RE the acoustic origin:
If you want to define the apparent acoustic center as the location in space where sound eminates from, it turns out this location is actually an inch or two IN FRONT of the driver (when mounted in a box). This has been determined an confirmed multiple times via this experiment: the on and off-axis reponses are obtained for the driver, in a horizontal plane intersecting the driver, every few degrees. Sound pressure measurements are performed for various frequencies, down to some very low frequency such as 10 Hz. At very low frequencies the system should be acting like a monopole, for which the sound pressure radiates spherically. This means that the pressure and SPL should be equal at all positions around the "center" of the source. When you make such a measurement at VLF, it turns out this "center" is actually in front of the driver by a couple of inches and not back at the voice coil or whatever. I first hear this from Siegfried Linkwitz back in the early 2000s and at the time it sounded wrong but it is indeed the real-world behavior.
Your own measurement shows some delay of a few usec (51.73usec). This corresponds to a length of about 0.7 inches behind the baffle, so a few inches behind the apparent center. This does not sound correct, unless this includes some crossover components, e.g. a highpass filter.
Anyway to answer your question about a "distance", you just need to calculate how far sound travels during the time interval in question. This is simply:
d = c * t
where d is the distance, c is the speed of sound, and t the time interval. The units must be the same for the numbers you plus in e.g. if you use c= 344 meters / second then d will be in meters and t should be in seconds.
If you want to define the apparent acoustic center as the location in space where sound eminates from, it turns out this location is actually an inch or two IN FRONT of the driver (when mounted in a box). This has been determined an confirmed multiple times via this experiment: the on and off-axis reponses are obtained for the driver, in a horizontal plane intersecting the driver, every few degrees. Sound pressure measurements are performed for various frequencies, down to some very low frequency such as 10 Hz. At very low frequencies the system should be acting like a monopole, for which the sound pressure radiates spherically. This means that the pressure and SPL should be equal at all positions around the "center" of the source. When you make such a measurement at VLF, it turns out this "center" is actually in front of the driver by a couple of inches and not back at the voice coil or whatever. I first hear this from Siegfried Linkwitz back in the early 2000s and at the time it sounded wrong but it is indeed the real-world behavior.
Your own measurement shows some delay of a few usec (51.73usec). This corresponds to a length of about 0.7 inches behind the baffle, so a few inches behind the apparent center. This does not sound correct, unless this includes some crossover components, e.g. a highpass filter.
Anyway to answer your question about a "distance", you just need to calculate how far sound travels during the time interval in question. This is simply:
d = c * t
where d is the distance, c is the speed of sound, and t the time interval. The units must be the same for the numbers you plus in e.g. if you use c= 344 meters / second then d will be in meters and t should be in seconds.
^^ Ah, that's the weighting you're referring to (before setting an average etc.).
To a point, but the fact is that USB microphones / measurement methods aren't ideal for this unless you use something like Jeff's approach, which has some limitations, but is good enough for most requirements.
Boxsim just uses whatever data is entered into it: it doesn't calculate or specify what the driver's Xmax (or Xmech / Xlim) is itself. The pre-loaded data in the software will be whatever Visaton specify for the relevant drivers since it's their software. Remember that other manufacturers will not necessarily specify Xmax in the same way though, since there is no standard defition, so to compare like with like you'll need to check they're using the same method, or calculate it yourself & use that for the data entry.
Not really, you need more than one source to get a decently accurate offset if you're measuring with a USB microphone.
Right, I've been using laser measures for years, as have many others; they make life a lot easier.
To a point, but the fact is that USB microphones / measurement methods aren't ideal for this unless you use something like Jeff's approach, which has some limitations, but is good enough for most requirements.
Boxsim just uses whatever data is entered into it: it doesn't calculate or specify what the driver's Xmax (or Xmech / Xlim) is itself. The pre-loaded data in the software will be whatever Visaton specify for the relevant drivers since it's their software. Remember that other manufacturers will not necessarily specify Xmax in the same way though, since there is no standard defition, so to compare like with like you'll need to check they're using the same method, or calculate it yourself & use that for the data entry.
Not really, you need more than one source to get a decently accurate offset if you're measuring with a USB microphone.
Right, I've been using laser measures for years, as have many others; they make life a lot easier.
The apparent acoustic center point of emanation and the time of arrival from the acoustic point of origin are two different and separate things.
This thread is more concerned with the time of flight from the acoustic point of origin.
As an illustration, a horn loaded woofer or tweeter will both have an acoustic center point of emanation near the horn mouth, but the acoustic point of origin could be anywhere from a fraction of a millisecond for a small tweeter to many milliseconds for a deep horn.
To time/phase align the drivers, the offset of the acoustic point of origin, not the acoustic center point of emanation is used.
