[NoEggu]

@1:54 "How do you measure all of that?"
Our brains (via our ears) measure all of that.

Note that the lemmings that worship Amir (the scientist impostor) claim that everything can be measured -- asserting if you can hear it, it can be measured.

Thank you, Paul, for addressing this measurement subject.

[PebblesChan]

LOL, the test bench LCD monitor says Yay!🎉

[freekwo]

In the lattest article from dCS for the Apex range, they are saying that they were had to create their own measurement tools because exsisting weren’t capable to measure the difference in sound that was probably audioable.

[spikeconley]

If it can be heard, it can be measured, in theory. At the end of the day, the human ears and brain are simply measurement devices. If you can hear something, that means sound waves are traveling to your ears. You can make a substitute for ears (for example a fake head with 2 microphones in it). Then it's just a question of putting some kind of algorithm on the soundwaves to figure out the difference. Is it worth creating? Maybe, maybe not.

The point is that the human mind can be tricked and biased, and a better analysis tool can remove that, in theory.

[sassafras]

“It’s so obvious when you hear it”. This one comment is key. If it’s obvious to listeners in double blind tests then you have science on your side. So easy to squash the “snake oil” accusation yet high-end audio manufacturers don’t do it. Perhaps it isn’t in their interests.

[joeythedime]

I agree with Paul the magic instrument sits on the left and right side of your head.

[howaboutataste]

I'm certain phase alignment can be measured. It is probably a situation similar to psychoacoustics, in that instrument measurments need to be translated into experience through well-described logarithmic transformation.
I really doubt there is much mystery left here, except that consumers don't read technical laboratory evaluations well, and manufacturers would MUCH prefer to sell on brand recognition and hype. Quantitatizion doesn't allow for profitablity the way psycho-emotional manipulation based marketing does.

[gotham]

This must have been filmed a while ago. Darren Myers has left PS Audio and now works for Parasound.

[PSA]

Audio is really interesting as it's a mix of hard and soft science, though some of the soft is probably just a matter of technology (or the combined use of several) that's not there yet, but will be in the future.

[ThinkingBetter]

Any audio company aiming at awesome audio fidelity needs to combine measurements with listening tests rigorously. Only ignorant people will think you can reach excellency with measurements alone or listening tests alone when it involves engineering of speakers and audio electronics.

[spacemissing]

In audio, one should Listen First,
then use test equipment to measure that which allows for it and use it mercilessly against the component being evaluated.

[biketech]

Sometimes I wonder if we've come all that far from the class A triode amplifier . More power, yes and lower measured distortion .

[tiffinytiffable]

I wonder if you could have an array of mics like a camera sensor in a standard size room and in standard listening locations. Have them record a standardized group of sound tones and timbres. These sounds are also evaluated by a board of audiophiles like a pier review group dedicated to the standardization of audio using this standardized room. We then correlate this panel's subjective experience with the empirical data from the mic array and this information becomes part of an audio standard ISO 20 - 20, 000. In time given enough data the community could approach some semblance of empirical audiophile attributes besides; open, airy, warm, cold etc. 🤓

[janinapalmer]

Good question 🤓😄😄..!
May I start by saying that soundstage and depth and transparency ALL start off with the recording !! If you can't get this bit right then it matters not WHAT the quality of the sound system reproducing it is ! I've heard a top quality recording streamed into a very mediocre amp and speakers .... and guess what ..🥹🤓..? It blew me away ... it was every bit as good as a high end system playing the mediocre recordings which the majority of us have to put up with !
The other end of the chain is our ears 👂... how good are they ... YES ..!! So often overlooked ... cleaning the wax build up in them is equally important as using a ribbon tweeter ...

[Scopolamine]

One piece of gear, or a different recording. Either one can make a substantial difference in transparency.

[noahbirdrevolution]

I want to see a $20k acoustic camera on a hifi system. Maybe more acoustic cameras than that would be needed. I don't have the funds.

[L.Scott_Music]

Seems to me that experienced listeners are the best tool for determining transparency and imaging. Sure, you get a subjective result but the results can be lessened by multiple listeners (even more so when they don't interact). After all, the final product is in the customers ears, so the final test equipment should be human ears.
This is what happens with recorded music. Mix engineers and mastering engineers are very good listeners. They are trained to hear details that others don't notice. I don't think a machine will ever be able to do that because machines aren't the customers.
Then there is the matter of A.I. That will be hit and miss until the A.I. has trained us to accept it's council. LOL. (shiver)

[howardskeivys]

Good musical reproduction is not all about straight line graphs and linearity. I think it’s more about striking a balance between transparency and harmonic richness or acoustic mass and those qualities countermand each other. Personally, I’m happy to sacrifice a little transparency in favour of tonal depth. Many audiophiles will lean the other way. That’s what keeps the audio industry, diverse and offering choices. There is no right and wrong, just personal preferences.

Enjoy the music and always remember music was! born analogue.

















CDC

[glenncurry]

With a degree in electronics, years of bench work, amp design, .... the biggest problem audio has in measurement is we are stuck in antiquated steady state sinewave measurements. And that has nothing directly to do with actual musical signals.

Back in the day it was even hard to produce a low distortion sinewave as a test signal! Impossible to capture and measure transient signals other than maybe clipping? In the video test world we at least had differential phase and gain measurements. Dynamic signal measurements. With audio at best THD, signals added harmonically related, IM, how two steady state signals interact, ... rise time/ slew ... of steady state squarewaves...

Even FFT's are based on accumulating information over a period of time to convert from time based to frequency based information. Which means starting at zero time means an infinite combination of frequencies. With actual frequency from time conversions requiring more time period of measurement for greater accuracy. So a single complex transient means nothing in FFT.

I was in highend test and measurement of video when digital hit that technology. Special time domain tests were invented. Digital test plates, zones, .... Other than TIM, audio has done little.