I had great sound with the SB 26ADC in an 5" Augerpro waveguide, crossed at around 2200Hz LR24, with the SB MW13TX 👍
Only put them aside, because I jumped KEF coax instead 🤓
The woofers, that I use now, are two SB Satori WO24P in a combined 75 liters closed cabinet. Best bass I've had yet in my apartment
Only put them aside, because I jumped KEF coax instead 🤓
The woofers, that I use now, are two SB Satori WO24P in a combined 75 liters closed cabinet. Best bass I've had yet in my apartment
@flex2
Please excuse me again:
I have checked with more attenction the filter to compensate the impedance here:
http://www.troelsgravesen.dk/BlieSMa_T34A-4.htm
At 2 Khz the peak of impedance is relavant, but at 3 Khz seem almost flat.
In your opinion if I cut to 3 Khz instead of 2 Khz I could avoid adding the RLC filter ? Or would it be better to add it anyway ?
Please excuse me again:
I have checked with more attenction the filter to compensate the impedance here:
http://www.troelsgravesen.dk/BlieSMa_T34A-4.htm
At 2 Khz the peak of impedance is relavant, but at 3 Khz seem almost flat.
In your opinion if I cut to 3 Khz instead of 2 Khz I could avoid adding the RLC filter ? Or would it be better to add it anyway ?
No excuses please...
I will give you my opinion if you want it.
The best is to add it, but I understand your concern.
Since midrange impedance is very flat you can expect true and full performance from the filter and that also means linear phase in the cross over region.
With a rising impedance caused by the phase response won't be linear and for a perfect filter the phase change for both midrange and tweeter must be as parallell and linear as possible.
If tweeter don't have the LCR you might have a -9dB/octave filter or even -6dB and then you will end up with a peak of 2-3 dB before the crossover point and a 2-3dB dip after. That will not sound good.
So linearity in both frequency response and in phase response in the crossover is important and phase slopes between mid and tweeter must be as parallell and linear over at least 2 octaves on each side of the crossover point.
A sneak version would be to make hybrid filter between -6 and -12 for the LP mid by increasing inductance, reducing capacitance and put a small resistor in series with the capacitor. The important thing is that you measure phase and keep them parallell.
That is the only way to have a straight frequency over the crossover point.
You do not need to have esoteric parts in this LCR circuit. Even bipolar caps are OK.
I also want to give my opinion on quality of crossover parts for the bass. Just for you to think about.
I have never been able to detect any benefit in sound quality between foil inductors or round solid core air inductors.
It's only for midrange a tweeter I noticed a more open and transparent sound with foil inductors and higher quality capacitors.
Thinking twice over your question, yes I think you should have the LCR link.
You have the potential to build a very good loudspeaker here, so no compromises.
To keep cost down you can use bipolar caps and iron core inductor if you want since only a small part of the energy will run through this circuit.
Most energy will run through the driver and parallell damping resistor since this tweeter needs to be attenuated around 7-9dB to match midrange.
I think 3kHz crossover will create other problems and would not go there. My choice would be to start at 2,5kHz and go downwards if necessary.
But I have very much components at home for trial and development purpouses and can change according to where I want to go.
Never ever have I hit a true -6 or -12dB filter. There is always tradeoffs and tweaks that needs to be done to get the best final result.
If you don't have that and must get it right on first trial you better spend some time with a cross over simulatior and build with cheap part until you are satisfied with the result. After that you can change to your desired esoteric components.
Inductors can be coupled in series and capacitors can be parallelled so you will benefit from having a few smaller values at home to adjust.
Great drivers is good to have, but it is in the cross over the magic happends, so I concider cross over design to be more important to over all sound than the drivers themself.
Usually I spend several weekends fine tuning crossover until satisfied.
For the mid/tweeter crossover you could use gated measurement using both drivers together placing microphone at a distance of 1/2 to 1 meter to verify smooth transition. This way you will detect unlinear phase problems.
I will give you my opinion if you want it.
The best is to add it, but I understand your concern.
Since midrange impedance is very flat you can expect true and full performance from the filter and that also means linear phase in the cross over region.
With a rising impedance caused by the phase response won't be linear and for a perfect filter the phase change for both midrange and tweeter must be as parallell and linear as possible.
