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

[restorer-john]

This is also why I have been critical of arim's reviews in the past. Measuring frequency response isn't going to tell you about transparency/imaging and other characteristics of the sound.

I too, have been critical in the past in relation to loudspeakers, but the reality is we don't have an objective measurement or measuring system that can test for imaging, placement and stability over the spectrum. We can hear it, but a Klippel cannot.

Sure, a systems could be created that has two precisely placed speakers, two identically placed microphones and sophisticated signal processing to determine this, but you'd also need a giant anechoic chamber or a precisely treated standardized room. And then all the complaints would roll in about the 'room' not being representative blah blah.

 

[DonH56]

I don't know what "transparency" means. If it means adding nothing to the source, then frequency response and distortion measurements will suffice. For imaging, frequency response (amplitude and phase), and dispersion measurements (e.g. on- and off-axis response) will tell you, along with the in-room response since room reflections contribute greatly. Don't know what "staging" means; to me, setting up for a play or musical, or a sequence of operations in a process. We can't measure how it sounds or feels to you, but if the terms are defined, we can almost always measure them. But taking and interpreting the measurements can take a lot of equipment, experience, and so forth. Probably easier to just listen, and since the effects of he room are so important, probably more logical.

 

[restorer-john]

For imaging, frequency response (amplitude and phase), and dispersion measurements (e.g. on- and off-axis response) will tell you

With respect @DonH56 , I don't completely agree.

Imaging is a result of two speakers, not one. Imaging depends on a whole lot of things, not the least of which are the tolerances of the drivers and crossover components from speaker to speaker. An image that suddenly shifts, loses focus or recesses during playback. A vocalist that isn't positionally stable on one pair of speakers but is on another. You know what I mean.

 

[preload]

My understanding is that spin-o-rama analysis is something like 87% correlated with loudspeaker preference.

That would mean it explains (0.87 * 0.87) of the variance, i.e. 76%.

Thank you I was just about to chime in with same, you beat me to it.

 

[DonH56]

With respect @DonH56 , I don't completely agree.

Imaging is a result of two speakers, not one. Imaging depends on a whole lot of things, not the least of which are the tolerances of the drivers and crossover components from speaker to speaker. An image that suddenly shifts, loses focus or recesses during playback. A vocalist that isn't positionally stable on one pair of speakers but is on another. You know what I mean.

I don't think we are all that far apart, John. If you measure the amplitude and phase around the speaker (needed to include dispersion) then it will tell you when and where different frequencies will combine constructively and destructively. That should tell you when the image shifts. But, room reflections will affect all of that, as well as listener placement, the impact of things in the room (which can also reflect and absorb sound), and so forth. So, while it may be measurable, in practice it would be a daunting task and perhaps impractical in a real situation.

 

[dc655321]

But have you ever seen an anechoically flat measurement? Or a coherent slope in an echoic measurement? Literally everything I have seen looks somewhere between the Appalachians and the Rockies. Makes me wonder why we fetishize "flatness", when apparently it's not even remotely approachable.

It's not fetish, it's fidelity to the source.
Perhaps you've not seen the Genelec or Neumann kh80 measurements on this site? Or maybe you're not familiar with limitations of mechanical devices (vs electronic)?

Here are the steady-state in-room response curves from Toole's book:

It is not clear to me that ANY of these curves are intuitive.

Those are preference curves - there's a human in the mix. People are messy! I interpreted your comment as being transducer-centric...

An anechoically flat, non-omni speaker will look a lot like those curves (along some axis in a typical room) once adjusted for human non-linear hearing characteristics.

Love your avatar by the way. Creeps me out every time I see it!

Mission accomplished! As GW would say

 

[escksu]

I too, have been critical in the past in relation to loudspeakers, but the reality is we don't have an objective measurement or measuring system that can test for imaging, placement and stability over the spectrum. We can hear it, but a Klippel cannot.

Sure, a systems could be created that has two precisely placed speakers, two identically placed microphones and sophisticated signal processing to determine this, but you'd also need a giant anechoic chamber or a precisely treated standardized room. And then all the complaints would roll in about the 'room' not being representative blah blah.

Long ago, I did propose to quantify this. I believe with a good 3D waveform analysis software, it may be possible to derive characteristics of sound that causes us to have these perception. By repeatedly playing certain segments of an audio clip with various speakers (eg. one that has extremely strong imaging vs one that has hardly any imagining), I feel we may be able to observe a pattern and what constitutes to strong imaging. Then test this pattern with various speakers to test if this is repeatable. If we are successful, then we will know what kinds of wareform causes this.

It can even go to test other aspects. The infamous cables...lol... Granted if cables do not make a difference, then the audio produce should have no difference as well. This will be the first time we can really quantify if speaker cables do make a difference or not. This is something which measuring 3 basic properties (ICR) and frequency response cant give us an answer.

Unfortunately, as you said, you need precisely treated standardized room. Not to mention cost of equipment (a standard amp/player as well) and time required etc.... This is just not possible to do on a personal capacity.

