Post by: Bob Somers on January 23, 2011, 11:22:25 PM
I would greatly value your observations on this from your own direct listening/measurements experience.
Also, if there are any well regarded academic research sources relating to this, a pointer to them would be great.
I have a second question.
For a home living room, say dimensions approximately 15' X 15' X 8'(height), which subwoofer (including those now available and also those that have been discontinued) or subwoofer combination (and if more than one of these is needed, please say so) would cover 125Hz down to the lowest audible frequency at say, +/- 3dB all at whatever the max SPL level is that the human ear can tolerate without pain (this may be about 120dB, not sure and may vary with frequency) and with inaudible distortion or if that is not achievable, then with the lowest currently available.
Of course, there are tradeoffs among the above criteria and I'm interested in your opinion on what equipment would give the most desirable results.
Please include these in what you consider if you have knowledge of their performance: labsub, submersive, contrabass, basstech 7, whatever the current most potent model is of the Danley Tapped Horn design, Thigpen Rotary Woofer (a fan with variable pitch blades), Paradigm Sub 2, and a real oldie: the Bose Acoustic Cannon.
Thank you very much!!!
Post by: James Feenstra on January 24, 2011, 02:24:55 AM
I guess if you felt like making a table (that you couldn't put anything on) out of a sub you could go with something like an adamson t21 w/ a lab gruppen fp7000 or equivilant
be prepared to spend $10,000+ PER SUB
one would probably do that kind of volume at that distance, or it'd kill you, one of the two
Post by: Ivan Beaver on January 24, 2011, 08:00:09 AM
http://www.danleysoundlabs.com/tapped_horn.asp?MODEL=DTS 10
For different size:
http://www.danleysoundlabs.com/tapped_horn.asp?MODEL=DTS 20
And the slim version:
http://www.danleysoundlabs.com/tapped_horn.asp?MODEL=TH SPUD
It is hard to distinguish between what you are hearing and what you are feeling at those freq.
But the louder it is, the lower you can hear it.
And at the lowest freq, are you hearing it or experiencing atmospheric pressure changes?
Post by: Art Welter on January 24, 2011, 02:27:15 PM
| Bob Somers wrote on Sun, 23 January 2011 21:22 |
I've seen all sorts of statements about what the lowest audible frequency by humans is and also, if there is a min SPL needed for it to be audible that is unusually high (such as 110dB, for example), please let me know what that value is. I would greatly value your observations on this from your own direct listening/measurements experience. Also, if there are any well regarded academic research sources relating to this, a pointer to them would be great. I have a second question. For a home living room, say dimensions approximately 15' X 15' X 8'(height), which subwoofer (including those now available and also those that have been discontinued) or subwoofer combination (and if more than one of these is needed, please say so) would cover 125Hz down to the lowest audible frequency at say, +/- 3dB all at whatever the max SPL level is that the human ear can tolerate without pain (this may be about 120dB, not sure and may vary with frequency) and with inaudible distortion or if that is not achievable, then with the lowest currently available. Of course, there are tradeoffs among the above criteria and I'm interested in your opinion on what equipment would give the most desirable results. Please include these in what you consider if you have knowledge of their performance: labsub, submersive, contrabass, basstech 7, whatever the current most potent model is of the Danley Tapped Horn design, Thigpen Rotary Woofer (a fan with variable pitch blades), Paradigm Sub 2, and a real oldie: the Bose Acoustic Cannon. Thank you very much!!! |
The Bose acoustic cannon, at about 106 dB at 25 Hz, does not qualify as a real contender in today's VLF arena. The 150 inch length is a bit of a problem too
.If you were to extend the lines to lower frequencies on the ISO 226 chart you would have a good idea of how loud low frequency needs to be audible at a particular level, and to be heard at equal loudness levels.
That said, there is another aspect to VLF hearing that is not often understood, at around 1000 Hz, a 10 dB change appears to be twice (or half) as loud, while at 20 HZ +5 dB appears twice as loud.
Because of this, and peoples perception, a slight increase in very low frequencies, or infrasound, can be annoying to one person and imperceptible to another.
I know I can hear idling trucks and generator VLF noise at SPL levels the chart says are below the minimum level. Wish I could say the same about my hearing response at 4K and above 15 K
.The levels required to damage hearing at very low frequencies are extreme, pretty much beyond the capability of any present transducers that could be fit in a room. Astronauts are subjected to over 160 dB of VLF on takeoff with no apparent hearing loss, there are no loudspeakers approaching that level at VLF.
One thing to watch out for is the tendency to turn up the volume when you have plenty of low, clean bottom.
