@jaakkopetteri
Thanks for sharing. I wonder who Speaker Symmetry is? The measurements from DATS-LA is interesting, but not the measurements taken in REW- the distortion measurement is affected by noise. (use of a miniDSP Umik-1, as suggested by the 7 digit microphone calibration file number)
Here's my measurement taken some time ago, in the nearfield, with 2 different microphones.
The level of imprecision starts to increase somewhere between 100-200Hz and fluctuates even more as the frequency gets higher. The usual rule of thumb of 4311/diameter of cone" = 4311 / 10" = 431Hz is a kind of soft boundary.
The upside of taking a nearfield measurement is that the signal to noise ratio is higher, so one can actually peek into the harmonic distortion more clearly. The downside is that the microphone is exposed to some 110dB... which may be result in microphone self distortion.
Here's the harmonics as measured by the Sonarworks Xref20 and Earthworks M23:
It unclear how important these measurement of harmonics is.
But for a start, we need to measure them properly if we are going to make meaningful correlations.
I have spoken about this previously here:
Here are 13 speakers measured by Sean Olive:
Unfortunately haven't found the original study. If anyone can point me to it, I would be most appreciative.
I'd like to know the measurement conditions. Were there limitations to the microphone, amplifier or measurement method? By method I mean- was this the Maximum Length Sequence (MSL- pre-1999) method, or log chirp (aka Sine Sweep- post 2000, after Alberto Farina presented his paper at AES)
Did anyone apart from @5th element notice that the speaker with highest preference rating also had the highest distortion rating (lowest distortion overall). 🥷
And for speaker L12, it's maximum 2nd order is lower than it's mean.
"I'm no maths major, but but how can the peak be less than the average?"
From my point of view, how did the measurement even manage to register a maximum H3 of 50 to 500 % for the 12 speakers?
Suppose it not due microphone or method, and this distortion is of the bass region, where distortion is typically the highest.
Measuring bass distortion is no trivial task.. and is reason why there are different proposals ow how to do it more accurately eg. CEA2010, CEA2010B, CTA2024-A.
Thanks for sharing. I wonder who Speaker Symmetry is? The measurements from DATS-LA is interesting, but not the measurements taken in REW- the distortion measurement is affected by noise. (use of a miniDSP Umik-1, as suggested by the 7 digit microphone calibration file number)
Here's my measurement taken some time ago, in the nearfield, with 2 different microphones.
The level of imprecision starts to increase somewhere between 100-200Hz and fluctuates even more as the frequency gets higher. The usual rule of thumb of 4311/diameter of cone" = 4311 / 10" = 431Hz is a kind of soft boundary.
The upside of taking a nearfield measurement is that the signal to noise ratio is higher, so one can actually peek into the harmonic distortion more clearly. The downside is that the microphone is exposed to some 110dB... which may be result in microphone self distortion.
Here's the harmonics as measured by the Sonarworks Xref20 and Earthworks M23:
It unclear how important these measurement of harmonics is.
But for a start, we need to measure them properly if we are going to make meaningful correlations.
I have spoken about this previously here:
Here are 13 speakers measured by Sean Olive:
Unfortunately haven't found the original study. If anyone can point me to it, I would be most appreciative.
I'd like to know the measurement conditions. Were there limitations to the microphone, amplifier or measurement method? By method I mean- was this the Maximum Length Sequence (MSL- pre-1999) method, or log chirp (aka Sine Sweep- post 2000, after Alberto Farina presented his paper at AES)
Did anyone apart from @5th element notice that the speaker with highest preference rating also had the highest distortion rating (lowest distortion overall). 🥷
And for speaker L12, it's maximum 2nd order is lower than it's mean.
"I'm no maths major, but but how can the peak be less than the average?"
From my point of view, how did the measurement even manage to register a maximum H3 of 50 to 500 % for the 12 speakers?
Suppose it not due microphone or method, and this distortion is of the bass region, where distortion is typically the highest.
Measuring bass distortion is no trivial task.. and is reason why there are different proposals ow how to do it more accurately eg. CEA2010, CEA2010B, CTA2024-A.