 

[dc655321]

I don't think we are all that far apart, John. If you measure the amplitude and phase around the speaker (needed to include dispersion) then it will tell you when and where different frequencies will combine constructively and destructively. That should tell you when the image shifts. But, room reflections will affect all of that, as well as listener placement, the impact of things in the room (which can also reflect and absorb sound), and so forth. So, while it may be measurable, in practice it would be a daunting task and perhaps impractical in a real situation.

In addition to all that good stuff, the input signal is another major variable. Oft neglected imo...

 

[Duke]

If you measure the amplitude and phase around the speaker (needed to include dispersion) then it will tell you when and where different frequencies will combine constructively and destructively. That should tell you when the image shifts.

Are you aware of any peer-reviewed research which demonstrates the successful prediction of loudspeaker spatial qualities from measurements? I'm not saying it can't be done, but I'd be very interested in reading about it if it has.

 

[Inner Space]

It's not fetish, it's fidelity to the source.
Perhaps you've not seen the Genelec or Neumann kh80 measurements on this site? Or maybe you're not familiar with limitations of mechanical devices (vs electronic)?

My point is given the gross alteration imposed by the room, a flat FR has only a random chance of producing happiness. A non-flat FR has an equal chance.

 

[escksu]

I don't think we are all that far apart, John. If you measure the amplitude and phase around the speaker (needed to include dispersion) then it will tell you when and where different frequencies will combine constructively and destructively. That should tell you when the image shifts. But, room reflections will affect all of that, as well as listener placement, the impact of things in the room (which can also reflect and absorb sound), and so forth. So, while it may be measurable, in practice it would be a daunting task and perhaps impractical in a real situation.

You got it all wrong. Imaging in audio terms isn't even about where frequencies combined constructively and destructively. This imaging is a perception created by our brain when we hear the sound.

 

[JSmith]

All I know is...

JSmith

 

[dc655321]

My point is given the gross alteration imposed by the room, a flat FR has only a random chance of producing happiness. A non-flat FR has an equal chance.

Fair enough...
But that seems like intuitive rather than objective reasoning. I don't think I could personally accept that without evidence.

 

[Inner Space]

Fair enough...
But that seems like intuitive rather than objective reasoning. I don't think I could personally accept that without evidence.

There are dozens of posts on ASR, showing super-smooth Genelecs and Neumanns in-room - and the curves are all over the place. Is that evidence?

The reasoning is common sense. If a room has dips where the speaker has peaks, the listener will hear a different result than if the room had peaks where the speaker also had peaks.

 

[dc655321]

There are dozens of posts on ASR, showing super-smooth Genelecs and Neumanns in-room - and the curves are all over the place. Is that evidence?

It is not evidence that an anechoically flat transducer is somehow a bad idea or frequency response should be left to chance.

 

[escksu]

All I know is...

View attachment 144970



JSmith

Lol....haha.... yeah, this is not the place to talk about cables.

 

[Duke]

An anechoically flat, non-omni speaker will look a lot like those curves (along some axis in a typical room) once adjusted for human non-linear hearing characteristics.

The default assumption seems to be that a spectral discrepancy between the direct and in-room responses is not only inevitable in a non-omni speaker but is ALSO the ideal, with the target anechoic response being flat and the target in-room response looking like one of the curves in post #13.

I understand why this relationship between the direct and in-room sound is subjectively preferred in conventional monopole loudspeakers, but do we KNOW that such is the only viable approach? I readily concede that it is by far the most widely-studied approach.

 

[Inner Space]

It is not evidence that an anechoically flat transducer is somehow a bad idea or frequency response should be left to chance.

I think it's evidence that FR is already left to chance. The curve that hits your ear is the far end of a sequence of chances. The first chance being flat has a random eventual outcome.

 

[dc655321]

I understand why this relationship between the direct and in-room sound is subjectively preferred in conventional monopole loudspeakers, but do we KNOW that such is the only viable approach? I readily concede that it is by far the most widely-studied approach.

So, you don't really understand the significance of an anechoically flat frequency response?

 

[Blumlein 88]

Long ago, I did propose to quantify this. I believe with a good 3D waveform analysis software, it may be possible to derive characteristics of sound that causes us to have these perception. By repeatedly playing certain segments of an audio clip with various speakers (eg. one that has extremely strong imaging vs one that has hardly any imagining), I feel we may be able to observe a pattern and what constitutes to strong imaging. Then test this pattern with various speakers to test if this is repeatable. If we are successful, then we will know what kinds of wareform causes this.

It can even go to test other aspects. The infamous cables...lol... Granted if cables do not make a difference, then the audio produce should have no difference as well. This will be the first time we can really quantify if speaker cables do make a difference or not. This is something which measuring 3 basic properties (ICR) and frequency response cant give us an answer.

Unfortunately, as you said, you need precisely treated standardized room. Not to mention cost of equipment (a standard amp/player as well) and time required etc.... This is just not possible to do on a personal capacity.

You could do some of the same stuff by altering response of only one channel(or both channels) in a sound editor and see how much in FR bumps and at what frequencies causes imaging to degrade (or maybe improve).