If tweeter don't have the LCR you might have a -9dB/octave filter or even -6dB and then you will end up with a peak of 2-3 dB before the crossover point and a 2-3dB dip after. That will not sound good.
So linearity in both frequency response and in phase response in the crossover is important and phase slopes between mid and tweeter must be as parallell and linear over at least 2 octaves on each side of the crossover point.
A sneak version would be to make hybrid filter between -6 and -12 for the LP mid by increasing inductance, reducing capacitance and put a small resistor in series with the capacitor. The important thing is that you measure phase and keep them parallell.
That is the only way to have a straight frequency over the crossover point.
You do not need to have esoteric parts in this LCR circuit. Even bipolar caps are OK.
I also want to give my opinion on quality of crossover parts for the bass. Just for you to think about.
I have never been able to detect any benefit in sound quality between foil inductors or round solid core air inductors.
It's only for midrange a tweeter I noticed a more open and transparent sound with foil inductors and higher quality capacitors.
Thinking twice over your question, yes I think you should have the LCR link.
You have the potential to build a very good loudspeaker here, so no compromises.
To keep cost down you can use bipolar caps and iron core inductor if you want since only a small part of the energy will run through this circuit.
Most energy will run through the driver and parallell damping resistor since this tweeter needs to be attenuated around 7-9dB to match midrange.
I think 3kHz crossover will create other problems and would not go there. My choice would be to start at 2,5kHz and go downwards if necessary.
But I have very much components at home for trial and development purpouses and can change according to where I want to go.
Never ever have I hit a true -6 or -12dB filter. There is always tradeoffs and tweaks that needs to be done to get the best final result.
If you don't have that and must get it right on first trial you better spend some time with a cross over simulatior and build with cheap part until you are satisfied with the result. After that you can change to your desired esoteric components.
Inductors can be coupled in series and capacitors can be parallelled so you will benefit from having a few smaller values at home to adjust.
Great drivers is good to have, but it is in the cross over the magic happends, so I concider cross over design to be more important to over all sound than the drivers themself.
Usually I spend several weekends fine tuning crossover until satisfied.
For the mid/tweeter crossover you could use gated measurement using both drivers together placing microphone at a distance of 1/2 to 1 meter to verify smooth transition. This way you will detect unlinear phase problems.
I had great sound with the SB 26ADC in an 5" Augerpro waveguide, crossed at around 2200Hz LR24, with the SB MW13TX 👍
Only put them aside, because I jumped KEF coax instead 🤓
The woofers, that I use now, are two SB Satori WO24P in a combined 75 liters closed cabinet. Best bass I've had yet in my apartment
Good info, and yes, SB 26ADC is a good tweeter. Have tried it too but find it a little coloured because of a bit lacking damping material compared to my prefered tweeter Seas 27TBC/G.
Include a simulation of your design with two WO24P (I assumed parallelled 8 ohm) and how it compares against the Monacor SPH250 KE that the tread owner intend to use in his design.
Red curve Monacor SPH250 KE in 74 liter bass reflex
Blue curve Monacor SPH250 KE in 74 liter sealed box
Green curve 2 x SB Satori WO24P 8 ohm in 75 liter sealed box. Very close to my taget curve for room at 45m2 (not included though).
Gray curve my target for 30m2 room.
Fully understand that you are satisfied with the performance of your Satoris...
How big is the room they are standing in ???
Thank you @flex2, now I understand how are important these factors that to be honest I didn't think they carried so much weight.
Please correct me I mistake:
Suppose 2 different cases:
1) Design a crossover normally with VituixCAD and Xsim with cut to 400Hz. At the end both software show an almost flat response, but these software are unable to calculate the baffle step deviation with accuracy.
2) Design a crossover with VituixCAD and Xsim with woofer cut to 200Hz and the Midange 400Hz. At the end both software show a non flat response.
In the first case when I put the my mic vertically (becouse Beyerdynamic MM1 it is a polarized condenser microphone) at 50cm distance at the center between woofer and midrange, in theory I should notice a peak at approximately around 300-500Hz of about 3db or more.
In the second case I should not notice a peak at approximately around 300-500Hz but there is a "lost" about 200/300 hz.
I need to increasy the crossover cut to find the right compromise and minimize these effects.
It is correct ?
Yes, I think you are right.
Designing crossover is an art where you DO NOT follow rules but tweak them to a good final result.