Headphones are the cheapest way to experience low frequencies down into the single digits, and without the coloration a room adds. A 20 Hz wave is over 56 feet long, so even in a good sized room it will reflect off the walls, floor and ceiling many times.
The DSL tapped horns are probably the most efficient VLF loudspeakers, and can provide serious level in the large units.
Their disadvantage is size, and they require DSP for optimum integration due to the long path length.
There are plenty of subs capable of low 20s and even single digit performance that are plenty loud enough to be heard and felt that are not huge, though they use more power to achieve the level.
I use a Mackie HRS 120 in my control room, placed in a corner it is “flat” (other than the huge dips and peaks caused by the small room dimensions) to below 20 Hz. The size is only 18 x 19 x 21 inches, but it produces enough LF output to excite wall resonant frequencies as low as 15 Hz.
It is interesting to hear a wall flapping when “scrubbing” a tape (well, actually, scrubbing a hard drive).
It is rated at 117 dB maximum SPL at one meter, which it probably can exceed in a corner of a small room. It has no problem keeping up with a pair of Tannoy PBM 6.5, small near field monitors.
I purchased it used after listening to several other small studio monitor subs, it was the only one that I heard that had much out put below 30 Hz.
A pair in a small room should allow 123 dB peaks down to 20 Hz, which would “keep up” to 80 dB at 1K. In music, that level disparity is not "normal", few forms of music (other than some electronic dance music) actually have 20 Hz levels recorded louder than the 60-120 Hz range.
I was also quite surprised by the Carver Sunfire, which can get down to 18 Hz in less than one cubic foot, though it’s not as loud as the HRS 120, even though it uses more than four times the power.
Most “good” PA subs are dropping at 24 dB per octave at around 40 Hz, so the output these little buggers have exceeds many full sized 2 x18” cabs at 20 Hz.
Art Welter
Post by: Jonathan 'JP' Peirce on January 25, 2011, 03:07:15 PM
JP
Post by: Bob Somers on January 25, 2011, 04:55:29 PM
I wonder what it would be like if the DTS-10 drivers were driven by the servo motor arrangement (or similar) that had been used in the Contrabass?
Post by: Art Welter on January 25, 2011, 05:47:59 PM
| Bob Somers wrote on Tue, 25 January 2011 14:55 |
I've been doing more research on the Net and the Danley products certainly appear to be very potent. I wonder what it would be like if the DTS-10 drivers were driven by the servo motor arrangement (or similar) that had been used in the Contrabass? |
Large advances have been made in conventional woofers linear excursion potential, the difference between conventional and servo drive speakers is not much with the present state of the art.
Servo drive cones had around 16 MM Xmax, conventional cones of that time more like 4 or 5.
Today, 16 mm Xmax is still quite respectable, but becoming more common in conventional sub woofers.
Due to supply and demand, the price of servo motors is greater than it used to be, while high Xmax drivers costs have gone down.
Post by: Ivan Beaver on January 25, 2011, 07:07:23 PM
| Art Welter wrote on Tue, 25 January 2011 17:47 | ||
Large advances have been made in conventional woofers linear excursion potential, the difference between conventional and servo drive speakers is not much with the present state of the art. Servo drive cones had around 16 MM Xmax, conventional cones of that time more like 4 or 5. Today, 16 mm Xmax is still quite respectable, but becoming more common in conventional sub woofers. Due to supply and demand, the price of servo motors is greater than it used to be, while high Xmax drivers costs have gone down. |
You pretty much got it nailed. There is just no advantage to using the servo technology with the state of todays drivers.
In its day the servos were very impressive. But look at the state of loudspeakers in general-as compared to back then
Post by: Don Gspann on January 26, 2011, 08:49:59 AM
Post by: Jeff Knorr - Cobra Sound on January 26, 2011, 09:08:49 AM
I have two that I assembled from the kits that Danley offered a while back. In an open floor plan basement, roughly 30'x75', they ROCK! In a typical sized living room, corner loaded, they would be devastating!
Jeff "has better sound at home than the local movie theater"
Post by: Jared Bartimus on January 26, 2011, 01:51:12 PM
http://jtrspeakers.com/home-audio/orbit-shifter-lf/
In a corner I think that should have 109dB sensitivity. I wonder if Jeff has a chart showing the actual max output at 25hz because 139dB corner loaded seems a 'bit' unrealistic.
Post by: Alan Singfield on January 26, 2011, 02:17:11 PM
Just a suggestion.
Edit: incorrect URL
Post by: Art Welter on January 26, 2011, 04:21:07 PM
| Alan Singfield wrote on Wed, 26 January 2011 12:17 |
http://www.rotarywoofer.com/ Just a suggestion. Edit: incorrect URL |
They want a lot of money for a maximum acoustic output of only 115dB between 1 and 20Hz.
