Why don't manufacturers provide frequency response data to consumers?

[Chrispy]

Seeing what Polk has (not) provided in specifics/measurements for various speakers/subs over the years in their own literature, the response isn't surprising. But thanks for asking and posting the response....

 

[Rntlee]

I recently emailed customer service for the manufacturer of my speakers (polk) requesting the frequency response data. I explained I wanted to use it so I could apply equalization to get my preferred response. I provided the speaker serial numbers as well.

Their polite response was:
"Unfortunately, the frequency response graph data, the spinorama data, and the anechoic chamber data is not something that we can give out to consumers. We apologize for the inconvenience."

my question is why? Is it intellectual property or trade secret that could be used by competitors? Anyone have any insight as to why this is not freely available to customers?

What speaker model?

 

[Kvalsvoll]

The question is why not provide the data upon request?

Now you certainly have a point here. You also stated you provided the serial numbers. Some manufacturers actually store measurements of all speakers, for quality control.

As for the intellectual property - yes, measurements can provide information that could be useful for competitors. But then it would need something like a complete scan - as amir does with the Klippel. On-axis frequency response alone does not tell much, and even a few off-axis to give a better picture, does not reveal anything important.

Today, anyone can do a measurement of a speaker that gives a fairly good indication of the speakers anechoic on-axis response. Many reviewers do this. It would be quite difficult to keep a frequency response as some sort of secret.

 

[jgazal]

Transducers and passive crossovers seem to be the weakest link in the chain.
I highly recommend the chat Erin (@hardisj) had with Dr. Earl Geddes.
For instance, they mention that some speakers behave quite well warming up, but then a bad crossover design or bad choice of components leads too poor performance at operational temperatures.
I see it as resistors have a temperature coefficients and some type of inductors may saturate (see the Kef LS50 meta white paper for a example of how to mitigate such saturation).

So even a well behaved frequency chart may be misleading when crossover heat up...
At least transducers and cabinet resonances and diffraction one can relate to frequency response.

 

[Beershaun]

What speaker model?

RTi100.

 

[Rick Sykora]

I would love to see it if it could be done. Here are some of the major challenges:

1. Determining a reasonable set of measures and being able to replicate conditions sufficiently to be sure the results can be compared...
2. Even if you managed #1 (multiple NRCs?), not every speaker is optimized for the same purpose. Some are for studio monitors, some are for music listening and other are for sound reinforcement, etc. While this might be the easiest thing to do...
3. If we just picked out home listening, it would be difficult to measure how a CBT compares to a monopole compares to a omnipole and then capture it for someone who has not heard the differences between them to decide using a measurement?

As if these were not enough, we know getting consensus among different manufacturers in different regions is unlikely. Without some clear common motivation or regulation, the goal seems pretty out of reach to me.

 

[valerianf]

In the 80's there was a French speaker company named 3A that was providing an individual respons curve for each speaker.
Nowadays they do not exist anymore.
In fact they were providing a copy of the measurement of the first manufactured speaker (the prototype) in place of measuring each speaker as advertised.
I am not sure there is any use at getting the measurement of the gold sample unit: it is a guidance but the unit that we buy today may diverge after several years of production. Manufacturers are making all the time changes to the production units (i.e.: Denon with low performing capacitors as Amir prove it during his tests).
Only measurement Of a current unit can tell what we get.

 

[richard12511]

You can't do Speaker correction based on in room measurments, otherwise there would be no need for anechoic measurments to begin with.

That first part isn't true, but even if it was, it wouldn't mean that there is "no need for anechoic measurements to begin with".

 

[Head_Unit]

my question is why?

Oh, the comments above. But as a loudspeaker engineer, I'll also say "because it is irrelevant to consumers." Why is that? Well, because every designer and factory has their own somewhat different physical and electronic and software setup for measuring. You really cannot compare the results too minutely. Therefore that information is mostly for internal use-when I wanted to compare to competitors I bought them and measured them on my own idiosyncratic setup. Other specs like frequency response ± specs, power handling, sensitivity are also pretty irrelevant for similar reasons plus often averaged, exaggerated, lied about by the marketing department.

As for EQ, let me build on what was alluded in a previous post. I attended an Audio Engineering Society talk given my Jamie something-or-other who had been sound guy for The Grateful Dead. They proudly were the first to take 1/3 octave EQs out on the road, EQd every little peak and dip flat. It sounded like [vomit emoji]. Turned out this is because the response is actually in two parts: the first direct sound, then all the reflections. A regular EQ cannot fix problems except grossly because it messes with both the direct and the reflected. Then you need a time-based EQ like Dirac or Audyssey. (Note that at the lowest frequencies, the direct and reflected sound blur together and old-school EQ can sometimes help problems).

 

[Chrispy]

Oh, the comments above. But as a loudspeaker engineer, I'll also say "because it is irrelevant to consumers." Why is that? Well, because every designer and factory has their own somewhat different physical and electronic and software setup for measuring. You really cannot compare the results too minutely. Therefore that information is mostly for internal use-when I wanted to compare to competitors I bought them and measured them on my own idiosyncratic setup. Other specs like frequency response ± specs, power handling, sensitivity are also pretty irrelevant for similar reasons plus often averaged, exaggerated, lied about by the marketing department.

As for EQ, let me build on what was alluded in a previous post. I attended an Audio Engineering Society talk given my Jamie something-or-other who had been sound guy for The Grateful Dead. They proudly were the first to take 1/3 octave EQs out on the road, EQd every little peak and dip flat. It sounded like [vomit emoji]. Turned out this is because the response is actually in two parts: the first direct sound, then all the reflections. A regular EQ cannot fix problems except grossly because it messes with both the direct and the reflected. Then you need a time-based EQ like Dirac or Audyssey. (Note that at the lowest frequencies, the direct and reflected sound blur together and old-school EQ can sometimes help problems).