Here a little description about how to (eventually) crossover during a baffle step compensation of 4 dB. Teoretically 6dB, but in practice 3-4dB.
Straight green = midrange level 4 dB lower than bass driver level to compensate for the transition between half space and full space distribution.
Straight red = Bass driver level without crossover.
Blue line is our target to compensate for the loss of front projected energy and keep a straight frequency response in our room and listening position.
As you can see in this draft bass driver will put out too much energy (yellow field) by it self since cross over point will be higher that expected 400 Hz.
My estimation for 200 Hz might have been a bit low, but the crossover point for the bass needs to be at a lower frequency.
More like the yellow line, and from the draft we can read approx. 300 Hz (Black arrow).
There might still be to much energy from the bass driver below 200 Hz, and that can be handled by increasing the inductance of the the bass drivers
low pass filter to reduce the energy and hit the pink line.
But this is only a idea for a starting point.
Right pink curve is of course tweeter.
At -20 to -24dB drivers stop contribute to the output so there is little need to care about what is happening after the -24dB point has been passed.
Sketch based on 2nd order filters which have -6dB at cross over point and midrange with reversed polarity.
Designing crossover is an art where you DO NOT follow rules but tweak them to a good final result.
Here a little description about how to (eventually) crossover during a baffle step compensation of 4 dB. Teoretically 6dB, but in practice 3-4dB.
Straight green = midrange level 4 dB lower than bass driver level to compensate for the transition between half space and full space distribution.
Straight red = Bass driver level without crossover.
Blue line is our target to compensate for the loss of front projected energy and keep a straight frequency response in our room and listening position.
As you can see in this draft bass driver will put out too much energy (yellow field) by it self since cross over point will be higher that expected 400 Hz.
My estimation for 200 Hz might have been a bit low, but the crossover point for the bass needs to be at a lower frequency.
More like the yellow line, and from the draft we can read approx. 300 Hz (Black arrow).
There might still be to much energy from the bass driver below 200 Hz, and that can be handled by increasing the inductance of the the bass drivers
low pass filter to reduce the energy and hit the pink line.
But this is only a idea for a starting point.
Right pink curve is of course tweeter.
At -20 to -24dB drivers stop contribute to the output so there is little need to care about what is happening after the -24dB point has been passed.
Sketch based on 2nd order filters which have -6dB at cross over point and midrange with reversed polarity.
@flex2 - you are a generous and patient contributor and the kind of person who helps make DIYAudio one of the very best places to learn about this challenging and rewarding pursuit. Tak!
@flex2
Many thanks for your very detailed explanation.
I assume this is electrical response that is similar to mine apart that with tweeter I have used 3° order filter and I have reversed the polarity of woofer.
It is clear that is prevision based on theory.
But I suppose the the real test must be done with measurement.
My question is: what should I expect (and should I see) if I measure with my mic if I cut to woofer and midrange at 400 Hz ?
As you teach: one thing is what you see in a computer-made simulation and another thing is to actually see it with actual measurements so you can adjust by seeing practically where the problem is and exactly to what measure.
Can you please send an example of therorical measure (cutted at 400 Hz) and another with woofer cutted at (200 Hz) ?
In theory both cases must be wrong and 300 hz can be the right compromise.
I have asked about the misuration becouse when you measure at 50cm or 1m environmental factors also come into play that could confuse the test.
So understand exactly what you must expect can be very useful.
I actually started with the Seas DXT, with a lot of inspiration from Heissmann acoustics 😀 What I found, was that the speakers I prefer the most, are the ones with an even response and dispersion - no matter the price or brand, so that is my reasoning to go that route. The ADC tweeter allowed me to play with different sizes of waveguides, best fitting the given midrange and cross over frequency.
My living room is around 50 sqm, but I always use subwoofers - 4 in my case - so the tactile feeling and impact of the sound, emanates from a combination of 2 x 12" + 2 x 15" subwoofers with 4 x 700W in 8 ohm and DSP behind them, slightly overlapping with the 4 WO24P's. The Satori plays quite deep on their own, but my apartment have a rather soft wooden floor, that somehow eats up way more bass than most "harder" rooms.
I used 4 SB 23NRX before - but I can easily say, that even though you might simulate and equal response with them under the same conditions - they play totally different. Two identical closed cabinets - one with the two NRX and one with two Satori, reveals a lack of "snap" with the NRX. The Satori seems much easier to EQ and give a realistic fullness to the sound of larger instruments and drums - which I simply can't make the NRX do. My only conclusion is, that the Satori is a better unit in general, and simply outperform the NRX - no matter what I try to do with the cabinet, power, EQ or placement.
