Detuned Demystified

[drmcclainphd]

This is a little white paper sort of thing with two purposes. It clarifies and corrects some ideas people have about "detuned", and contains a little test I did on a minimal detuned design. The first part was written mostly for those on guitar forums I frequent, as they tend to run on repeated buzzwords rather than numbers. Feel free to skip that if you like. I just thought the testing on the design would be of interest here.

=====

Speakers, or more accurately, the drivers (which along with an enclosure make up a speaker) have physical and acoustic characteristics according to their design and construction. These characteristics can be used to determine how and what size of an enclosure will produce a desired result, such a flat response, accentuated bass response, etc. When a speaker is constructed according to these parameters, with an internal volume that best produces the desired result, it is considered to be “tuned”. The simplest of these is an entirely closed enclosure. Some designs call for a small opening called a vent, usually with a tube that extends into the enclosure. This is called a tuned port. In both this design, and the previous, the tuning depends on matching the internal volume to the characteristics of the drivers. The wrong size of enclosure, the wrong size port, or replacing one of more drivers from a tuned speaker with one or more with different characteristics results in a poorly tuned speaker.

Some speakers, particularly those in guitar “combo” amplifiers (combined amplifier and speaker) are open in back. The back side of the drivers are open to the air. The internal volume effectively becomes infinite. Such a design has no tuning involved, and is considered “detuned”. Although there is no optimization of the speaker involved, in a speaker of this type there is typically much control over the signal to make it sounds as desired, as well as being noticeably louder than other designs due to sound from the back side of the driver being able to freely escape the enclosure.

Some have come to consider that a guitar speaker enclosure with half the drivers removed (one of two, or two of four) to be detuned. This is only partially accurate. To be truly detuned, the area of opening to let the “rear” sound escape should be larger than that of the drivers. A 12” driver has an area of 113 square inches. To be truly detuned the back side must have an open area greater than this. The actual size of the enclosure is irrelevant, only the driver and the size of the open back matter. Too small a “rear” opening and there will be resonance effects, emphasizing some frequencies and dampening others.

The term “detuned” has also been used to describe a design with the “rear” opening being redirected by the enclosure towards the front, producing more volume being directed towards the listeners. If constructed with the proper opening size, this design is detuned, but detuned does not necessarily mean this redirection takes place. It's simply a very good idea, and is the basis for having half the driver mounting holes left open in a regular speaker.

Many very detailed detuned designs have been attempted, some with adequate reasoning, others simply a matter of convenience or taste. The fact is the design requires little more than the rear opening area being larger than the driver area. I attempted to produce a minimal detuned design as a proof of concept test of this point. The “enclosure” consisted solely of a 12” plastic bowl, a 12” x 24” x 1/2” sheet of “expanded metal” steel grating, and appropriate fasteners and wiring. The driver was an 8” GRS “BOFU” clone.

[See picture 1]

The driver was attached to the grill,

[See pictures 2 and 3]

which was then suspended from a horizontal bar by hooks made of stiff wire.

[As in picture 5]

The back was left open to the air, outdoors, to minimize reflected sound affecting measurements.

A 1kHz sine wave tone was generated by computer and amplified to 1 watt (measured as 2.8 volts in parallel with the 8 ohm driver impedance). The volume of the sound produced was measured with a hand held SPL meter held 1 meter away from the center of the driver, on axis, integrated over the 30 second duration of the tone.

The driver is rated at 91 dB for this measure. The measurement on this driver and configuration was 93 dB. This is well in keeping with the fact that manufacturers often state parameters conservatively in case a particular piece does not perform as well as intended. I have obtained slightly higher than stated response from these drivers in other tests.

The driver/grill assembly was then attached to the lip of the bowl. The driver was positioned off-center to minimize resonance effects. The 8” driver has a frontal area of 50 square inches, and the bowl 113 square inches. The difference, 63 square inches for the “rear” opening is approximately 25% greater than the driver's area.

This assembly was then hung in the same manner as for the first test.

[See pictures 4 and 5]

The same signal at the same power level was tested in the same manner. The response of the system in this detuned,
forward facing output was 101 dB. Both theoretical and measured increases in response for detuning a given speaker is 3 to 6 dB. I speculate that this greater improvement, nearly double the volume, was due to the use of an approximately hemispherical reflector held in close proximity and nearly surrounding the driver from the mounting flange down. (Driver depth, flange to bottom of magnet = 3.875”; bowl depth 4.5”).

In any case, I believe this demonstration (NOT an experiment – there are important differences) supports the concept of detuned enclosures for greater SPL output from a given driver configuration. Also, not only is a minimal rear enclosure space adequate, it may be beneficial if configured to produce greater response.