Last edited:
Not sure where the cross between affordable and not is but some options.
https://www.beyma.com/speakers/Fich...akers-data-sheet-low-mid-frequency-12BR70.pdf
SBacoustic/SBaudience has a good range of 12's
https://www.redcatt.net/web/product.html?nominal=12
https://www.punktkilde.com/search?products_search[0]=transducer_size_inches:(min:11,max:14)
https://stx.pl/en/w-32-500-8-mc.html
https://stx.pl/en/w-30-500-2x4-mc.html
https://www.ciare.com/en/products/lf-driver?category=lf-driver
Faital 12RS series if they are not out of budget range.
It's been verified previously-
Here.
and here,
From what I understand, it's been done since 1989, by Ken Kriesel in his push-pull subwoofers:
Designed by Kreisel, a new innovation in subwoofer technology: Push-Pull drivers, used in M&K’s new MX Series. Push-Pull is achieved by mounting one driver physically backwards in the cabinet and wiring it backwards as well, causing both driver cones to move mechanically in the same direction. The result is acoustic cancellation of even order harmonic distortion, cleaner, tighter and more accurate musical bass reproduction, and significantly improved high-output linearityReference: Miller and Kreisel timeline
Here's a modern incarnation-
You can do it in dipoles too:
Last edited:
Is that 82% or 70%?
Because even for subwoofers, I don't find 70% very useful anymore.
Things like symmetry and stability are very important as well.
The last one is very rarely shown at public LSI results.
Thanks Arez detailed as usual 🙂
Can you rank them a little?
Allways looking for candidates with
FS under 30 hz and xmax of +- 10 mm or more.
88 dB or higher, 200-400 watt and as small VAS as possible.
Use 4 scan speak discovery 12inch in 2 dos subs, the peerless xxls is almost the same design. Good drivers
That's would be the one. Same 4 ohm, alu cone.
Models the same, apart from minor sensitivity difference (-1dB). Though this could certainly be within margin of error that would be inconsequential in typical use as a 4 ohm subwoofer.
My 835017 free air impedance sweep:
Here's the Scan-
from Scan-Speak datasheet:
https://www.scan-speak.dk/datasheet/pdf/30w-4558t00.pdf
Indeed a close match to Tymphany datasheet eg.
The Scan has different basket and dust-cap (fibreglass). It may be a better performer in the midrange.
Unfortunately I do not have any samples to nitpick it's comparative performance, so I will refrain from further speculations.
Besides, if you are using below 150 Hz only (subwoofer applications) practical considerations may be more important eg. 4 ohms vs 8 ohms, availability etc.
Models the same, apart from minor sensitivity difference (-1dB). Though this could certainly be within margin of error that would be inconsequential in typical use as a 4 ohm subwoofer.
My 835017 free air impedance sweep:
Here's the Scan-
from Scan-Speak datasheet:
https://www.scan-speak.dk/datasheet/pdf/30w-4558t00.pdf
Indeed a close match to Tymphany datasheet eg.
The Scan has different basket and dust-cap (fibreglass). It may be a better performer in the midrange.
Unfortunately I do not have any samples to nitpick it's comparative performance, so I will refrain from further speculations.
Besides, if you are using below 150 Hz only (subwoofer applications) practical considerations may be more important eg. 4 ohms vs 8 ohms, availability etc.
Last edited:
Lessons:
1) SPL must be specified (apples to apples to comparison)
2) Microphone must be specified (different mics distort at different SPLs)
2) Process must be specified
ie. different results from calculations eg. Maximum Length Sequence eg. JustMLS, MLSSA, Clio,
Or sine sweep (Clio, ARTA, REW) or stepped sine (STEPS) or FSAF (REW 5.40b40)
3) if higher order harmonics are bunched up- suspect that the Noise floor is limiting ability to measure at.
Without this- beware of distortion measurements. Actual driver distortion is probably a lot lowerthan what's shown.
Last edited:
The Peerless xxls driver has a vas of 136liter the Scan Speak Discovery or Revelator subwoofers even more, arround 200liter vas.
Higher vas means looser surround and spider it wil play small details in music better.
Peerless stw-350 for example is a driver with heavy cone and 58 liter vas it wil be not that good for music, but way better for home cinema if you need that spl level.
But if you dont need the spl i think the xxls or discovery drivers wil also sound better with movies.
Higher vas means looser surround and spider it wil play small details in music better.
