Hypothetically, let's say a crossover is designed with the speaker in an anechoic chamber. The resulting frequency response is flat, or whatever you prefer and design it to be. How would that frequency response change when that speaker is used in a real room?
Would it have more treble in a real room because when it was designed, the anechoic chamber was sucking up all non-direct high frequency radiation?
Same question for bass, in a real room there would be some reinforcement that was absent in the anechoic chamber. Would that cause increased bass in a real room?
Would it have more treble in a real room because when it was designed, the anechoic chamber was sucking up all non-direct high frequency radiation?
Same question for bass, in a real room there would be some reinforcement that was absent in the anechoic chamber. Would that cause increased bass in a real room?
The measurement of drivers and systems is by convention gated to remove room behaviour, so in effect loudspeaker systems are typically designed in an 'anechoic' environment, even without the need for anechoic measurements.
However the human ear / brain auditory system integrates energy over a ~10-30mS arrival window, which includes primary, secondary and even higher order reflections in a typical listening room. So clearly there is a significant divergence between how loudspeakers are typically designed and how human beings hear things.
A very small number of loudspeaker designers historically worldwide have grasped this dilemma and attempted to match the power response (i.e. total radiated power versus frequency) of loudspeakers to that of typical real acoustic sources in the same acoustic space. There is no 'one size fits all' solution, but avoiding the power response holes of the conventional loudspeaker design approach is an incredibly worthwhile pursuit IMHO.
However the human ear / brain auditory system integrates energy over a ~10-30mS arrival window, which includes primary, secondary and even higher order reflections in a typical listening room. So clearly there is a significant divergence between how loudspeakers are typically designed and how human beings hear things.
A very small number of loudspeaker designers historically worldwide have grasped this dilemma and attempted to match the power response (i.e. total radiated power versus frequency) of loudspeakers to that of typical real acoustic sources in the same acoustic space. There is no 'one size fits all' solution, but avoiding the power response holes of the conventional loudspeaker design approach is an incredibly worthwhile pursuit IMHO.
Last edited:
the ideal anechoic chamber "sucks up" all non-direct frequency radiation.
how the speaker sounds in the real room will depend on the room (dimensions, reflectivity), positioning of speaker and listener in the room and the directivity behavior of the speaker.
edit: for a speaker with high directivity or for small listening distance in relation to the room dimensions the speaker will sound similar in the real room compared to the anechoic chamber, because the reflected sound energy is small compared to direct sound.
for a low directivity speaker and/or big listening distance the sound of speaker will more depend on off-axis response of the speaker and the room.
Last edited:
I disagree. In a three-way speaker system for example, level matching of the three drivers is essentially equalizing. Level matching via added series and/or parallel resistance is part of crossover design. Other passive component design choices also affect frequency response curves.
Last edited:
Yes. But I am wondering which part(s) of the frequency spectrum would be more significantly affected when a design tuned in an anechoic chamber is moved to a real room
It is understood that all rooms are different. But typical rooms will be more like each other than they are like the anechoic chamber. There are no black-and-white answers. But I believe there are some generalizations that could be made.
I am assuming that few of us have the opportunity to design using an anechoic chamber. But it would be great if someone who has actually done this sees this thread and reports their experience.
Thanks to all who have responded so far
That's how most (good) loudspeakers are designed. Even if the designer does not have an anechoic chamber at his disposal, he/she will hopefully use anechoic measurements (gated impulse response technique) as a basis for the design. Let's assume by "frequency response" you are referring to the on-axis SPL response. How it all sounds to a human sitting in a real room depends on the acoustics of the room and the dispersion of the speaker. This is why a good loudspeaker design will not only consider the on-axis frequency response, but also off-axis.
During the past few years, people (including me) learned that it's necessary to consider the off-axis SPL response, and attempt to make those "flat", too (by "flat" I mean a straight line without any dips or peaks, and the line may/will be sloped). A good way to estimate the "in-room" sound of a loudspeaker is to look at the power response, which is the SPL response integrated over "all angles". Or, in other words, the power response indicates the sound power vs. frequency that is emitted into the room, and which may come back from the room to the ears of the listener. Determineing the power response requires numerous measurements of the SPL response at different angles (horizontal, vertical), which is a bit of work -- but it it's worth it, because it provides the designer with a very good indication of how the speaker will sound in the end.