Questions welcomed: [email protected]
Snide remarks go to: [email protected]

Dr. Dennis McClain PhD
“I am a scientist, but I don't play one on TV.”

[billfitzmaurice]

Re: Detuned Demystified

I'm afraid that almost everything contained above is fundamentally wrong, and for those of you not well schooled in how loudspeakers actually function the best advise I can offer is to ignore all of it. :(
I don't have the time or inclination to explain point by point why the conclusions drawn are erroneous, but if any of the others here who do want to bother, by all means have at it.

[drmcclainphd]

Re: Detuned Demystified

I'm afraid that almost everything contained above is fundamentally wrong, and for those of you not well schooled in how loudspeakers actually function the best advise I can offer is to ignore all of it. :(
I don't have the time or inclination to explain point by point why the conclusions drawn are erroneous, but if any of the others here who do want to bother, by all means have at it.

I fully expected you to respond first, and in this vein.

You declined to provide anything of substance the last time you made such an assertion, relying on requoting perfect-world theoretics and neither refuting the numbers themselves nor providing any of your own with respect to real world acoustics. I replied back with more numbers, and still nothing. Once again you express ire instead of numbers. Engineering is a matter of mathematics, Bill, not adjectives, and real world engineering is nothing like point sources floating in empty space. And relying on perceived authority in lieu of actual data is a major logical flaw. I heartily concur that any who wish to examine the situation do so and draw their own conclusions.

As for the test itself, it's simple and cheap enough for anyone to replicate and extend, and they can draw their own conclusions on that too. I would like little more than for someone to provide numbers refuting it. Thus knowledge advances. Besides, I only posited that the effect was reflected sound. I'd be interested to know to what extent the larger opening caused it to act as a horn, emphasizing the effect.

[mzisserson]

Re: Detuned Demystified

A 1Khz sine wave has a lambda of 13.56 inches. Since that is larger than the driver itself, the back wave off the driver will begin to cancel out the forward axis wave in free air. This could throw off your measurement at 2.83V at 8 ohms, assuming the BOFU is 8 ohms at 1K. So you basically but a bowl around it creating U type of baffle but less of a baffle since the driver was still open to the sides. Assuming you did not move the SPL meter because it was affixed to a mic stand, and all the work you did around the driver sis not disturb its position and you were exactly on axis for your second measurement, gain would be expected since you are actually preventing some of that back wave from canceling the front with the bowl both widening the "baffle" and providing a chamber to extend the time it takes for the rear wave to interact with the front wave therefore driving the point of cancellation down in frequency.

It's just math as you say. Bill is very right, and that's only the beginning.

To me, you demonstrated that blocking the rear wave or preventing it from interacting from the front INCREASES gain. Also if you are drawing a correlation between a fixed point and a variable, the method used to derive the information to conclude a correlation indeed lends itself to become an experiment, not a demonstration.

[killa]

Re: Detuned Demystified

Your "detuned" more commonly called open back speaker might be louder at higher frequencies due to reflections but your lows will be gone and the highs sound will be highly dependant upon placement and room.
Your second type of detuned speaker is going to sound horrible with a woofer size hole and power handeling will be greatly reduced due to the driver reaching xmax easily. This would just be a horribly tuned port which is not good.
Your third example sounds more like a horn loaded enclosure but if it is not optimized it will also run into power issues.
The lack of understanding of equipment from guitar players (I am one myself) drives me crazy. If they would take the time to learn a little they could save a lot of money without going through a bunch of cabs and equipment.

[drmcclainphd]

Re: Detuned Demystified

A 1Khz sine wave has a lambda of 13.56 inches. Since that is larger than the driver itself, the back wave off the driver will begin to cancel out the forward axis wave in free air. This could throw off your measurement at 2.83V at 8 ohms, assuming the BOFU is 8 ohms at 1K. So you basically but a bowl around it creating U type of baffle but less of a baffle since the driver was still open to the sides. Assuming you did not move the SPL meter because it was affixed to a mic stand, and all the work you did around the driver sis not disturb its position and you were exactly on axis for your second measurement, gain would be expected since you are actually preventing some of that back wave from canceling the front with the bowl both widening the "baffle" and providing a chamber to extend the time it takes for the rear wave to interact with the front wave therefore driving the point of cancellation down in frequency.

It's just math as you say. Bill is very right, and that's only the beginning.

To me, you demonstrated that blocking the rear wave or preventing it from interacting from the front INCREASES gain. Also if you are drawing a correlation between a fixed point and a variable, the method used to derive the information to conclude a correlation indeed lends itself to become an experiment, not a demonstration.