Probably really cool for sound effects, but there is not much music that goes below 16 Hz.
For a fraction of that cost, there are many subs with more output down to below 10 Hz.
Post by: Tim Padrick on January 27, 2011, 01:10:30 AM
| Don Gspann wrote on Wed, 26 January 2011 07:49 |
.... I'm sure the transient response is much better on the new subs,..... |
Given the slow rise time of the frequencies that are fed to subwoofers, I think that the term "transient response" is not applicable.
Post by: Art Welter on January 27, 2011, 12:07:50 PM
| Tim Padrick wrote on Wed, 26 January 2011 23:10 | ||
Given the slow rise time of the frequencies that are fed to subwoofers, I think that the term "transient response" is not applicable. |
Transient response describes the behavior of a system following a sudden change in its input.
It is not limited to high frequencies.
Many sub designs suffer from the persistence of cone movement or stored energy in the system after the signal has stopped, underdamped transient response.
Listening with two systems equalized for identical response, a sub with poor transient response, as described above, will subjectively sound “tubby” or “slow”. Percussive notes tend to blend together, bass lines become less recognizable.
I hate it when that happens
.
Post by: Tim Padrick on January 29, 2011, 03:38:21 AM
| Art Welter wrote on Thu, 27 January 2011 11:07 | ||||
Transient response describes the behavior of a system following a sudden change in its input. It is not limited to high frequencies. Many sub designs suffer from the persistence of cone movement or stored energy in the system after the signal has stopped, underdamped transient response. Listening with two systems equalized for identical response, a sub with poor transient response, as described above, will subjectively sound “tubby” or “slow”. Percussive notes tend to blend together, bass lines become less recognizable. I hate it when that happens . |
My thinking is that the term transient response describes the DUT's performance on the leading edge only - that anything that happens after the peak is reached has to be described by another term. No?
Post by: Art Welter on January 29, 2011, 02:01:17 PM
| Tim Padrick wrote on Sat, 29 January 2011 01:38 | ||||||
My thinking is that the term transient response describes the DUT's performance on the leading edge only - that anything that happens after the peak is reached has to be described by another term. No? |
No.
I have not found a single definition of transient response that does not take the tail end of a signal into account as well as the front end.
Webster’s definition of Transient: “passing especially quickly into and out of existence.”
McGraw-Hill Science & Technology Dictionary defines Transient Response as: “The behavior of a system following a sudden change in its input.”
Wikipedia: “In Electrical Engineering and Mechanical Engineering, a transient response or natural response is the response of a system to a change from equilibrium. Specifically, transient response in Mechanical Engineering is the portion of the response that approaches zero after a sufficiently long time (i.e., as t approaches infinity). “
But even looking at the transient response at the start of a signal, there are differences that can be observed between woofers.
A 10” woofer with a powerful magnet can accelerate a light cone faster at the onset of a signal than an 18” heavy cone woofer with a weak magnet. The former would have better transient response than the latter.
Post by: Mac Kerr on January 29, 2011, 02:25:53 PM
| Art Welter wrote on Sat, 29 January 2011 14:01 |
McGraw-Hill Science & Technology Dictionary defines Transient Response as: “The behavior of a system following a sudden change in its input.” Wikipedia: “In Electrical Engineering and Mechanical Engineering, a transient response or natural response is the response of a system to a change from equilibrium. Specifically, transient response in Mechanical Engineering is the portion of the response that approaches zero after a sufficiently long time (i.e., as t approaches infinity). “ But even looking at the transient response at the start of a signal, there are differences that can be observed between woofers. A 10” woofer with a powerful magnet can accelerate a light cone faster at the onset of a signal than an 18” heavy cone woofer with a weak magnet. The former would have better transient response than the latter. |
Of course, there are no sudden changes in input in a sub. All changes are gradual because of the LPF eliminating those pesky rapid changes. Obviously there can be mechanical systems that are so slow in response that even a relatively slow change to the input of the system can't be accurately tracked by the output. Since it is a case of the output not tracking the input I suppose that could be thought of as transient response. It isn't transient response in the sense that that term has in any other part of the system.