There is no technical difference between a gated measurement and the same measurement in an anechoic chamber. Anechoicity is defined by compliance with the inverse square law, i.e. a doubling of distance = -6dB in level. By definition an appropriately gated MLS measurement in an ordinary room will give exactly the same result as the measurement in an "anechoic" room of the same dimensions. That is merely a consequence of basic physics. All current acoustic measurement systems use gated measurements by default, unless the operator is using them incorrectly.
you are right that very few are listening to the speaker in an anechoic chamber in the design process.
but most people measure frequency responses etc. in "quasi anechoic" conditions (gated/nearfield).
you can perfectly tune the on-axis and off-axis response in an anechoic chamber. you could even listen to the speaker on-axis and off-axis in the anechoic chamber.
off axis response (energy response) should not be too different from the on-axis response, as far as i know. it is considered standard to have a downward slope of energy response.
and there may be frequencies where the human hearing is more sensitive.
also as far as i know (off axis) peaks will be more audible than dips.
Even fewer people have typically had the opportunity to use a reverberation (echoic) chamber for design, which is a real shame becasue total radiated power versus frequency (which is measured in a reverberation chamber) is more representative of human auditory perception than anechoic on axis response IME.
Allow me to rephrase the question, and make it very specific:
There is a crossover (XO) redesign posted on the web that I am considering using. I don't want to identify it, but will describe it.
The XO is for a well known, vintage era, 3 way speaker. The XO was designed using state of the art measuring equipment perhaps about 10 years back. It was designed with the speaker in an anechoic chamber, with measurement mic on axis with the tweeter (which is situated above the other drivers) and 1 meter from the tweeter. The resultant frequency response curve is essentially flat, with no downward slope. Nothing is mentioned about power response, or off-axis response, figuring into the design of this XO.
The speakers would be used a room 14' x 10', not acoustically treated but with more than average quantities of carpet, furniture, curtains, etc. But obviously not close to the same conditions as an anechoic chamber.
I know that precise answers are not possible, and don't expect any. I am also not overly concerned about too much bass as I could do something to reduce that if necessary.
I am really only looking informed opinions as to whether those anechoic-flat-on axis frequency response speakers would sound overly bright in this room?
.
There is a crossover (XO) redesign posted on the web that I am considering using. I don't want to identify it, but will describe it.
The XO is for a well known, vintage era, 3 way speaker. The XO was designed using state of the art measuring equipment perhaps about 10 years back. It was designed with the speaker in an anechoic chamber, with measurement mic on axis with the tweeter (which is situated above the other drivers) and 1 meter from the tweeter. The resultant frequency response curve is essentially flat, with no downward slope. Nothing is mentioned about power response, or off-axis response, figuring into the design of this XO.
The speakers would be used a room 14' x 10', not acoustically treated but with more than average quantities of carpet, furniture, curtains, etc. But obviously not close to the same conditions as an anechoic chamber.
I know that precise answers are not possible, and don't expect any. I am also not overly concerned about too much bass as I could do something to reduce that if necessary.
I am really only looking informed opinions as to whether those anechoic-flat-on axis frequency response speakers would sound overly bright in this room?
.
Last edited:
Except for the power response hole of all order 2 or higher order HF crossovers to the tweeter, typically in the 2 - 5 kHz range, which (extremely unfortunately) coincides with the frequency range of NIHL (noise induced hearing loss which I prefer to call "exposure induced" hearing loss) or the wear and tear that everyones' ears experience over time. That frequency band is how humans seperate the consonant sounds, so a room that is deficient because of a power response hole is a bummer.
it is absoultely impossible to predict how the speakers would sound without off axis- or power responses.
Last edited:
Hmmm...the directivity is often ( if not always, but let's not count the baffle ) obtained by the means of a horn, which reduces the angle of projection but not the energy, so when it pin-balls on a reflective surface the wave still contains lots of it, so, no, not IME.
The anechoic chamber would need profound rifts to extend its action to the bass, but also the peaks and all the surfaces need to be acoustical inert ( someone said...Felt ?!)
if the horn would provide constant directivity over the complete speakers range you are right. but that is not very likely for real horns.
i think we are speaking of an ideal (theoretical) anechoic room