Thank you for the rational and challenging response.

It is a demonstration, and not an experiment, because it is a single data point. There can be no inference because there is neither enough instances ("subjects") nor iterations of the variable representing the volume and/or "rear" opening size, to be able to conduct any inferential statistic such as correlation. I heartily recommend Collins and Pinch's second book, "The Golem Unleashed" for a clearly presented, insightful and entertaining exposition into the subject of the difference between. I mean, crashing an airplane, and running a locomotive into a nuclear waste container, these are some exciting attempts to make a point, but all they did was make a single point. Feynman's famous dipping an O ring into ice water in front of congress is in there too. A Nobel prize winner made a valid point, but only one point, and I have no doubt he knew it. I've always used both their books as required reading in my methodology courses.

As for your comments regarding the design, I believe I understand what you're saying. However, it is not clear to me whether you have taken into account the fact that the reflected waves will be phase reversed, and that the driver was mounted off-center (perhaps I failed to emphasize this; if so I apologize) by two inches. The distance between the cone and the bowl varied by 4" in the 12" cavity. No doubt this would have some effect on any resonance effects (constructive or destructive), though how they would interact with the front signal due to the fact that they would vary constantly across the range of frequencies (wavelengths) represented by the cone-to-bowl distance.

After having been challenged on some of the effects, and whether they were confounds, illusory or simply misguided, I extended the measurements to octaves +/- 3, and included a condition with a 24x24" baffle surrounding the driver. I suppose there may be enough here to conduct an ANOVA/multiple regression, but I believe the fact that some of the results contradict some of the purely theory driven assertions made in some comments warrants examination and contemplation prima facia, which is, after all, how said assertions were made and thus apparently apparently intended to be taken.


All in dB/1 w/1 m
f bowl baf open
8 k 84 95 82
4 k 92 90 94
2 k 94 98 99
1 k 100 95 92
500 92 90 85
250 82 70 72
125 72 60 68



To actually be an experiment I would test several different instances of the same driver. But all I had intended to do was demonstrate that the effect occurred. I believe I did so, and in response to criticisms then showed that the loss of bass response many predicted did not, in fact, happen. Quite the opposite in fact.

[drmcclainphd]

Re: Detuned Demystified

Your "detuned" more commonly called open back speaker might be louder at higher frequencies due to reflections but your lows will be gone and the highs sound will be highly dependant upon placement and room.
Your second type of detuned speaker is going to sound horrible with a woofer size hole and power handeling will be greatly reduced due to the driver reaching xmax easily. This would just be a horribly tuned port which is not good.
Your third example sounds more like a horn loaded enclosure but if it is not optimized it will also run into power issues.
The lack of understanding of equipment from guitar players (I am one myself) drives me crazy. If they would take the time to learn a little they could save a lot of money without going through a bunch of cabs and equipment.

Thank you also for the response. I will approach it in the order presented.

An open back speaker is detuned, as I said in the original. However, there are other detuned configurations which differ significantly. I defer to O'Connor's book as my source for the operational definition which is an open back larger than the frontal driver area, reflected forwards. My 3 piece PA (two 2x6x5.5" towers + a center unit with 4x4x5.5" in a "Sweet 16" arrangement), all of detuned design (per O'Connor), have been performing admirably in both indoor and outdoor venues, over which I had no control on placement, and certainly not environment, with quite adequate bass.

I concur that simply leaving a driver hole open is non-optimal, as an understatement. I also said this in the original.

The third assertion is, I believe, something that is answered by my recently posted expanded testing. I would like to request specifically what you mean by "power issues".

As for the last, I most heartily concur. As stated in the original, this was precisely my purpose in conducting this test. I would only wish to emphasize that "learning a little" should include learning from testing real things embedded in reality, not only from books or from "common wisdom". The latter is particularly subject to the "Chinese Telephone" effect in the retelling, but also to the logical flaw of appeal to authority, where one recounts something they have come to believe because a perceived authority said it. I'm reminded of an old Doonesbury cartoon. Here's the description from a paper decrying a common classroom environment:

"In a recent Doonesbury cartoon a teacher is upset because his students seem more intent on writing down what he says than on listening and understanding. He delivers a series of outrageous statements culminating with "Jefferson was the Antichrist! Democracy is fascism! Black is white! Night is day!" while students scribble frantically without pause. As the teacher collapses, saying "Teaching is dead," one student says, "Boy, this course is really getting interesting." Another answers, "You said it. I didn't know half that stuff."