Mac
Post by: Art Welter on January 29, 2011, 03:18:04 PM
| Mac Kerr wrote on Sat, 29 January 2011 12:25 | ||
Of course, there are no sudden changes in input in a sub. All changes are gradual because of the LPF eliminating those pesky rapid changes. Obviously there can be mechanical systems that are so slow in response that even a relatively slow change to the input of the system can't be accurately tracked by the output. Since it is a case of the output not tracking the input I suppose that could be thought of as transient response. It isn't transient response in the sense that that term has in any other part of the system. Mac |
An underdamped woofer can persist in movement for several cycles after the signal is stopped, an example of a electro mechanical system that is so slow in response that a change to the input of the system is not accurately tracked by the output.
I guess we have a different idea of what "sudden changes" are.
100 "changes" a second is not sudden?
Of course, that 100 changes would be only the upper fundamental frequency, when a woofer is responding to a musical signal, there are multiple frequencies involved, the "changes" it needs to respond to are in the thousands.
When are changes sudden, at 1000, 10,000 Hz ?
How does transient response apply to one set of frequencies and not to another?
Post by: Mac Kerr on January 29, 2011, 04:17:34 PM
| Art Welter wrote on Sat, 29 January 2011 15:18 |
100 "changes" a second is not sudden? Of course, that 100 changes would be only the upper fundamental frequency, when a woofer is responding to a musical signal, there are multiple frequencies involved, the "changes" it needs to respond to are in the thousands. When are changes sudden, at 1000, 10,000 Hz ? How does transient response apply to one set of frequencies and not to another? |
Yes, the changes it needs to respond to are in the thousands, but all of them are below about 100Hz. In a sub, maybe an impulse with a 2.5ms 1/4 wave rise time or greater can be called "transient". What often gets mentioned in discussions about sub "transient response" in subs is the first snap of the beater on the bass drum. It you compare the rise time of that first impulse as heard by the mic to the signal as heard by the sub, the signal at the sub won't look so much like a "transient". If the "snap" of a bass drum is around 4kHz it has a rise time 1/40th of that of 100Hz, or 0.0625ms.
If "transient response" is the ability of the output to track the input, there is no lower limit to what a transient is, it is defined by the mechanical system's ability to track the input. If a "transient event" is defined by the frequency content of the original full range source, what happens at the sub is slow motion.
Mac
Post by: Art Welter on January 30, 2011, 02:26:30 PM
| Mac Kerr wrote on Sat, 29 January 2011 14:17 | ||
Yes, the changes it needs to respond to are in the thousands, but all of them are below about 100Hz. In a sub, maybe an impulse with a 2.5ms 1/4 wave rise time or greater can be called "transient". What often gets mentioned in discussions about sub "transient response" in subs is the first snap of the beater on the bass drum. It you compare the rise time of that first impulse as heard by the mic to the signal as heard by the sub, the signal at the sub won't look so much like a "transient". If the "snap" of a bass drum is around 4kHz it has a rise time 1/40th of that of 100Hz, or 0.0625ms. If "transient response" is the ability of the output to track the input, there is no lower limit to what a transient is, it is defined by the mechanical system's ability to track the input. If a "transient event" is defined by the frequency content of the original full range source, what happens at the sub is slow motion. Mac |
Exactly, accurate transient response is the ability of the output to track the input, particularly in respect to time, rather than frequency response.
If a "transient event" is defined by the frequency content of the original full range source, the ability of a sub to accurately reproduce the low frequency portion of the original "transient event" source related to time constitutes good transient response.
Lagging on the initial portion of the signal or "ringing" after would be poor transient response.
Post by: John Halliburton on February 01, 2011, 02:43:59 PM
| Don Gspann wrote on Wed, 26 January 2011 07:49 |
... but I do have 2 ContraBass subs in my living room. I'm sure the transient response is much better on the new subs, but for the size and weight and the sheer wall/floor shaking, knick-nak knocking off shelves enjoyment, I'm not giving them up anytime soon! A Telarc/DTS disc that has a recording of a space shuttle launch is too much fun to play for people. At first, all your you hear is the windows rattling, and feel the house shaking! Love them! |
Transient response of a Contrabass is actually some of the best you'll ever come across, even impressive if it didn't have a -3db point at 16hz. The design is very, very good, and nothing that small has that kind of ELF output either.
I've had mine out of pro "retirement" for the last couple of events I've done, and one was a movie presentation in a small gym of Toy Story 3. Nice fx from Disney/Pixar, the Contrabasses worked like a charm.
Best regards,
John
Post by: drewgandy on February 03, 2011, 01:53:43 PM
drew
Post by: Bob Somers on February 03, 2011, 02:22:42 PM
The question will be if an ideal speaker system, including a subwoofer, is fed a 100Hz tone that starts at time t=0 and is a cosine wave (so that at t=0, it starts out at max value which means it has an infinitely fast rise time) and it lasts for 1 second and then abruptly stops with an infinitely fast fall time, what should be the response of a perfect subwoofer?
