Question re: limitations of Frequency Response measurements (you there Sir Terrence?)

[kexodusc]

How much information does a frequency response plot really give us?

I recently had an opportunity to try out a few speakers with considerably different measurements and design philosophies.
On the one hand, I spent a good few hours trying out some of my uncle's Audio Note speakers, which typically measure very tight, +/- 1.5 dB if memory serves.
Then I stumbled into some Focus Audio FS-688 and 788 speakers that cost as much as the AN K's and AN E's, but measure +/- 3 dB's and get even sloppier in their low frequencies.

As much as I think of the Audio Notes, I was really wowed by the Focus Audio speakers.
Intuitively, one might expect the better measuring speakers to sound better, but I'm sure we've all heard speakers that measure well and sound bad, and vice-versa.

This got me to thinking about what's really being captured in these measurements. Obviously the fundamental tones are being measured, but are the harmonics and overtones being accounted for too? What about the impact of attack and decay? I mean, we could have a singer, a tuba, and a guitar making sound at 300 Hz, but the all sound different because of the harmonics or timbre. Does a frequency response plot account for this? And if so, to what degree?

Could we infer from a plot that a speaker with a good response should output the harmonics of an instrument within it's limitations (ie: +/- 3dB or whatever) as well as the fundamental frequency? (provided of course that the recording captured this well)

Thoughts?

[Woochifer]

A frequency response is a good indicator of what a speaker might sound like in person. But, when you're doing the actual listening, there are a whole slew of other factors that you need to account for. First off, a published frequency response measurement is typically done either in an anecholic chamber or outdoors in order to minimize the effect of room interactions. Obviously, at home the speakers are very much subject to the room interactions. The frequency response is an indicator of what you might perceive when you listen to something, but it's not the only thing to account for.

For example, if you look at Stereophile's speaker measurements, you'll see that they do a whole series of measuremens such as the off-axis frequency response, the spectral decay, cabinet vibration, and the impulse response, among others. Each of those measurements add to what the frequency response says. A decent frequency response from an on-axis measurement is good, but if it does not hold up in the off-axis measurements, then the speaker might tend to beam if it cannot be optimally setup. The spectral decay measurement identifies whether certain frequencies decay at different rates than others -- a "notch" in the graph can correlate to audible resonance. Here are some links to what their measurements mean.

http://www.stereophile.com/features/99
http://www.stereophile.com/features/100
http://www.stereophile.com/features/103

The thing to keep in mind is that every speaker out there is making compromises somewhere, and the ones that we individually prefer are the ones whose compromises are mean the least to us. Speakers that more obsessively focus on the time domain accuracy will tend to image better, but might lose some frequency response accuracy as a trade off. Other speakers purposely roll off the highs to put more emphasis on the midrange, while others boost the midbass to compensate for deficiencies in the lower bass range, etc. No speaker can perfectly replicate the source signal, so all the subjective evaluations have to do with what we prefer and what we're willing to live with.

Of course, the other part of all this is the room effects. The "slap echo" that you hear in an empty room is a time domain distortion that has audible effects on the sound. The shape and the size of the room, and the location of your listening position will influence how standing waves affect what you hear. All of these things have separate measurements and corrections that go along with them.

[Quagmire]