If not clear, I too am a guitar player. And luthier, and do my own recording/production, and performing myself as well as doing live sound for others. I started playing in 1964, the same year I started working in my dad's TV shop. My first amplifier was a gutted TEAC reel to reel with a 10 watt tube amp running into a 12" corner-sitting Karlson. In the last half century I've built more equipment than I've bought. It's only in the last 20 years that I've applied my scientific training and tested the concepts I was considering using, to find out if they were actually correct or just "really getting interesting". In my neuroscience work, nothing gave me more pleasure that refuting a "common wisdom" with hard data, and bursting the bubble of an expert old fart. I intend to continue, in all my endeavors, and accepted wisdom be damned, and with good reason, far more frequently than I'd expected,

[killa]

Re: Detuned Demystified

Did you keep the meter at 1m away from the center of the cone in the second measurement? You have basically horn loaded your driver and I believe the low end measured the way it did due to the meter getting close to the horn as it was possibly moved to keep same distance to driver. I would set the meter 1m away form the expanded steel for all measurments as this would be where the imaginary baffle would be.

[billfitzmaurice]

Re: Detuned Demystified

You have basically horn loaded your driver .

Not with the throat that large and the mouth that small. Mzisserson is correct, un-baffled the driver was simply experiencing major front/rear wave cancellation. The 'cup' increased the distance the front and rear waves had to travel before meeting, lowering the cancellation frequency. There was no gain, there was just less cancellation sourced sensitivity loss. As for 'detuning', there's no such thing. All speakers are tuned, having a resonant frequency, even raw drivers, where the resonant frequency is Fs. Open back cabs have an Fb, and also suffer from front/rear wave cancellations, as do all open baffle designs, which open back guitar cabs are merely a variety of.

[badman]

Re: Detuned Demystified

Not with the throat that large and the mouth that small. Mzisserson is correct, un-baffled the driver was simply experiencing major front/rear wave cancellation. The 'cup' increased the distance the front and rear waves had to travel before meeting, lowering the cancellation frequency. There was no gain, there was just less cancellation sourced sensitivity loss. As for 'detuning', there's no such thing. All speakers are tuned, having a resonant frequency, even raw drivers, where the resonant frequency is Fs. Open back cabs have an Fb, and also suffer from front/rear wave cancellations, as do all open baffle designs, which open back guitar cabs are merely a variety of.

Actually, it IS effectively hornloaded- not "effectively" as in "Useful" but effectively as in "In effect". The device is sufficient to achieve pattern control and quarterwave behavior, to some extent, at the measurement frequency (1khz).

I agree that "detuned" is a funny way of saying "OB".

[killa]

Re: Detuned Demystified

Not with the throat that large and the mouth that small. Mzisserson is correct, un-baffled the driver was simply experiencing major front/rear wave cancellation. The 'cup' increased the distance the front and rear waves had to travel before meeting, lowering the cancellation frequency. There was no gain, there was just less cancellation sourced sensitivity loss. As for 'detuning', there's no such thing. All speakers are tuned, having a resonant frequency, even raw drivers, where the resonant frequency is Fs. Open back cabs have an Fb, and also suffer from front/rear wave cancellations, as do all open baffle designs, which open back guitar cabs are merely a variety of.

I don't know why but I was thinking there would be a steeper rolloff at lower frequencies after they get long enough to cancel (like when you have a port) But thinking with a less tired brain I realize this is not the case.
As for the horn part maybe we should call it a waveguide :D Whatever you want to call it my point was mainly about proper distance of the spl meter.
I would call it poorly tuned not detuned :D

[billfitzmaurice]

Re: Detuned Demystified

I would call it poorly tuned not detuned :D

Not at all. Guitar drivers don't go all that low, and to get what's perceived as good bass response, though it's actually in the midbass, a high Q enclosure is required. That's what open back is. But to call it detuned means that whoever did so probably doesn't even know what Q is.

[killa]

Re: Detuned Demystified

Not at all. Guitar drivers don't go all that low, and to get what's perceived as good bass response, though it's actually in the midbass, a high Q enclosure is required. That's what open back is. But to call it detuned means that whoever did so probably doesn't even know what Q is.

I wasn't talking about the open back design specifically. But the design with one woofer removed would be bad.

[billfitzmaurice]

Re: Detuned Demystified

I wasn't talking about the open back design specifically. But the design with one woofer removed would be bad.

It also would be high Q, so it may not be totally without merit.

[killa]

Re: Detuned Demystified

It also would be high Q, so it may not be totally without merit.

You may be correct. I have never tried it so I can't really say, but in my head it seems like a bad idea. Of course we are talking about a guitar cab so it doesn't necessarilly need to be high fidelity.