Per Fourier analysis, all signals can be matched by a sum of sine waves. I also assume the subwoofer is fed by a perfect crossover set to pass everything at or below 100Hz and nothing above it. Possibly, this crossover is an active digital crossover. I'm rusty on some of my theory and so if implementing such a perfect filter is not possible, then the filter used will be assumed to be the best that can be realized and will be set so that nothing above 100Hz is passed to the sub even if that means also attenuating some frequencies at or below 100Hz.
This is now fed through a perfect (or near perfect) crossover to a subwoofer.
All the sub has to do therefore is reproduce frequencies at or below 100Hz. Therefore, nothing in the signal being sent to the sub can move fast. In fact, any sharp rise or fall times in the signal will actually be reproduced by the tweeter and/or midrange, not the woofer!
The fastest possible changes the sub must reproduce would be those of a 100Hz sine wave at the maximum amplitude the sub can handle with acceptably low audio distortion. Any frequency/amplitude combination below that changes even slower, so can be ignored.
If a sub can't "track" a 100Hz sine wave it max amplitude, isn't that simply a measure of distortion??? Stated another way, if a woofer can't reproduce some part of a 100Hz signal when the signal stops or starts, it won't be able to properly reproduce any cycles of that signal, not just the ones that start up the signal or stop it. If the start and stop of a signal can't be composed of anything higher than 100Hz, then there is no difference between that part of the signal and the "steady state" part that might be in between the start and stop. Therefore, the woofer will have distortion on all parts of the signal, not just the start and stop of it.
It seems to me there is no such thing as transient response and that what is incorrectly called transient response is really included within the definition of distortion.
Please give me your thoughts about my views on this.
Thanks,
Audioresearch
Post by: Art Welter on February 03, 2011, 08:42:03 PM
| Bob Somers wrote on Thu, 03 February 2011 12:22 |
It seems to me there is no such thing as transient response and that what is incorrectly called transient response is really included within the definition of distortion. Please give me your thoughts about my views on this. Thanks, Audioresearch |
I think you are trying to change the definition of transient response.
Any variation of a signal by the driver under test could be considered distortion, but transient distortion is different from harmonic, intermodulation or phase distortion.
Post by: Don Gspann on February 03, 2011, 08:54:03 PM
I'd like to add, that from my experience with them, and based on the use of the servo motor, they actually have an easier time going lower, rather than higher. Conventional cones still have a harder time the lower they go.
Post by: Don Gspann on February 03, 2011, 09:00:42 PM
natural philosophy, physics - the science of matter and energy and their interactions; "his favorite subject was physics"
oscillation, vibration - (physics) a regular periodic variation in value about a mean
Post by: jeff harrell on February 10, 2011, 06:58:55 PM
Post by: Bob Somers on March 02, 2011, 05:48:24 PM
If I'm wrong about that, and I may be, which off the shelf woofers now available could equal or exceed or at least come very close to the displacement volume of the Danley-type external motor with attached cone that was used in, say, the Contrabass?
Post by: Art Welter on March 08, 2011, 10:30:13 AM
| Bob Somers wrote on Wed, 02 March 2011 15:48 |
It just seems to my eye (just an educated guess not based on any measurements) from photos I've seen that the external servo motor Danley used is a much, much larger motor than what could fit into the "motor" portion of a conventional voice coil speaker and I would think that a cone driven properly by such an external motor would be likely to move a lot more air than a conventional voice coil driver with the same cone area because of what appears to be the much larger power that a big external motor could provide. If I'm wrong about that, and I may be, which off the shelf woofers now available could equal or exceed or at least come very close to the displacement volume of the Danley-type external motor with attached cone that was used in, say, the Contrabass? |
The servomotors really were not that big (or heavy) compared to huge slab magnets used on big woofers, much smaller considering one motor drives two cones in some of the designs.
The Servodrive cones (depending on which brochure I look at)have 1" to 1.25 inch linear peak to peak excursion, an Xmax of 12.5 to 16 mm. That was a big deal in the 1980's when most 15-18" woofers had less than half that excursion. It takes about a 6 dB increase in power to double excursion.
The relatively inexpensive dual slab magnet Eminence Lab 12 has 13 mm Xmax.
The B&C BC18SW115 has 14-16mm Xmax.
A JBL 2256G has an Xmax of 20.3 mm.
Speakers with more Xmax potential, and higher Bl than Servodrives are increasingly common.
Small, powerful Neodymium magnets make it possible to concentrate high flux density over a longer gap with far less size and weight, and new coil winding schemes make it possible to get more linear response with long excursions.