I think the larger question that you raise in your post, whether intentional or not, is if the science can accurately quantify not only the listening experience, but the equipment responsible for bringing us that experience as well -- not just limited to speakers but to the entire audio chain. I believe this is at the heart of the disagreement between the objectivist subjectivist groups. Though traditionally, speakers have been one area where there is little disagreement over the measured and perceived audible differences of audio equipment, I remember encountering a self proclaimed objectivist on these boards who's position was that there was no such thing as speaker "timbre". Even when I agreed with him that timbre is in a sense "distortion" in that it does distort the original audio signal, he absolutely refused to accept that any good quality speaker could have timbre. It didn't seem to matter to him when I pointed out that his much touted use of an equalizer was correcting for frequency response problems which, according to him, shouldn't exist in good quality speakers. It also didn't matter when I pointed out that no amount of EQing was capable of making horn loaded speakers sound like say... electrostatics, and that this was due to the specific timbre of each type of speaker. He wasn't hearing any of this. As far as he was concerned timbre was just a fancy way of saying "distortion" and that from his point of view any distortion of the source signal, save for output levels, was undesirable and shouldn't be tollerated by those buying speakers. So he was understandably pissed when I pointed out that his use of an EQ was a direct alteration (distortion) of the source signal prior to it every arriving at the speakers; and that unless he was using a very expensive EQ, he was like creating other sonic anomalies that he wasn't even aware of. From his objectivist point of view, everything he needed to know about his system and the listening experience could be evaluated and quantified by "correcting" the frequency response of his speakers. I doubt that there are many objectivist here who would agree with him, but IMO it is an example of how we get locked into our positions and simply won't budge. My position is that the raw data may not ALWAYS be able to quantify the experience or the equipment; and that because of this, we shouldn't sumarily dismiss all "claims" of perceived differences as being false. I understand that the reasonable response is to say "then why not use some objective test method to determine this instead of relying of subjective "biased" means?" -- and I agree. But I also believe that it is possible for the test methodology to be insufficient to completely evaluate the experience and the equipment, even though on face value, it may seem obvious that it is. I just raise the posibillity that it may not be and that in the future, as even better test methods are developed, some of the things which have been dismissed as pure audio myth may find some scientific foundation. Personally, I think it would be beneficial for those in both "camps" to get a healthy dose of humility and to say "We just don't know for sure."

My rant for the day.

Q

[kexodusc]

Q, I agree, just because we can't develop tests that capture all the info we hear and measure everything, doesn't mean there aren't differences.
I used the whol bumble-bee analogy and got flamed big-time...Scientists can't explain, replicate, and understand how a bee flies...yet it does....the only proof we have is our sense of sight.

But, I wasn't really headed in that direction with my post. Just wondering how much of a speaker's sound is captured by frequency response...good points though.

[Monstrous Mike]

Q, I agree, just because we can't develop tests that capture all the info we hear and measure everything, doesn't mean there aren't differences.
I used the whol bumble-bee analogy and got flamed big-time...Scientists can't explain, replicate, and understand how a bee flies...yet it does....the only proof we have is our sense of sight.

But, I wasn't really headed in that direction with my post. Just wondering how much of a speaker's sound is captured by frequency response...good points though.

Here is an interesting quote on this subject:

"That difficulty (explaining bumblebee flight) has even made its way into an urban legend of science, typically recounted as "a scientist 'proved' that a bumblebee can't fly" and often cited as an inspiring example for persevering in the face of overbearing dogma.

The fact is that the scientist in 1934 who claimed it is impossible for bees to fly used a steady state dynamic model, i.e. fixed wings like an aircraft. But bumblebees and hummingbirds have wings that move. So over the years, this myth has been conclusively disproven.

However, it is still referred to today when people claim something (like sonic differences in wires) that science cannot "prove" or says should be impossible. Like author of the qoute stated, myths like this are invoked by people with a dogmatic approach to life.

[kexodusc]

Monstrous Mike, I'm no scientist, but I do have several Biology textbooks (for example: Biology, Campbell and Reece, 3rd edition, Benjamin / Cummings Publishing Co.) from the late 90's (a very common 1st and 2nd year textbook in most US universities). At that time there was still no conclusive explanation explaining how the geometry, displacement, and mass of bee's body can allow for it to fly.
All it said was that scientist "assume it has something to do with the movement of the bees wings". Duh!
It goes on to say the US military is even studying it (as of that date) and hope to be able to incorporate some of natures wonders in future designs.
Assuming these folks aren't liars, I wouldn't call this an urban legend. So they've disproved one theory that claimed a bee "couldn't fly". They haven't not adequately explained the physics and reasons as to why a bee "does fly".
A difference you can surely accept.

[kexodusc]

Hey my thread got moved...D'oh! Old habbits die hard I guess.
I s'pose this is a better place for it, but I suspect the hit count will drop dramatically.

[E-Stat]

... but I suspect the hit count will drop dramatically.

I don't understand that perception. Am I that unusual by looking through most of the forums and threads, even if I don't have anything in particular to add?

rw

[kexodusc]

It's quite easy to understand, my question was about something quite relevant to the performance of speakers, not of a derogatory, or inflammatory nature. Though it was technically based, it is quite logical to think that the hit count could be higher in a forum dedicated to speakers specifically.