Peerless stw-350 for example is a driver with heavy cone and 58 liter vas it wil be not that good for music, but way better for home cinema if you need that spl level.
But if you dont need the spl i think the xxls or discovery drivers wil also sound better with movies.
I do not really understand this reasoning. Unless spider and surround are heavily unlinear the stiffness should not affect small details different from big ones.
(And yes, I know Bernd Timmermanns keeps repeating this)
All the moving soft parts in a speaker ie. spider; surround, even the cone can, and do make a difference, for sure.
Let’s go back to the original definition of Vas. It was defined publicly for the first time in 1961 by Neville Thiele of the Australian Broadcasting Corporation.
P being density of air
C being speed of sound
Sd- Equivalent piston area of the driver diaphragm
Cms- Compliance of the driver's suspension, in metres per newton
At first glance, it might appear that these are fixed quantities. However, we now know that in cone movement-
Cms is variable (Klippel et. Al)
Sd is variable (Steve Mowry)
So Vas is in fact variable with cone movement.
Footnote:
T/S parameters were designed help predict low frequency performance when putting a speaker into a particular cabinet size, and how large and long the bass reflex port (if it is used) should be.
Thiele’s two-part article is so important that it was later republished in the Journal of the Audio Engineering Society in 1971, 10 years after being published for the first time. Turns out he didn’t publish it in the AES until 1971, hoping he’d give Aussies a (10 year) head start.
Even the team at Peerless couldn’t get it to work. Perhaps it because “around 1962 Peerless drivers were using open weave dust caps. Meaning that air could penetrate through the dust cap, creating a lossy design”
It was Richard Small’s moving to AU to study that he was exposed to Neville work, and publishing in the AES that it started getting picked up in the 1970s.
Reference:
https://audioxpress.com/article/loudspeaker-bass-reflex-alignments
Let’s go back to the original definition of Vas. It was defined publicly for the first time in 1961 by Neville Thiele of the Australian Broadcasting Corporation.
P being density of air
C being speed of sound
Sd- Equivalent piston area of the driver diaphragm
Cms- Compliance of the driver's suspension, in metres per newton
At first glance, it might appear that these are fixed quantities. However, we now know that in cone movement-
Cms is variable (Klippel et. Al)
Sd is variable (Steve Mowry)
So Vas is in fact variable with cone movement.
Footnote:
T/S parameters were designed help predict low frequency performance when putting a speaker into a particular cabinet size, and how large and long the bass reflex port (if it is used) should be.
Thiele’s two-part article is so important that it was later republished in the Journal of the Audio Engineering Society in 1971, 10 years after being published for the first time. Turns out he didn’t publish it in the AES until 1971, hoping he’d give Aussies a (10 year) head start.
Even the team at Peerless couldn’t get it to work. Perhaps it because “around 1962 Peerless drivers were using open weave dust caps. Meaning that air could penetrate through the dust cap, creating a lossy design”
It was Richard Small’s moving to AU to study that he was exposed to Neville work, and publishing in the AES that it started getting picked up in the 1970s.
Reference:
https://audioxpress.com/article/loudspeaker-bass-reflex-alignments
Last edited:
This is true, the suspension obviously plays a big role in high-signal scenarios, when there is a a significant amount of cone excursion involved.
However, a non-linear compliance only has effect on lower frequencies (see Klippel, they have some very good papers on this).
Therefor does not, and will never effect non-linear inter-modulation distortion in the mid-range.
A non-linear Sd can have some effect on this.
Although, since the positive and negative excursion complement each other (unlike non-linear compliance and BL), this effect will be less strong.
When there is no cone excursion involved, none of this is an issue anymore = no/little intermodulation distortion.
What's left is issues from heat, power and current modulation.
So one of the best ways is just simply avoid excursion = bigger Sd and don't let your mid-range do low-end.
Reducing cone movement is a good thing for reducing distortion.
The problem is when you reduce it to zero, you get no sound (TM).
Everything else being constant (it never is), we know that by increasing cone area, we can increase
maximum SPL <-> dynamic range.
The challenge then is how to increase SPL? More magnet. Or more cone?
A larger cone will increase directivity <-> reduces dispersion.
(This is why a large cone has diminished high frequency ability, particularly off axis).
