Frequency Response Discrepancy

[AcuDefTechGuy]

We all know for a fact that speakers have to be heard.
We can't just say that since a speaker has a great Frequency Response that it is automatically a great speaker.

We are all familiar with the "+/-3dB @ 90dB w/m" F.R. standard.
However, some companies don't include these specifications unless we specifically ask for them. Definitive Technology and MQ Quart are just two such examples. They only list the "Overall" F.R. of their speakers. Some people even consider these specs dishonest and say that their F.R. are just plain wrong.

Def Tech list the F.R. of the BP7000Sc as "11 Hz - 30 kHz". When ask, Def Tech states that the +/-3dB F.R. of the BP7000SC is 20 Hz - 20 kHz. Home Theater Magazine independent measured the F.R. of the BP7000SC to be 22 Hz - 20 kHz +/-3dB. So is Def Tech lying?

I have compiled a quick chart comparing the manufacturers' +/-3dB F.R. versus the 3rd Party Independent +/-3dB F.R. of 76 speakers. Interestingly enough, only three speakers (DefTech CLR3000, Linn Akurate 242, Paradigm S8) matched the Independent 3rd Party's F.R.
Four speakers actually had better F.R. when tested by 3rd Party (DefTech BP10B & BP8B, Linn Komponent 110, Polk LSi25).
All the rest were off by as many as + 41 Hz!

http://www.freewebs.com/acudeftechguy/frequencydiscrepancy.htm

 

[Alex2507]

Great work. I love knowing that about Gallo's. As another member stated, "There is no substite for displacement."

 

[abboudc]

We all know for a fact that speakers have to be heard.
We can't just say that since a speaker has a great Frequency Response that it is automatically a great speaker.

We are all familiar with the "+/-3dB @ 90dB w/m" F.R. standard.
However, some companies don't include these specifications unless we specifically ask for them. Definitive Technology and MQ Quart are just two such examples. They only list the "Overall" F.R. of their speakers. Some people even consider these specs dishonest and say that their F.R. are just plain wrong.

Def Tech list the F.R. of the BP7000Sc as "11 Hz - 30 kHz". When ask, Def Tech states that the +/-3dB F.R. of the BP7000SC is 20 Hz - 20 kHz. Home Theater Magazine independent measured the F.R. of the BP7000SC to be 22 Hz - 20 kHz +/-3dB. So is Def Tech lying?

I have compiled a quick chart comparing the manufacturers' +/-3dB F.R. versus the 3rd Party Independent +/-3dB F.R. of 76 speakers. Interestingly enough, only three speakers (DefTech CLR3000, Linn Akurate 242, Paradigm S8) matched the Independent 3rd Party's F.R.
Four speakers actually had better F.R. when tested by 3rd Party (DefTech BP10B & BP8B, Linn Komponent 110, Polk LSi25).
All the rest were off by as many as + 41 Hz!

http://www.freewebs.com/acudeftechguy/frequencydiscrepancy.htm

I wouldn't trust any manufacturer's stated performance of their own product. They test under ideal conditions, and have a stake in the outcome. It doesn't really matter to me how accurate their measures are nor does it necessarily correlate to the quality of their product.

 

[jaxvon]

We all know for a fact that speakers have to be heard.
We can't just say that since a speaker has a great Frequency Response that it is automatically a great speaker.

We are all familiar with the "+/-3dB @ 90dB w/m" F.R. standard.
However, some companies don't include these specifications unless we specifically ask for them. Definitive Technology and MQ Quart are just two such examples. They only list the "Overall" F.R. of their speakers. Some people even consider these specs dishonest and say that their F.R. are just plain wrong.

Def Tech list the F.R. of the BP7000Sc as "11 Hz - 30 kHz". When ask, Def Tech states that the +/-3dB F.R. of the BP7000SC is 20 Hz - 20 kHz. Home Theater Magazine independent measured the F.R. of the BP7000SC to be 22 Hz - 20 kHz +/-3dB. So is Def Tech lying?

I have compiled a quick chart comparing the manufacturers' +/-3dB F.R. versus the 3rd Party Independent +/-3dB F.R. of 76 speakers. Interestingly enough, only three speakers (DefTech CLR3000, Linn Akurate 242, Paradigm S8) matched the Independent 3rd Party's F.R.
Four speakers actually had better F.R. when tested by 3rd Party (DefTech BP10B & BP8B, Linn Komponent 110, Polk LSi25).
All the rest were off by as many as + 41 Hz!

http://www.freewebs.com/acudeftechguy/frequencydiscrepancy.htm

While you are correct about not being able to trust manufacturer specs, you are obsessing about frequency response and extension. These are not sole determining traits of speakers. As for the Vandersteen VSM being "+41dB off", that is good to note, but does not say anything about the quality of the loudspeaker, it just means there was an error of measurement on one party or both.