I have no solid answer for you since I do not know if you will try to adopt a baffle step compensation or not.
The last sketch I sent the goal would be a +4dB in bass compared to midrange with a transition from around 100Hz to 700 Hz.
The goal would be a frequency response that are similar to the blue target curve.
OK, thank you please confim if I have understand.
In this case the baffle step compensation is raccomandated becouse the woofer is not closed to midrange.
But if for example the woofer is only to 2cm from the midrange in theory the baffle step compensation is not so important do you confirm ?
In this case the baffle step compensation is raccomandated becouse the woofer is not closed to midrange.
But if for example the woofer is only to 2cm from the midrange in theory the baffle step compensation is not so important do you confirm ?
I'm sorry, but no. It has nothing to do with placement of drivers and it has nothing to do with time alignment which I think you are talking about here.
Just for you to fully understand the concept of baffle step compensation.
Lets say you have a perfect frequency linear loudspeaker from 20-20000 Hz when measuring near field as you do.
When you measure like that you do not include the effect of the baffle.
At 10 kHz the wave length are 3.4 cm which is smaller than the baffles width of let say 25 cm.
So the baffle works as a wall that only permits sound to be distributed forward and not backwards behind the driver.
At some point the wave length is equal to the baffles width, and with a width of 25 cm that occurs at 1360 Hz, and when you comes to half the wave length at 680 Hz the driver starts to spread it's energy backwards behind the loudspeaker.
This is a gradual process down to around 100Hz where the driver spreads it's energy 360° around the speaker and equally much behind the loudspeaker as in front of the loudspeaker into the room.
This means that the energy you measured near field will be spread in full space 360° pattern below 100 Hz and in half space 180° above 1360 Hz.
When measuring from far field and the listening position you theoretically will have 6dB lower frequency response below 100 Hz compared to frequency above 1360 Hz, but in practice it's around 4 dB because of rooms reflective surfaces and reverb.
The region between 100 Hz to 1360 Hz is the transition region where spread pattern gradually changes from 180° to 360°.
In the region from half wave length (680 Hz) to full wave length (1360 Hz) of the baffle width the edges of the baffle will contribute to the energy sent out and that is what we call diffraction, and diffraction is also energy sent out, so that is why the raise in amplitude better is calculated from half wave at 680 Hz or slightly above like 750 Hz. To minimize diffraction, edges of the baffle should be rounded or chamfered to reduce peaks and dips in this region created by out of phase energy.
So to be able to have a straight frequency response inside the room at listening position you need to compensate for this loss and have 4 dB higher amplitude below 100 Hz that gradually decreases to meet the midrange at 4 db lower amplitude at 750 Hz and then midrange and tweeter should be a straight line to 20000 Hz. Remember that this must be measured near field to work out properly.
Look at the sketch I sent you. That will help you understand the concept of baffle step compensation.
Just for you to fully understand the concept of baffle step compensation.
Lets say you have a perfect frequency linear loudspeaker from 20-20000 Hz when measuring near field as you do.
When you measure like that you do not include the effect of the baffle.
At 10 kHz the wave length are 3.4 cm which is smaller than the baffles width of let say 25 cm.
So the baffle works as a wall that only permits sound to be distributed forward and not backwards behind the driver.
At some point the wave length is equal to the baffles width, and with a width of 25 cm that occurs at 1360 Hz, and when you comes to half the wave length at 680 Hz the driver starts to spread it's energy backwards behind the loudspeaker.
This is a gradual process down to around 100Hz where the driver spreads it's energy 360° around the speaker and equally much behind the loudspeaker as in front of the loudspeaker into the room.
This means that the energy you measured near field will be spread in full space 360° pattern below 100 Hz and in half space 180° above 1360 Hz.
When measuring from far field and the listening position you theoretically will have 6dB lower frequency response below 100 Hz compared to frequency above 1360 Hz, but in practice it's around 4 dB because of rooms reflective surfaces and reverb.
The region between 100 Hz to 1360 Hz is the transition region where spread pattern gradually changes from 180° to 360°.