Perhaps irrelevant to a certain number of consumers, but why not make the data available in a somewhat standardized format?

 

[SIY]

Perhaps irrelevant to a certain number of consumers, but why not make the data available in a somewhat standardized format?

That certain number is likely 99%.

 

[Beershaun]

Oh, the comments above. But as a loudspeaker engineer, I'll also say "because it is irrelevant to consumers." Why is that? Well, because every designer and factory has their own somewhat different physical and electronic and software setup for measuring. You really cannot compare the results too minutely. Therefore that information is mostly for internal use-when I wanted to compare to competitors I bought them and measured them on my own idiosyncratic setup. Other specs like frequency response ± specs, power handling, sensitivity are also pretty irrelevant for similar reasons plus often averaged, exaggerated, lied about by the marketing department.

As for EQ, let me build on what was alluded in a previous post. I attended an Audio Engineering Society talk given my Jamie something-or-other who had been sound guy for The Grateful Dead. They proudly were the first to take 1/3 octave EQs out on the road, EQd every little peak and dip flat. It sounded like [vomit emoji]. Turned out this is because the response is actually in two parts: the first direct sound, then all the reflections. A regular EQ cannot fix problems except grossly because it messes with both the direct and the reflected. Then you need a time-based EQ like Dirac or Audyssey. (Note that at the lowest frequencies, the direct and reflected sound blur together and old-school EQ can sometimes help problems).

Thanks for the insight. Yes I can see how the information can be non-standard and not representative of the in room characteristics and therefore not something you can use outside the testing environment.

 

[preload]

4. It's way easier to sell a speaker with a nice story behind it, full of praise and adjectives, rather than with it's performance data.
First is less expensive and easier to do, second is more expensive and harder to achieve.

This

 

[preload]

Also, the spin data only help you rule out really horrible speakers. It's fairly useless in differentiating good and great speakers from one another, particularly when people are eyeballing the charts instead of applying mathematical analysis.

 

[Newman]

Hey all. Let's set my motivations aside for a minute. The question is why not provide the data upon request?

My speaker has a published frequency response spec. So someone took a measurement to determine the published spec.
36hz-27khz +/-3db
90db 1w1m
8ohm

Because then you go on the internet and say you got individual graphs upon request. Suddenly Polk have countless ‘individual requests’.

 

[sigbergaudio]

In the 80's there was a French speaker company named 3A that was providing an individual respons curve for each speaker.
Nowadays they do not exist anymore.
In fact they were providing a copy of the measurement of the first manufactured speaker (the prototype) in place of measuring each speaker as advertised.
I am not sure there is any use at getting the measurement of the gold sample unit: it is a guidance but the unit that we buy today may diverge after several years of production. Manufacturers are making all the time changes to the production units (i.e.: Denon with low performing capacitors as Amir prove it during his tests).
Only measurement Of a current unit can tell what we get.

We measure every single speaker and subwoofer and compare it to the reference. I'm reasonably sure most other speaker companies do the same. But actually providing individual printouts to every shipped speaker would add to cost and logistics, and we/they don't necessarily do that type of control testing in an anechoic chamber (that would significantly add to cost), so the control measurement may not accurately portray the frequency response of the speaker, it would just indicate that the variance is acceptable. Useful for internal quality control, not necessarily for publication.

 

[restorer-john]

Personally, I think individual QC test reports for top of the range products are not unreasonable and should be provided. Plenty of gear in the vintage era came with signed off individual test cards in the box.

But for normal run of the mill products, specifications are sufficient.

 

[Frgirard]

I recently emailed customer service for the manufacturer of my speakers (polk) requesting the frequency response data. I explained I wanted to use it so I could apply equalization to get my preferred response. I provided the speaker serial numbers as well.

Their polite response was:
"Unfortunately, the frequency response graph data, the spinorama data, and the anechoic chamber data is not something that we can give out to consumers. We apologize for the inconvenience."

my question is why? Is it intellectual property or trade secret that could be used by competitors? Anyone have any insight as to why this is not freely available to customers?

Neumann KH420 https://en-de.neumann.com/kh-420#technical-data

PSI Audio which displayed the curves of frequencies, distortion, directivity on its site no longer seems to be displayed. Each PSI audio speaker was or is delivered with its response curve.

When George Cabasse managed Cabasse, The measurements of the speakers was available.

 

[Koeitje]

I was looking into some cheap active monitors and Mackie was on my shortlist. Emailed them if they had any measurements available. They didn't. So I went for the JBL, which has its measurements printed on the retail packaging.

Personally, I think individual QC test reports for top of the range products are not unreasonable and should be provided. Plenty of gear in the vintage era came with signed off individual test cards in the box.

But for normal run of the mill products, specifications are sufficient.

Yep, just get me the measurements of the reference model they have for a line and I'm already happy.

 

[PierreV]

Personally, I think individual QC test reports for top of the range products are not unreasonable and should be provided. Plenty of gear in the vintage era came with signed off individual test cards in the box.

But for normal run of the mill products, specifications are sufficient.

I am a bit of a cynic when it comes to those. For all I know, this is part of the brand image, just as "built from hand-picked dry-aged Oregon wood" can be. I am assuming, per the factory tour videos, that Focal has at least some level of quality control and that modern production techniques make sure that parts behave in a mostly consistent way but I doubt that it has any impact on products that aren't defective.

I remember the vintage era when it wasn't rare to have a printed and signed bench results report, sometimes on a dot matrix printer but they never were especially useful to me. Nowadays, I believe it is more about creating an "unboxing experience" and putting the user in a "I am about to hear or use something special" mood than about real quality control.

But then, as I said, I am a cynic.