The acoustic point of origin is generally near the voice coil former and cone juncture, which on a woofer is generally behind the baffle.
The speed of sound through the solid voice coil former is transmitted far faster than the speed of sound in air, so the location of the coil center in the magnetic gap is seldom the exact acoustic point of origin- it will be closer to the midpoint of the cone for woofers.
0.7 inch behind the baffle sounds about right as the acoustic point of origin for a 7" speaker.
Art
That's how I thought about it when I came across the idea - an apparent point source.
When boxsim is used with Visaton drivers it sets these. Eg a 6.5" 47mm and a tweeter 6mm. These are referenced to the surface of the baffle so that any mix or number of chassis can be used in the package.
Latest thing the package has taught me is that I am unlikely to be able to do what I want to do with any 6.5" chassis. Xmax limitation and F's lower than I am interested in coming along and trying to achieve an SPL of say 100dB. Things also rather tight in the lower regions of the range I am interested in. LOL Time to look at 8".
Head scratching. Xmax limiting SPL at low frequencies. If I look at designs I have seen even 10Hz ok at 28w with a certain Xmax limit. Then I compare different speakers and in many areas there is not much of a difference in spec other than Xmax and a larger Xmech.
Taking one with a smaller Xmax and I can increase it. At 40Hz ~28w is ok. maybe more but at 20Hz the limit is about 75dB and a lower wattage. Use 2 which means doubling the cab volume doesn't change things all that much in this area. I smell a difference in simulation and specifications so have decided to go ahead with a 6 1/2" but in a tower about 1.2 meters tall with a volume of 50L. That leaves scope for modification on chassis size and make in the future. It's based on a Boxsim design. I needed to change the xover to restore the response to what was shown in this design. Seems the drivers spec has changed since 2014. Software too. Nice and simple. Just 2 2nd order xovers and one correction on the tweeter. A simpler one than others that I have seen which do the same thing. Probably build as per the simulation to see what happens. The chassis costs are on the low side as well.
Taking one with a smaller Xmax and I can increase it. At 40Hz ~28w is ok. maybe more but at 20Hz the limit is about 75dB and a lower wattage. Use 2 which means doubling the cab volume doesn't change things all that much in this area. I smell a difference in simulation and specifications so have decided to go ahead with a 6 1/2" but in a tower about 1.2 meters tall with a volume of 50L. That leaves scope for modification on chassis size and make in the future. It's based on a Boxsim design. I needed to change the xover to restore the response to what was shown in this design. Seems the drivers spec has changed since 2014. Software too. Nice and simple. Just 2 2nd order xovers and one correction on the tweeter. A simpler one than others that I have seen which do the same thing. Probably build as per the simulation to see what happens. The chassis costs are on the low side as well.
75dB SPL @ 20Hz could be achieved with a 6.5" driver with about 3.7mm Xmax.
75dB at 20Hz is just barely audible (to average hearing 18 year olds), equivalent in perceived equal loudness to 50dB at 40Hz and about 5dB at 1000Hz.
A quiet room will usually be above 25 dB...
More Xmax requires stiffer (heavier) cones, longer (heavier) higher power voice coils, limiting the useful upper frequency response range and reducing sensitivity.
In a sealed cabinet, SPL is simply related to excursion and Sd (cone area).
Doubling equally powered (same voltage applied to both) drivers in double the cabinet volume adds +6dB SPL.
Two drivers can produce the same low frequency SPL as one with half the excursion.
Four times the excursion is required for the same SPL for each octave drop, 40Hz to 20Hz.
http://www.baudline.com/erik/bass/xmaxer.html
Smaller boxes require more power to reach a given excursion.
If you want loud and even low frequency below around 80 Hz, better off using larger high excursion subwoofers that can be spread around the room and hidden.
Art
75dB at 20Hz is just barely audible (to average hearing 18 year olds), equivalent in perceived equal loudness to 50dB at 40Hz and about 5dB at 1000Hz.
A quiet room will usually be above 25 dB...
More Xmax requires stiffer (heavier) cones, longer (heavier) higher power voice coils, limiting the useful upper frequency response range and reducing sensitivity.
In a sealed cabinet, SPL is simply related to excursion and Sd (cone area).
Doubling equally powered (same voltage applied to both) drivers in double the cabinet volume adds +6dB SPL.
Two drivers can produce the same low frequency SPL as one with half the excursion.
Four times the excursion is required for the same SPL for each octave drop, 40Hz to 20Hz.
http://www.baudline.com/erik/bass/xmaxer.html
Smaller boxes require more power to reach a given excursion.
If you want loud and even low frequency below around 80 Hz, better off using larger high excursion subwoofers that can be spread around the room and hidden.