In the region from half wave length (680 Hz) to full wave length (1360 Hz) of the baffle width the edges of the baffle will contribute to the energy sent out and that is what we call diffraction, and diffraction is also energy sent out, so that is why the raise in amplitude better is calculated from half wave at 680 Hz or slightly above like 750 Hz. To minimize diffraction, edges of the baffle should be rounded or chamfered to reduce peaks and dips in this region created by out of phase energy.
So to be able to have a straight frequency response inside the room at listening position you need to compensate for this loss and have 4 dB higher amplitude below 100 Hz that gradually decreases to meet the midrange at 4 db lower amplitude at 750 Hz and then midrange and tweeter should be a straight line to 20000 Hz. Remember that this must be measured near field to work out properly.
Look at the sketch I sent you. That will help you understand the concept of baffle step compensation.
Thank you for this explanation
"The region between 100 Hz to 1360 Hz is the transition region where spread pattern gradually changes from 180° to 360°.
In the region from half wave length (680 Hz) to full wave length (1360 Hz) of the baffle width the edges of the baffle will contribute to the energy sent out and that is what we call diffraction, and diffraction is also energy sent out, so that is why the raise in amplitude better is calculated from half wave at 680 Hz or slightly above like 750 Hz. To minimize diffraction, edges of the baffle should be rounded or chamfered to reduce peaks and dips in this region created by out of phase energy."
I have looked the link that you have sent me.
There is a clear example about what happen if the edges are flat and what happen if are rounded or chamfered.
My actual speakers have edges rounded and chamfered and certainly the next speakers will build in the some way.
But this reduce the problem not solve it.
However the solutions proposed by that link refer to how to design the box to prevent the problem, but do not refer to possible interventions on the crossover.
I like to share my experience about one mistake that I have done:
After design a crossover with relative flat response, after do measurement I saw a rather high peak at about 200 Hz, that in my simulation there isn't.
So I was thinked this peak depend by a insufficient damping materials.
I was added more damping material and the peak was reduced.
If I had more experience I would not have made this mistake.
But at the sometime, if I had not made an accurate measurement at only few cm distance perhaps that peak would have been covered by any ambiant type resonances and I wouldn't have even noticed the mistake.
My general life experience tells me that theory is important but by itself is not enough.
The electrical response of the crossover actually I don't consider it much, because unfortunately speakers are nonlinear transducers and therefore cannot be treated as simple “resistors”.
What I have always been advised by everyone has been to take accurate measurements and take action based on what the measurement says.
My opinion regarding baffle step compensation is to try to work on the causes as much as possible (as suggested by the link you posted) and not on the effect.
But if I really have to intervene on the effect, it is the measurements that can tell me whether and to what extent to intervene not the crossover electrical response that do not take into account the complex principles of acoustics.
In the case of the damping material error, it was a stupid error that was easy to detect and solve.
The case on the other hand of baffle step compensation does not seem so easy to interpret, looking at other kit designs for example those made available by visaton (where even the software I used Boxsim allows you to enter the box the size and speakers position in 3d) the simulated response was flat.
Maybe because their software to some way in their calculation they also considered the effect of the baffle step, or maybe in their kits they did not consider this factor.
Hope you undestand my point of view and thank you again for your patient.
"The region between 100 Hz to 1360 Hz is the transition region where spread pattern gradually changes from 180° to 360°.
In the region from half wave length (680 Hz) to full wave length (1360 Hz) of the baffle width the edges of the baffle will contribute to the energy sent out and that is what we call diffraction, and diffraction is also energy sent out, so that is why the raise in amplitude better is calculated from half wave at 680 Hz or slightly above like 750 Hz. To minimize diffraction, edges of the baffle should be rounded or chamfered to reduce peaks and dips in this region created by out of phase energy."
I have looked the link that you have sent me.
There is a clear example about what happen if the edges are flat and what happen if are rounded or chamfered.
My actual speakers have edges rounded and chamfered and certainly the next speakers will build in the some way.
But this reduce the problem not solve it.
However the solutions proposed by that link refer to how to design the box to prevent the problem, but do not refer to possible interventions on the crossover.
I like to share my experience about one mistake that I have done:
After design a crossover with relative flat response, after do measurement I saw a rather high peak at about 200 Hz, that in my simulation there isn't.
So I was thinked this peak depend by a insufficient damping materials.
I was added more damping material and the peak was reduced.