After all, as eric has stated, the purpose of the audio lab and moderators was to cut down on the nasty posts, and to allow hobbyists to post in the relevant forums (in this case speaker forum) worry-free, without being asked to "prove anything", without it degenerating into a naysayer-yeasayer argument.
As has been posted by several people, some will refuse to view the Audio Lab, for varying reasons. First, it's new, and often forgotten about by myself...bad habbit, as I said in my post. Second, some people avoid it because they rightfully don't care about the super-technicalities of this.
Posting my thread in the Audio Lab would make it more likely that these people would not see it. Undeniable.

Besides, it was my understanding from Eric himself that merely talking about speakers in a technical nature was not reason enough to post this in the Audio Lab. In fact, if you READ my first post, I was pointing out how scientific measurements DON'T necessarily capture everything.

Besides E-Stat, look at the traffic in the speaker forum compared to this one...even a completely compulsive liar wouldn't dare suggest it's even close.

But anyway, that's just my thinking. Now it's linked in the Speaker forum AND here...which I would never have done (dual posts) on my own. Thanks E-Stat (or whoever moved it)

[Geoffcin]

It's quite easy to understand, my question was about something quite relevant to the performance of speakers, not of a derogatory, or inflammatory nature.
But anyway, that's just my thinking. Now it's linked in the Speaker forum AND here...which I would never have done (dual posts) on my own. Thanks E-Stat (or whoever moved it)

I am moderator for the Speaker Forum and I didn't move your post. Heck, I wasn't even consulted. I don't see any good reason for moving it. At least you got double links now!

[kexodusc]

Oh, I wasn't complaining...this is kinda technical, though I didn't intend for it to be, but looking back, I'm not sure what else I could have expected...I just defended my logic for choosing the speaker forum in the first place...no biggy.
Like I said, this'll probably take some getting use to for everyone.

I do feel less people frequent the Audo Lab though...so far. Maybe if we called it the "Warm-n-fuzzy-happy-feel- good Lab"?

[ToddB]

Oh, I wasn't complaining...this is kinda technical, though I didn't intend for it to be, but looking back, I'm not sure what else I could have expected...I just defended my logic for choosing the speaker forum in the first place...no biggy.
Like I said, this'll probably take some getting use to for everyone.

I do feel less people frequent the Audo Lab though...so far. Maybe if we called it the "Warm-n-fuzzy-happy-feel- good Lab"?

I moved your post. The gist of your question was sufficiently technical that I thought freeing it from the constraints of the rest of the board would benefit the answers. If you'd prefer, I'll move it back.

As to your question, I've found frequency response charts to be useful for giving me an idea of general tonal balance, and little else. Not soundstage width/depth/specificity, resolution, macro/micro dynamic ability, speed, etc. Other tests might help account for some of these other aspects, but then there's your comment:

just because we can't develop tests that capture all the info we hear and measure everything, doesn't mean there aren't differences.

which I fully agree with. I hope someday that every aspect of what we can hear will be measured and quantified, because it will make things so much easier, but that day doesn't seem to have arrived yet.

[E-Stat]

Besides E-Stat, look at the traffic in the speaker forum compared to this one...even a completely compulsive liar wouldn't dare suggest it's even close.

No, but it could as easily suggest the true level of interest for those who wish to evaluate numerical results.

rw

[Geoffcin]

I hope someday that every aspect of what we can hear will be measured and quantified, because it will make things so much easier, but that day doesn't seem to have arrived yet.

That might help. One I would like to see is Square Wave performance. I've seen this only done regularly on amps, but it can be applied to ALL components right down to the wires. Cardas uses it as a measurement of his cables, and it's easy to see the difference in response between cables using it. Speaker testing would also be well served by seeing the closed mic response of a square wave. Transient attack, delay, and overshoot are all easily displayed by a square wave.

[ToddB]

That might help. One I would like to see is Square Wave performance. I've seen this only done regularly on amps, but it can be applied to ALL components right down to the wires. Cardas uses it as a measurement of his cables, and it's easy to see the difference in response between cables using it. Speaker testing would also be well served by seeing the closed mic response of a square wave. Transient attack, delay, and overshoot are all easily displayed by a square wave.

Yeah, I've seen those Cardas measurements, and other issues aside, they make me wonder if their reputation for having warm-sounding cables is really justified, or if their cables are just so pure-sounding that they reveal the warmth in the equipment they're connected to.