Those of you who believe measurements aren't the whole story, do you have a hypothesis why that is?

[Killingbeans]

Right now I've spent much more time listening to stereo rather than studying the science of it. I'm not the only one that finds the midrange sounds much louder than the fq's above and below even tho' each tone burst measures 75 db? That tells me the most natural sounding speaker would indeed have the BBC dip or smiley face curve to compensate for our hearing yet speakers that do this and get favorable reviews also get slammed by the science community?

Well... none of the naturally generated sounds around you, including live music, get adjusted with a smiley face curve.

Compensating for the varying senitivity of human hearing across the spectrum might make it easiers to judge relative differences in volume, and I can only guess that's the reason why the BBC chose that curve for their monitors. But it's not natural. Enjoyable maybe, but not natural.

 

[steve59]

Well... none of the naturally generated sounds around you, including live music, get adjusted with a smiley face curve.

Compensating for the varying senitivity of human hearing across the spectrum might make it easiers to judge relative differences in volume, and I can only guess that's the reason why the BBC chose that curve for their monitors. But it's not natural. Enjoyable maybe, but not natural.

I have a second system that does have a midrange dip. It causes me to turn up the volume that in turn produces a less forward sound with better detail at the fq extremes. I’ve never considered opening the db meter app on my phone at a concert but your reply does make me wonder.

 

[Wanman]

And since our meausuring devices are far more sensitive than average human hearing.

In terms of, say, for example, HD/IMD caused by an unchanging slightly curved transfer function, and or something like noise floor? Asking because those are some types of things that can fairly easily be measured with spectral analysis.


OTOH, and this is just my personal viewpoint, not sure if anyone has ever measured some other things such as for example, close-in phase noise convolved with audio coming out of a sigma-delta dac. Its something that can sort of be visualized using spectral analysis, but its not a quantitative measurement that way. Since to the best of my understanding its hard to measure quantitatively, its hard for me to see in a strict scientific sense if it can be shown to be audible or not to the average ear (the usual thresholds being in relation to "the average ear").

 

[MarkS]

My own personal standard is that "audibility" first needs to be established by ABX blind listening tests. I want there to be a listener who can routinely and repeatedly get it right 90% of the time. Then and only then will I believe there is something "audible".

And then we can look for measured differences between A and B. I very confidently predict that differences would be easily found.

 

[KLi]

I have a maybe naive question: Say, you have a WAV-file oft a recording. One piano and a saxophon playing side by side. Is it possible to create a digital image on screen to show where they stood while the recording was made? A literal image?

 

[30 Ounce]

I have a maybe naive question: Say, you have a WAV-file oft a recording. One piano and a saxophon playing side by side. Is it possible to create a digital image on screen to show where they stood while the recording was made? A literal image?

Go to about 3:40…he does almost exactly what your asking.

 

[kemmler3D]

Is it possible to create a digital image on screen to show where they stood while the recording was made? A literal image?

A vectorscope is an analyzer they use in studios to look at stereo image, although it's not a picture of a piano and saxophone placed along a left-right scale per se, it does show you that visually.

(this is a mac app that costs $3)

 

[KLi]

Thanks guys @kemmler3D and @30 Ounce . Very very interesting. This leaves even less room for the esoteric crowd indeed. Have a nice sunday!