If I had more experience I would not have made this mistake.
But at the sometime, if I had not made an accurate measurement at only few cm distance perhaps that peak would have been covered by any ambiant type resonances and I wouldn't have even noticed the mistake.
My general life experience tells me that theory is important but by itself is not enough.
The electrical response of the crossover actually I don't consider it much, because unfortunately speakers are nonlinear transducers and therefore cannot be treated as simple “resistors”.
What I have always been advised by everyone has been to take accurate measurements and take action based on what the measurement says.
My opinion regarding baffle step compensation is to try to work on the causes as much as possible (as suggested by the link you posted) and not on the effect.
But if I really have to intervene on the effect, it is the measurements that can tell me whether and to what extent to intervene not the crossover electrical response that do not take into account the complex principles of acoustics.
In the case of the damping material error, it was a stupid error that was easy to detect and solve.
The case on the other hand of baffle step compensation does not seem so easy to interpret, looking at other kit designs for example those made available by visaton (where even the software I used Boxsim allows you to enter the box the size and speakers position in 3d) the simulated response was flat.
Maybe because their software to some way in their calculation they also considered the effect of the baffle step, or maybe in their kits they did not consider this factor.
Hope you undestand my point of view and thank you again for your patient.
Last edited:
No problem, you are welcome.
I see that you still resist the concept and do not accept it. It's fine and there might be a good reason for it.
What you can do is to measure one driver and increase distance between driver and microphone.
Start at near field about 5 cm from the driver.
Increase distance to 20cm, 50cm, 80cm, 120cm and 150cm and then smooth the curves and scale them to be equal in response over 1000Hz.
When you then look at the curves you will see that bass increase depth and keep the amplitude better than the region 100 Hz to 700 Hz.
That way you will see what happends in this region and that amplitude is falling more rapidly than below 100 and above 1000 Hz.
The curves you see presented for different kits at Visaton are far field measurement and then they should be linear.
If you measure them in near field they will look different and probably quite close to the sketch I made.
If you design your loudspeaker and aim at the straightest frequency response with near field measurement you will have loss of amplitude in the bass.
Is this maybe why you like some 4-6 dB higher bass than shown in simulations?
Is this the reason you like Monacor SPH250 KE in bass reflex configuration instead of sealed as it was designed for?
You need that excessive bass amplitude to compensate total bass level ?
With that extra bass you might be close to a straight in room response.
But then you might have a dip in the 150-300 Hz region where output from bass reflex tube is less.
I think you will find your way of designing and you should do it your way.
I would have compensated the full 4dB in bass level and for sure selected the sealed application for SPH250 KE that have 3-5 dB less output in the lower bass region.
That way I believe you would end with a more balanced and linear bass response also in the 150-300 Hz region.
You know what...
I will make some measurements that show you the effect later today.
I see that you still resist the concept and do not accept it. It's fine and there might be a good reason for it.
What you can do is to measure one driver and increase distance between driver and microphone.
Start at near field about 5 cm from the driver.
Increase distance to 20cm, 50cm, 80cm, 120cm and 150cm and then smooth the curves and scale them to be equal in response over 1000Hz.
When you then look at the curves you will see that bass increase depth and keep the amplitude better than the region 100 Hz to 700 Hz.
That way you will see what happends in this region and that amplitude is falling more rapidly than below 100 and above 1000 Hz.
The curves you see presented for different kits at Visaton are far field measurement and then they should be linear.
If you measure them in near field they will look different and probably quite close to the sketch I made.
If you design your loudspeaker and aim at the straightest frequency response with near field measurement you will have loss of amplitude in the bass.
Is this maybe why you like some 4-6 dB higher bass than shown in simulations?
Is this the reason you like Monacor SPH250 KE in bass reflex configuration instead of sealed as it was designed for?
You need that excessive bass amplitude to compensate total bass level ?
With that extra bass you might be close to a straight in room response.
But then you might have a dip in the 150-300 Hz region where output from bass reflex tube is less.
I think you will find your way of designing and you should do it your way.
I would have compensated the full 4dB in bass level and for sure selected the sealed application for SPH250 KE that have 3-5 dB less output in the lower bass region.
That way I believe you would end with a more balanced and linear bass response also in the 150-300 Hz region.
You know what...
I will make some measurements that show you the effect later today.
