[konut]

I know that plotting frequency response graphs will show differences in actual SPLs at differing frequencies, just as using a meter can be used to measure the resistance, inductance, and capacitance in a cable, but how does one go about conducting, what seems to be the gold standard, a null test?

[JohnR]

I think you have to stick one piece in each ear.

[Mitsuman]

I know that plotting frequency response graphs will show differences in actual SPLs at differing frequencies, just as using a meter can be used to measure the resistance, inductance, and capacitance in a cable, but how does one go about conducting, what seems to be the gold standard, a null test?

A very sophisticated and sensitive instrument.......................your ears. 

[konut]

I think you have to stick one piece in each ear.

Are you sure one piece will be enough? And what about my other bodily orifices? 

A very sophisticated and sensitive instrument.......................your ears. 

 It seems that ears are not a reliable enough mechanism for repeatable, verifiable data. We're talking science here.

[JohnR]

Are you sure one piece will be enough? And what about my other bodily orifices? 

I advise against eating too much acoustic treatment. Especially if it's only wafer-thin...



But seriously, what is that you would be trying to determine with such a test?

[konut]

But seriously, what is that you would be trying to determine with such a test?

A number of things. I see a great deal of hyperbole being used to describe the changes that such and such room treatments make. It would be a good tool to correlate one manufacturers effectiveness to another, or whether the cost can be plotted against effectiveness. Also, it might be able to discern to what degree a component change might influence if a change is measured. For example, if you changed to speakers with a different directivity pattern you might influence the need for different degrees of reflection, absorption, diffusion, etc. A null test would quantify the results instead of relying on unreliable anecdotal opinions.

[neekomax]

Wouldn't simply measuring frequency response at the listening postion with some sort of RTA software be sufficient? You could do this with and without various treatments in place. I think the differences would be great enough that a null test wouldn't be necessary.

It is my understanding that a null test is best used to detect very small variations (or lack thereof). Treating a room will normally create large, easily detectable FR differences.

[konut]

A null test would quantify changes in a more dynamic environment such as with real music instead of test tones. Reverberation, echo, and ringing might also be more effectively measured. FR is only one measure that does not present the whole picture. 

[neekomax]

Ok, then you might consider getting in touch with Ethan Winer, who seems to have experience and expertise with acoustic treatments and is familiar with running null tests.

[bpape]

I would agree that there is a LOT more to it than just frequency response.  That said, static consistent measurements are really the only way to accurately measure FR and decay times, impulse responses, etc. 

I would also, however, agree that listening must take a place.  There are simply things that we do not yet understand how to quantify. 

Bryan

[konut]

Ok, then you might consider getting in touch with Ethan Winer, who seems to have experience and expertise with acoustic treatments and is familiar with running null tests.

I was kind of hoping he would comment in this thread.

I would agree that there is a LOT more to it than just frequency response.  That said, static consistent measurements are really the only way to accurately measure FR and decay times, impulse responses, etc. 

I would also, however, agree that listening must take a place.  There are simply things that we do not yet understand how to quantify. 

Bryan

I would agree with you. Others might not.

[Quiet Earth]

[Mitsuman]

Are you sure one piece will be enough? And what about my other bodily orifices? 

 It seems that ears are not a reliable enough mechanism for repeatable, verifiable data. We're talking science here.

Listening and what sounds good are not science. You can measure all you want but the real test is how it sounds. $.02 

[jtwrace]

Listening and what sounds good are not science. You can measure all you want but the real test is how it sounds. $.02 

  Not to an engineer.

[Mitsuman]

Not to an engineer.

I am an engineer, so what does that prove. Tell you what, you develop a machine/method of measuring beauty or love, and then we'll talk. 

[MaxCast]

Good sound is in the ear of the listener. 

[JohnR]

A number of things. I see a great deal of hyperbole being used to describe the changes that such and such room treatments make. It would be a good tool to correlate one manufacturers effectiveness to another, or whether the cost can be plotted against effectiveness. Also, it might be able to discern to what degree a component change might influence if a change is measured. For example, if you changed to speakers with a different directivity pattern you might influence the need for different degrees of reflection, absorption, diffusion, etc. A null test would quantify the results instead of relying on unreliable anecdotal opinions.

I'm not sure I would call what you are suggesting a null test, but as I understand it your suggestion is that the idea of differencing two signals might provide a way to get a handle on the magnitude of the difference. For example, eyeballing waterfall plots is pretty eye-glazing stuff - having just the "waterfall difference" would make it easier to see how much option A improved decay times vs option B. Is that the kind of thing?

I'm not sure what the difference of a measured response when playing music would show, as it's more difficult to analyze.

At any rate, I'm game to try it. Do you have measuring capability so we can compare results?

[konut]

I'm not sure I would call what you are suggesting a null test, but as I understand it your suggestion is that the idea of differencing two signals might provide a way to get a handle on the magnitude of the difference. For example, eyeballing waterfall plots is pretty eye-glazing stuff - having just the "waterfall difference" would make it easier to see how much option A improved decay times vs option B. Is that the kind of thing?

I'm not sure what the difference of a measured response when playing music would show, as it's more difficult to analyze.

At any rate, I'm game to try it. Do you have measuring capability so we can compare results?

A waterfall plot would be good, but I was thinking more in terms of a sound file that played the difference or a waveform file. I have no measurement capability. I was merely asking if this kind of test was being performed, and if not, why not?

[JohnR]

Hm, well, I guess because the intent of the null test is to show that there is no difference. You are however wanting to use it to quantify the size of a difference. I don't see how that can be done with a random music signal - short of running it through a frequency analyzer, but then you may as well use a frequency sweep so you have better analysis tools.

[rollo]

   Can we agree that a room correction device measures and corrects accordingly ? If you agree so far to date not one room correction device we have tried sounds good. The only area that makes a substansial difference is correcting the bass with say a Berhinger.
   In a live hall say Carnergie vs Lincoln center. One presents a bright presentation the other full and rich. both are treated acoustically yet have different characters. Which one is correct ? Both actually with different sonic characters. I would say that listening to the results of room treatments from measurements still leaves are subjective ears.
  We cannot rely on measurements alone. The biggest offense is when someone tells you that you are not hearing what you hear. The bottom line is its sound. How one arrives at that is the Manf's choice.
  Measure away but in the end it is what you here that determines your purchase. happy listening.

charles
SMA