Of course, whether this is a good or bad thing, depends on your goals.
***
Do you want to direct your sound?
Or you want to spray it around and let it wash over you?
Do you want to minimize the effects of the room, or have a bit of it, or a lot of it?
Well that depends on your goals eg. outdoors, limited boundaries, high background noise vs.
indoors- boundaries, low background noise.
Or the kind of music you listen to eg. classical (purely acoustic, made organically, captured and recorded by microphones in rooms <----> electronic (synthetic "made in the (computer) box", no organic equivalent exists)
Indoor commercial (think theatre, auditorium) or domestic speakers (home theater, living room, apartment corner, bedroom) will interact with their room they are in.
Headphones- no room influence (the boundaries are now the external ear (pinnae, HRTF), with lower background noise, lower SPL needed.
In-ear phones - ear canal are now boundaries, even lower background noise, lower SPL needed
Direct (Vestibulocochlear/CN VIII) nerve conduction- bypass ears completely, no boundaries, no mechanical to electrical conversion necessary, highest signal to noise.
Note:
At a whopping 105dB@1m ie. lifelike / THX /Atmos "reference" levels,
At 80Hz the cone excursion is just 1mm for the 12TXB100
At 80Hz, the cone excursion is 3mm for the 830845
If you don't want separate subwoofers than the 830845 is a better solution, as it is 9dB higher SPL at 80Hz.
But with the use of separate multiple subwoofers, a 12TBX100 in a 50L closed cabinet is a better solution from 100Hz upwards is better from an purely acoustics point of view. The SPL and dynamic range will be higher, the amplitude modulation distortion will be lower.
In C21, when real estate is increasingly expensive, but amplifier watts is cheap.
The separate subwoofers + mains approach is the right approach, if we are listening to music via speakers.
Bose Loudspeakers
Miller and Kriesel Loudspeakers
Geddes Loudspeakers got it right decades ago. @gedlee
The problem is when you reduce it to zero, you get no sound (TM).
Everything else being constant (it never is), we know that by increasing cone area, we can increase
maximum SPL <-> dynamic range.
The challenge then is how to increase SPL? More magnet. Or more cone?
A larger cone will increase directivity <-> reduces dispersion.
(This is why a large cone has diminished high frequency ability, particularly off axis).
Of course, whether this is a good or bad thing, depends on your goals.
***
Or you want to spray it around and let it wash over you?
Do you want to minimize the effects of the room, or have a bit of it, or a lot of it?
Well that depends on your goals eg. outdoors, limited boundaries, high background noise vs.
indoors- boundaries, low background noise.
Or the kind of music you listen to eg. classical (purely acoustic, made organically, captured and recorded by microphones in rooms <----> electronic (synthetic "made in the (computer) box", no organic equivalent exists)
Indoor commercial (think theatre, auditorium) or domestic speakers (home theater, living room, apartment corner, bedroom) will interact with their room they are in.
Headphones- no room influence (the boundaries are now the external ear (pinnae, HRTF), with lower background noise, lower SPL needed.
In-ear phones - ear canal are now boundaries, even lower background noise, lower SPL needed
Direct (Vestibulocochlear/CN VIII) nerve conduction- bypass ears completely, no boundaries, no mechanical to electrical conversion necessary, highest signal to noise.
***
I recently simulated a 12" B&C midwoofer in a 50L cabinet, similar to what I used the 830845 in (50L)
Note:
At a whopping 105dB@1m ie. lifelike / THX /Atmos "reference" levels,
At 80Hz the cone excursion is just 1mm for the 12TXB100
At 80Hz, the cone excursion is 3mm for the 830845
If you don't want separate subwoofers than the 830845 is a better solution, as it is 9dB higher SPL at 80Hz.
But with the use of separate multiple subwoofers, a 12TBX100 in a 50L closed cabinet is a better solution from 100Hz upwards is better from an purely acoustics point of view. The SPL and dynamic range will be higher, the amplitude modulation distortion will be lower.
In C21, when real estate is increasingly expensive, but amplifier watts is cheap.
The separate subwoofers + mains approach is the right approach, if we are listening to music via speakers.
Bose Loudspeakers
Miller and Kriesel Loudspeakers
Geddes Loudspeakers got it right decades ago. @gedlee
Last edited: