I have no issue whatsoever about the published work of O'Toole and Olive in regards to loudspeaker preferences. But remember they define preferences not accuracy or correctness. I would contend there is more to a good loudspeaker than just complying with the findings of O'Toole and Olive, although that is a pretty good start. (Oh - if only most designers got this far....!)
I won't argue. I will simply say that I asked a very specific question which was basically ignored in many of the replies. The interaction between speakers and room acoustics is complex, something that I was already well aware of. And there is an abundance of knowledge and theory that is available, some of which has been echoed in this thread. And that is all good. But my question had nothing to do with accuracy, correctness, preferences, or what constitutes a good loudspeaker or an ideal room curve.
I got the answer I was looking for from SpeakerBob's link. Thanks again for your input johnmath.
Ha ha I am guilty of not reading your first post. Glad you found what you needed! (Now with my tail between my legs...)
Some off-topic posts removed. Please keep responses respectful.
It would have HUGE boomy bass and be so bright you would run from the room screaming.
Room gain and "Peter Walker Curve"
Unless you were designing a PA system for outdoor use mounted high on a pole, in which case the chamber data would be reasonable.
Chambers are useful for driver design. That's about it. The do make great graphs for advertisements and to show how "scientific" the company is, which has no relation to how a speaker will sound in your home. Speaker companies have an impossible task, to design for an unknown environment. At least some companies used to try. Many speakers used to have an L-Pad on the tweeter and AR even had a bass contour adjustment on I think the AR-10. "Purist" advertising killed these useful features.
Agree with this statement.
I do not agree with this statement, although it is possible I have misunderstood your point.
j.
Well, 40 years of building speakers I understand the physics reasonably well. At one time, I too thought things like chambers were useful. I learned.
A chamber has no room gain, so flat bass would be a HUGE boom in bass. about 6 dB/octave from mid-bass down. This is why I perfecto build sealed woofers as the blend into a real room very well. Now, for an outside PA system or arena where there is no room gain, then the measurements would be valid. Not in a home. An in-room sealed speaker with a Q of under .7 , F3 about 60 turns out to be very realistic in a average room without eq.
And as was suggested, with no reflections supplementing the top end, flat in the chamber would be horribly bright.
Peter Walker ( Quad) published some suggestions on what you want at the listening position. It is quite rolled off. If you look at a lot of drivers, you will see the off-axis response falls off steeply. This works out well with speakers pointed strait, or pointed crossed in front of you, but listening on-axis is not very pleasant without substantial eq.
You don't need a chamber to fine tune crossovers. We have invented technologies like MLS ( the gated response mentioned above) that allow it to be done about anywhere. ( and free software to boot) That is why good speakers don't need to be designed in a chamber. But those big rooms do look impressive in advertisements. I don't know about you, but my living room does not look like that. Yes, we need to understand on and off axis in all dimensions. Many forget the vertical. We could guess a floor and a ceiling average, but that is about all.
Another problem with anechoic chambers is we do not really understand what is it that makes your brain think pressure waves are music. If we really knew, we could have a target curve and tune to that. Kind of like the valiant, but crude attempt with the Harmon curve. A step forward, but incorrect as it does not take into account enough of various ear shapes to give us localization information. We use frequency, phase and time. Actual good speakers are fine tuned by ear in a variety of real rooms by a variety of people.
I am not even including we have several generations on average with substantial hearing loss.
My generation, Who concerts. Then the ear bud generations which have been shown to cause damage at what you think is not very loud, Clubs, military service and workplace noise. It usually attacks the high end first, so maybe over-bright is needed to compensate on average. Add my generation getting old. As you age, the fluid in the inner ear thickens solely reducing your high end hearing. More fun? There are learned, cultural and ethnic based preferences on tonal balance. Speakers targeted for the far East will tend to be brighter than for the West.
It's complicated.
A chamber has no room gain, so flat bass would be a HUGE boom in bass. about 6 dB/octave from mid-bass down. This is why I perfecto build sealed woofers as the blend into a real room very well. Now, for an outside PA system or arena where there is no room gain, then the measurements would be valid. Not in a home. An in-room sealed speaker with a Q of under .7 , F3 about 60 turns out to be very realistic in a average room without eq.
And as was suggested, with no reflections supplementing the top end, flat in the chamber would be horribly bright.
Peter Walker ( Quad) published some suggestions on what you want at the listening position. It is quite rolled off. If you look at a lot of drivers, you will see the off-axis response falls off steeply. This works out well with speakers pointed strait, or pointed crossed in front of you, but listening on-axis is not very pleasant without substantial eq.
You don't need a chamber to fine tune crossovers. We have invented technologies like MLS ( the gated response mentioned above) that allow it to be done about anywhere. ( and free software to boot) That is why good speakers don't need to be designed in a chamber. But those big rooms do look impressive in advertisements. I don't know about you, but my living room does not look like that. Yes, we need to understand on and off axis in all dimensions. Many forget the vertical. We could guess a floor and a ceiling average, but that is about all.
Another problem with anechoic chambers is we do not really understand what is it that makes your brain think pressure waves are music. If we really knew, we could have a target curve and tune to that. Kind of like the valiant, but crude attempt with the Harmon curve. A step forward, but incorrect as it does not take into account enough of various ear shapes to give us localization information. We use frequency, phase and time. Actual good speakers are fine tuned by ear in a variety of real rooms by a variety of people.
I am not even including we have several generations on average with substantial hearing loss.
My generation, Who concerts. Then the ear bud generations which have been shown to cause damage at what you think is not very loud, Clubs, military service and workplace noise. It usually attacks the high end first, so maybe over-bright is needed to compensate on average. Add my generation getting old. As you age, the fluid in the inner ear thickens solely reducing your high end hearing. More fun? There are learned, cultural and ethnic based preferences on tonal balance. Speakers targeted for the far East will tend to be brighter than for the West.
It's complicated.
Dont know if this is sarcasm but I don't think I've ever seen instructions or recommendations on what kind of speakers a musician required their stuff to be listened on, and I listen to a wide range of music.
As far as bass being much stronger in a room than in an anechoic chamber, I think it is due to half-space (on floor) , quarter-space (against wall) , or eighth space (in corner) loading.
And seldom due to low frequency "room gain" ....as often also called "cabin gain", or "pressure gain". All calculated from the "longest wavelength supported" giving a "room cut-off frequency".
If the room is a concrete bunker, or even sealed up up like a car, then yeah, some cabin gain. But most residential construction is loose as a goose, and passes bass thru the walls, ceilings, and often even floors, pretty easily.
Interestingly, dipoles like electrostats, are not supposed to be able to achieve any low frequency room gain below room cut-off, even in a bunker, because forward and rear pressures cancel.
+1 to it's complicated
And seldom due to low frequency "room gain" ....as often also called "cabin gain", or "pressure gain". All calculated from the "longest wavelength supported" giving a "room cut-off frequency".
If the room is a concrete bunker, or even sealed up up like a car, then yeah, some cabin gain. But most residential construction is loose as a goose, and passes bass thru the walls, ceilings, and often even floors, pretty easily.
Interestingly, dipoles like electrostats, are not supposed to be able to achieve any low frequency room gain below room cut-off, even in a bunker, because forward and rear pressures cancel.
+1 to it's complicated
"If the room is a concrete bunker, or even sealed up up like a car, then yeah, some cabin gain. But most residential construction is loose as a goose, and passes bass thru the walls, ceilings, and often even floors, pretty easily."
Well not exactly. Now in my first trailer, I would say yes as 1 x 2 walls with paneling and aluminum were pretty much transparent. but if this was true, then half-space, quarter etc. would have no effect.
Yea, I have to laugh when beginners think a OB full range would be an easy first project.
Well not exactly. Now in my first trailer, I would say yes as 1 x 2 walls with paneling and aluminum were pretty much transparent. but if this was true, then half-space, quarter etc. would have no effect.
Yea, I have to laugh when beginners think a OB full range would be an easy first project.
As jReave said, the standard for a well designed speaker is to measure flat anechoically. To elaborate a little, this means either with a gated quasi-anechoic measurement or an anechoic chamber measurement. The chamber measurement is the gold standard, the gated quasi-anechoic is a reasonable compromise that gets us down to about 300 Hz.
It's an illusion like stereophonic sound in general. The popularity of those tells that many people like the "extra spaciousness" they give. As well I know many people who prefer horn speakers which work in another way, to minimize reflections and give sharp stereo image. I think it is really strange that many people prefer listening to headphones...
Dipole or omnispheric radiation pattern are just ways to manipulate reflections in a room, they don't give anything extra in an anechoic chamber or oudoors. Tonal balance must be fixed with dipoles, low freq behaves different from monopoles.
I think that it should be pointed out that the "room curve" that Toole talks about in his writings is measured not at 1 or 2 meters from the front baffle of the loudspeaker, but more like 3-4 metres (10-12 feet)--with all the all-pass, floor bounce, ceiling bounce, and sidewall bounce reflections, etc. thrown in. I haven't found where Toole really says at what conditions his "room curve" was measured via calibrated microphone. This is pretty important to note.
BTW, I see a lot of people are trying to replicate Toole's house curve without realizing Toole himself shows pretty much flat SPL response of the loudspeakers that he referenced under test (of course, forgetting about the other half of the transfer function). What you'll find is that the room curve occurs naturally when the loudspeakers are set up for flat SPL response on-axis at 1 m with plenty of absorption on the floor to control floor bounce to be able to use the phase information.
Toole apparently just talks about whether or not the "room curve" is steep enough or shallow enough when measured with greater distance and room reflections, etc. thrown in , and recommends adjusting it...but curiously, he doesn't mention anything about the effects of directivity. In my room, my loudspeakers are set up for flat response, but my loudspeakers have full-range directivity down to the room's Schroeder frequency. I measure their response at 1 m, not 3 m, and correct their SPL and phase response to flat.
When I turned on Dirac...voila!, there was my "room curve" in the measurements when measured at the listening position instead of 1 m. (BTW: I abandoned Dirac later when I realized it could not distinguish between room reflections--which can't really be EQed--and direct-arrival minimum phase acoustic energy at midbass frequencies.)
It seems like I'm splitting hairs here, but I'm not. Trust me.
Loudspeakers that lose their directivity below 800-2200 Hz (depending on the diameter of the woofer) will typically need more "room curve", otherwise called "salt & pepper EQ", to sound correct in-room even if they are hybrid horn/direct radiating loudspeakers which increasingly lose directivity at woofer frequencies below the crossover frequency and spray the near field with acoustic energy at lower frequencies.
(By the way, it's Toole--not O'Toole--who's one heck of a nice guy who's spent his career on loudspeakers and small room acoustics. I think we should get his name right.)
Hope this helps.
Chris
I guess we will have to disagree. I build speakers that sound "natural" in the intended environment. I once thought the pure anechoic was the standard, and some still do, but I learned better. It is only a "gold" standard if the advertisers say so. There is no accepted standard.
I am with Cask05 on the auto eq mess. It is getting far better and I would pick Dirac over Audessey any day, but I still put on music and tune to where a cymbal sounds right, where male and female voices sound tight. I pay most attention above 1K.
One thing you can't predict is when you walk into a room, your brain makes an acoustic map and it will have expectations on the eq. If you try to force some pre-determined curve that is too far from your brain's expectations, it will not sound right, even if it is in theory perfect.
I am with Cask05 on the auto eq mess. It is getting far better and I would pick Dirac over Audessey any day, but I still put on music and tune to where a cymbal sounds right, where male and female voices sound tight. I pay most attention above 1K.
One thing you can't predict is when you walk into a room, your brain makes an acoustic map and it will have expectations on the eq. If you try to force some pre-determined curve that is too far from your brain's expectations, it will not sound right, even if it is in theory perfect.
The op asked about crossovers. The point is actually that it is impossible to design a proper crossover in a measuring environment that has interference from reflections or room gain. An anechoic room is of course best, except for really low crossover frequencies. Most anechoic rooms don't go below 100 Hz or so without calibration to take account of room gain.
Hi, What do you mean by MLS?
(the only abbreviation i know means 'maximum length signal' which is just a particular form of test signal, with the inevitable plusses and minuses for what/how it measures.)
Getting back to room construction and bass loss....ime, it takes one hell of a stiff construction to produce real low frequency room gain.
Takes at least brick walls, or concrete walls, be they block or solid.
I've found 2x6 framing with any kind of wood or composite-plank siding to be quite leaky.
And we haven't even got to the ceiling!, or floors!