Power response...why is it important?

[Coytee]

This is less of a question and more of a commentary as to at least one aspect that I think I'm finally beginning ot understand. I'll make the statement and if my understanding is wrong, then please, tell me how, where and why.

Power response is important for, several reasons. One of which is (or might be?), as the sound front leaves the speaker, it's going to get sprayed into the room. It's then going to start hitting walls, ceilings, floors and bounce around, turning into smaller and smaller bits of reflections.

If the speaker has an even power response, then in a sense, each one of these little sound bounces will contain most of the signal that was eminated from the speaker and you will have hundreds and thousands of little "sound bites" bouncing all around the room once it becomes a reverberent field.

To the degree the speaker has an even power response, each time one of these little sound bites hits your ears, you will hear the decay of the sound and each decay will in a sense, be a (smaller) representation of the entire sound spectrum that once eminated from the speaker. Sort of like "green" instead of "blue and red"

If on the other hand, the speaker has a very unbalanced power response (btw, what would that be called?)

anyways, if the speaker has a very unbalanced power response, then we're still going to have the same sound bites running around the room, HOWEVER, now, each sound bite will contain different bits and pieces of the total spectrum. You might have one sound bite that is heavy in the HF section and another one that is heavy in the LF spectrum, less like green and the blue/red would be visible components because they're not very well mixed (balanced). You would never be able to specifically single these out, but none the less, they're bouncing around your room. As they hit your ears, if they happen to hit simultaneously, then you might get a balanced sound to your ear but since they are random and you don't know which bite holds what sound spectrum... you might get a partial amount of the total output to your ears.

If that is reasonably accurate, then your typical results from using a speaker with a balanced verses unbalanced power response, MIGHT be...

?

Lack (or increase) of clarity?

Lack (or increase) of intelligiblity?

Lack (or increase) of spatial cues?

Lack (or increase) in ???

[gaspr]

My understanding of this is that it is important that the early lateral boundary reflections or the sound boucing off the walls, contain as much of the spectra of the on axis sound or the direct sound as possible. ( in the typical house, all reflections are "early") The reason that this is important is to maintain the "precedence effect". It seems that our ear brain processors can ignore or mask early reflections in a way that becomes pleasant to our ears. It seems to add a sense of spaciousness that we like. It is why we humans prefer to listen to music indoors. It is why speech becomes more intelligible in a good room.

If the reflected sound is missing some of the spectra that is in the direct sound, then our ear/brain starts to perceive it as a seperate sound and the pleasant "precedence effect" is lost. Note that thin absorbers wiil remove only some of the spectra and will cause the same type of problem.

This effect seems to be most concerned with frequencies above 300hz. Below this, room modes must be dealt with in some way.

[Arkytype]

Coytee,

Think of power response (PR) as the average of many anechoic frequency response measurements taken in a 360 degree sphere around a loudspeaker.

For sound system designers, the PR data is far more useful than a single axial frequency response curve especially when designing an array for specific coverage angles in a large venue. Of course, directivity curves, polar plots and other data are used as well in the design process.

The performance of a "perfect" loudpeaker (measurement-wise) in an imperfect acoustic space will suffer the indignities of uncontrolled reflections and poorly spaced room modes.

If you are satified (at least for now :>)) with the loudspeakers you have but aren't satified with how they sound, try to quantify what you are hearing or missing. If the bass is excessively boomy (or thin sounding) and relocating the loudspeakers to a different wall or moving them away from (or toward) the wall doesn't help, you probably need to invest in or construct some diaphragmatic-type bass treatment. OTOH if the bass balance seems about right but the highs seem too prominent, try some wall-mounted absorption products.

As for your last questions regarding how clarity, intelligibility, spatial clues correlate with a loudspeaker's PR; while there may be a parallel, e.g. a flatter power response might signify higher speech intelligibility, I don't think PR data alone can be used as an acoustic benchmark.

I wouldn't rely only on PR data to make a purchasing decision. Besides, how many manufacturers even publish PR data for their products? If you read the articles linked below, you'll learn that a loudspeaker's PR and frequency response don't necessarily correlate. The author of the first link's article notes that a loudspeaker with a flat PR won't necessarily have a flat axial frequency response and conversely, a loudspeaker with a flat axial frequency response may not have a flat PR. Kinda like the saying, while all Port is wine, not all wine is Port.