 

[TLS Guy]

Frequncy response

i agree with jaxvon. There seems to be an obsession in these forums with low freqency extension.

The most important first attribute of any speaker is to have a very smooth midband response. If it has it does not follow that it will be a good speaker. However if it has a rough mid band response it will be a bad speaker.

If I had to have one item to look at to make a speaker short list it would be the waterfall response. Unfortunately most manufacturers would not like their waterfall plots seen by the public. The waterfall plot is revealing of axis, room response, phase and time delay problems. However there is no substitute for a listen. As B & W point out even with all the best technology, speakers still have to be voiced and tuned by ear after very extended listening tests.

 

[Seth=L]

It could mean they used different testing methods, maybe one or the other has a more effective way of testing or accurate. I wouldn't fret about 2hz much. I do find it a little dishonest that they publish specifications like 11hz-20khz, it doesn't help anyone to differ from standards.

 

[fmw]

The other issue is the listening room. The room acoustics dominate what speakers do. Measurements need to be done in anechoic chambers and none of us listen to speakers in anechoic chambers. Measuring speakers in a normal listening room is meaningless because the measurments would only correspond to that room. That's why speakers don't usually get same sort of frequency response measurements that electronics do. If a flat frequency response in a real listening room is the goal, then equalization is the only answer.

But anything you or anyone else can do to help clarify speaker measurements certainly can't hurt.

 

[TLS Guy]

The other issue is the listening room. The room acoustics dominate what speakers do. Measurements need to be done in anechoic chambers and none of us listen to speakers in anechoic chambers. Measuring speakers in a normal listening room is meaningless because the measurments would only correspond to that room. That's why speakers don't usually get same sort of frequency response measurements that electronics do. If a flat frequency response in a real listening room is the goal, then equalization is the only answer.

But anything you or anyone else can do to help clarify speaker measurements certainly can't hurt.

Room effects do have an effect. Now I have been chastised for this before, but speakers vary greatly in their sensitivity to room problems.

No we don't listen in anechoic, and the sound would be terrible if we did.

In most rooms conversational speech sounds fine and the household piano if there is one. As I have pointed out before for most rooms the reflected sound is returned in the Hass fusion zone, so the reflected sound can not be detected by the ear separate from the direct sound. Now if a speaker has an off axis response that does not mirror the axis response, it will be more severely affected by the room as compared to one that does. Again a look at the waterfall plot will be helpful here.

The other issue is bass standing waves. The worst case scenarios tend to involve cubic or close to cubic rooms. If there is a dominant dimension it makes a room less prone to this problem. If the width of the room can be roughly half the length this really helps.

However even in ambient rooms speakers can still give a good account with out room treatment.

Now these speakers are in a suboptimal location in a room that is not promising at all. In fact it has a distinct echo!, so there are pronounced reflections out of the fusion zone. However I managed to voice them to sound very good in the space. I'll admit it was more work than I bargained for.

http://mdcarter.smugmug.com/gallery/2424105#127080824

I only like to resort to equalizers for very tough situations. They have a whole subset of problems all to themselves.

 

[pzaur]

The website just shows how important it is to have a decent sub to mate with your setup.
I would venture to say that a vast majority of home systems have a sub, so the low end frequency extension for any speaker isn't really that important. As long as there is some type of low end in the speaker to allow overlap with a sub, the discrepancy isn't a major issue.

-pat

 

[TLS Guy]

The website just shows how important it is to have a decent sub to mate with your setup.
I would venture to say that a vast majority of home systems have a sub, so the low end frequency extension for any speaker isn't really that important. As long as there is some type of low end in the speaker to allow overlap with a sub, the discrepancy isn't a major issue.

-pat

In general I agree with that comment. However the crossover to the sub is still a crossover. Now that crossover is not as critical as a crossover at the usual 2.5 to 3KHZ range. However it is still a crossover with the problems that go with it. For superlative results I still feel there is benefit in an integrated system designed from the ground up, versus any speaker, any sub and set the crossover frequency with out regard to the order of the crossover to mains and sub.