Art
Last edited:
Maybe I wasn't clear. The set up gives F3 at ~40Hz. 20Hz is ~20dB down. This with the usual 2.83 vrms drive.
I have a plot of max SPL / max power which is limited by Xmax. At 20Hz this relates to an spl output of ~75dB. It's about 95dB at 40Hz which I think will be ok. So say a signal from the amp wants 95dB at 40Hz. Fine but if that happens at 20Hz whoops.
A discussion elsewhere demonstrated a certain driver with a higher Xmax limit and the simulation showed that Xmax was ok with a drive of 28w down to 10Hz. I can increase the Xmax on the above to the same level and it could be driven with >28w but not at 20Hz, same problem remains. It would have to be a much lower power level at 20Hz.
There are a few factors that make think that I am worry about something that wont happen ie the real Xmech being used due to the signal levels at 20Hz. One is other designs.
If I want to get wet I have to step in the water.
Yeah, if you are not using a high pass filter, you will definitely hear the "whoops" when the program material demands it.
Most low frequency drivers just make flapping noises much past Xmax, there is little "push" left regardless of input voltage when the coil leaves the gap. Vocals will sound like they are "gargling" before the cone goes that far.
It's a tricky area for someone who wants to get their feet wet for the first time in this area. I read for instance that 100 and something dB at 20Hz needs 1kw of amp. This leads to the question of who is going to produce music that needs that? Few could listen to it and who knows what it would do to their kit.
It's also possible to look at the bass range of musical instruments even a bass drum and the others. Also voice. Male bass >80Hz.
Then the chassis. A linear magnetic motor. More or less linear over it's stated Xmax but beyond that the efficiency in terms of cone movement will get worse and worse. How much pass.
Streamed movies. Some one mention that they have probably been filtered to limit the bass end to 30Hz. Dolby however state all kit used to produce will function at 20Hz. I don't use surround sound so wondered what these streamed movies used. Tricky but it seems if Dolby is being used there will be a logo on the screen. I have never seen one. Lots and lots of movies. They just chuck the effects channel, Bit of a problem noise wise if speech is set to usual listening levels, say 60dB. Vaguely related. One of our lesser known politician made a speech about actors mumbling in films. Has to stop. It can be hard to make out the words. Mix of diction and people knowing that the effects channel will blast out now and again. This seems to have got worse and worse over the years and bought kit lacks an effective loudness control.This used to be available on some analogue gear long ago.
It's also possible to look at the bass range of musical instruments even a bass drum and the others. Also voice. Male bass >80Hz.
Then the chassis. A linear magnetic motor. More or less linear over it's stated Xmax but beyond that the efficiency in terms of cone movement will get worse and worse. How much pass.
Streamed movies. Some one mention that they have probably been filtered to limit the bass end to 30Hz. Dolby however state all kit used to produce will function at 20Hz. I don't use surround sound so wondered what these streamed movies used. Tricky but it seems if Dolby is being used there will be a logo on the screen. I have never seen one. Lots and lots of movies. They just chuck the effects channel, Bit of a problem noise wise if speech is set to usual listening levels, say 60dB.
Vaguely related. One of our lesser known politician made a speech about actors mumbling in films. Has to stop. It can be hard to make out the words. Mix of diction and people knowing that the effects channel will blast out now and again. This seems to have got worse and worse over the years and bought kit lacks an effective loudness control.This used to be available on some analogue gear long ago.
A speaker with 70dB sensitivity at 20 Hz would require 1000 watts to produce 100dB SPL.
A large tapped horn subwoofer like the DSL DTS-20 has about 93dB sensitivity (half space, outdoors) down that low, using about 4 watts (5.65 volts) to produce 100dB SPL at 20Hz.
"Need" is a relative term- as a kid I listened to music on transistor radios that couldn't produce any meaningful level below 160Hz. I could certainly appreciate the difference between them and the piano and Hammond organ/Leslie speaker in the living room.
The low "B" note on a 5 string bass is 30.87 Hz, the low "A" on piano 27.5Hz.
I listen to a variety of music that has content at 25 Hz and below, for example the "Raising Sand" album by Robert Plant and Alison Krauss.
Enya's "Watermark" and "Long Ships" tracks have pedal notes in the 15 Hz range, like pipe organs.
Tracks like Bass Mekanik's "Bass I Love You" have notes that drop down to 7Hz, as low as the largest pipe organs.
Some movies have content below that.
With advancing age and loss of HF hearing, the octave(s) below 40Hz are more audible to me than the octave above 10kHz.
In most music, the harmonics are louder than the fundamental, so you can still get a sense of low frequency even when it's not there, as the spacing between each harmonic is multiples of the fundamental.
That said, until you use subwoofers capable of low frequency at audible levels, you don't know what you are missing.
Art