The author of the second link's article raises the oft-overlooked issue of how should a loudspeaker be energized when making axial frequency response measurememnts---constant voltage or constant power source?

The third link is to a paper, Basic Acoustics (written by John Vanderkooy) and is a good reference (with dumbed down math) for understanding acoustics and loudspeakers.

http://svconline.com/mag/avinstall_sound_power/

http://svconline.com/mag/avinstall_ambiguous_frequency_response/

http://www.linkwitzlab.com/LF-SL/Basic_Acoustics_P773.pdf

IMHO the acoustic arrival time from the loudspeaker drivers (woofer, squawker, tweeter) are of at least equal importance to the on-axis frequency response. I'm not just referring just to the arrival time difference caused by the physical front-to-back separation of say a K-55 diaphram from that of a K-77.

The acoustic arrival time from an individual driver can vary with frequency. The effect is that the driver is moving toward and away from the listener in a frequency dependent manner. Now that can have a marked (an measurable) effect upon clarity, intelligibility, spatial clues, etc.!!

Richard C. Heyser, father of Time Delay Spectrometry (TDS), wrote Determination of Loudspeaker Signal Arrival Times: Parts 1-3 which was published in the Audio Engineering Society's Journal in 1971. By using the then-new TDS measurement techniques he invented while working at JPL, Heyser makes the case for coherent arrival times in the design of a loudspeaker system. If there is interest in reading some of his insightful papers, I can post a few.

Lee

Lee

[RockOn4Klipsch]

....I wish I knew what was going on......lost in my own world

[Tempestking]

I may be lost as well, but this is wonderful information. Thank you for the articles and papers

~TK

[Coytee]

Yikes... I read most of that and got lost after about the third word... []

[DrWho]

http://svconline.com/mag/avinstall_sound_power/

http://svconline.com/mag/avinstall_ambiguous_frequency_response/

http://www.linkwitzlab.com/LF-SL/Basic_Acoustics_P773.pdf

Hmmm, I don't think any of those address the concept of power response as it pertains to a home environment (small room acoustics).

I wouldn't rely only on PR data to make a purchasing decision. Besides, how many manufacturers even publish PR data for their products? If you read the articles linked below, you'll learn that a loudspeaker's PR and frequency response don't necessarily correlate. The author of the first link's article notes that a loudspeaker with a flat PR won't necessarily have a flat axial frequency response and conversely, a loudspeaker with a flat axial frequency response may not have a flat PR. Kinda like the saying, while all Port is wine, not all wine is Port.

I think what Richard is referring to is the goal of having both a flat on-axis response AND a flat power response at the same time. In other words, constant directivity.

The author of the second link's article raises the oft-overlooked issue of how should a loudspeaker be energized when making axial frequency response measurememnts---constant voltage or constant power source?

I don't think that's an issue at all unless you're trying to find a way for the measurement to look like a flat line.....but that's as ignorant as zooming out really far so that it looks flat too.

In other words, the manner of taking the measurement directly influences the interpretation of the measurement. Both methods should result in the same conclusion, or you're doing something wrong.

But ultimately, speakers are designed for constant voltage sources. I think it makes it a lot easier to visualize what to expect from a measurement taken in the way your amplifier will behave.

[DrWho]

If that is reasonably accurate, then your typical results from using a speaker with a balanced verses unbalanced power response, MIGHT be...

?

Lack (or increase) of clarity?

Lack (or increase) of intelligiblity?

Lack (or increase) of spatial cues?

Lack (or increase) in ???

I don't think it really affects any of the above.

I think it mostly affects the perception of tonal balance.

I suppose it might sorta affect the spatial cues in the sense that it's kinda "confusing" (for lack of a better description) when the direct sound has a very different timbre than the reflected sound. However, in order for the reflected sound to have the same timbre, your room is going to need to have a "flat power response"...which isn't going to be the case. So the end user needs to choose whether or not the acoustics of the room will be addressed in a manner that will preserve the tonal balance, or if the power response of the speaker should be tilted to offset the natural behavior of the room. Whatever approach is taken, the on-axis frequency response should still be maintined...which ultimately means that EQ is not a perfect solution, just a band-aid.