Last edited:
Hi @marcop999,
just a quick remark - be aware that monacor sph250ke will not be produced anymore. The taiwanese manufacturer stops (or already stopped) making them. If you are quick you will still get a pair somewhere.
(German) reference here:
https://www.donhighend.de/?p=10273
just a quick remark - be aware that monacor sph250ke will not be produced anymore. The taiwanese manufacturer stops (or already stopped) making them. If you are quick you will still get a pair somewhere.
(German) reference here:
https://www.donhighend.de/?p=10273
Last edited:
Thank you @flex2,
Now is more clear the concept.
"The curves you see presented for different kits at Visaton are far field measurement and then they should be linear."
I was not referring to the measured curves, but to those simulated by Visaton's Boxsim design software.
At 3cm the measured curve of my woofer still conditioned by external resonances in fact the measurements of the two channels quite different.
In the past I have tried measuring at greater distances but came up peaks very high, dependent to a greater extent on the surrounding environment.
My measures show that the reflex mainly in the 40-200 Hz range, beyond which the crossover filter intervenes which cuts the woofer.
I like the bass at 60-100 Hz, above I would like it to flatten as it happens in my measurements
So at 3cm of distance my measure don't match with your drawing but perhaps if, as you say, if I measure from a greater distance I could have a similar answer to your drawing.
But if this were to happen the point is to understand what the ideal situation should be.
Maybe a perfect speaker should show the same response even at greater distances?
Here comes the difference between the theory and what then happens in practice.
Now is more clear the concept.
"The curves you see presented for different kits at Visaton are far field measurement and then they should be linear."
I was not referring to the measured curves, but to those simulated by Visaton's Boxsim design software.
At 3cm the measured curve of my woofer still conditioned by external resonances in fact the measurements of the two channels quite different.
In the past I have tried measuring at greater distances but came up peaks very high, dependent to a greater extent on the surrounding environment.
My measures show that the reflex mainly in the 40-200 Hz range, beyond which the crossover filter intervenes which cuts the woofer.
I like the bass at 60-100 Hz, above I would like it to flatten as it happens in my measurements
So at 3cm of distance my measure don't match with your drawing but perhaps if, as you say, if I measure from a greater distance I could have a similar answer to your drawing.
But if this were to happen the point is to understand what the ideal situation should be.
Maybe a perfect speaker should show the same response even at greater distances?
Here comes the difference between the theory and what then happens in practice.
Last edited:
So an example of baffle step effect:
Focus on response out of 1 kHz as reference point. @60cm 91,5 dB, @120cm 88 dB, @240cm ~85dB.
So -3dB for each doubling of distance as you would expect.
Basically you have the same relationship of -3dB for each doubling of distance @ 100 Hz and below.
But not so at 150 and 200 Hz. Almost -3dB between 120 and 240 cm which is to be expected since we basically are measuring in free field.
@60cm we are more in the near field area and and have a drop of 5 dB between 60cm and 120cm.
That is 2 dB extra.
Unfortunately I had no measurement at 30 cm but believe me when I say that we would have even clearer difference of -6dB.
At 230/250 cm distance you also can see room gain have kicked in.
With near field measurement you do not see the effect of baffle step and room gain, so that is the reason one have to set up a target curve for near field measurement that as good as possible compensate for the baffle step loss and added bass because of room once the loudspeaker are placed in a room.
Focus on response out of 1 kHz as reference point. @60cm 91,5 dB, @120cm 88 dB, @240cm ~85dB.
So -3dB for each doubling of distance as you would expect.
Basically you have the same relationship of -3dB for each doubling of distance @ 100 Hz and below.
But not so at 150 and 200 Hz. Almost -3dB between 120 and 240 cm which is to be expected since we basically are measuring in free field.
@60cm we are more in the near field area and and have a drop of 5 dB between 60cm and 120cm.
That is 2 dB extra.
Unfortunately I had no measurement at 30 cm but believe me when I say that we would have even clearer difference of -6dB.
At 230/250 cm distance you also can see room gain have kicked in.
With near field measurement you do not see the effect of baffle step and room gain, so that is the reason one have to set up a target curve for near field measurement that as good as possible compensate for the baffle step loss and added bass because of room once the loudspeaker are placed in a room.
OK, thank you for your graph that explain clearly what happens by measuring at different distances.
Now, suppose that ideally each speaker is not monted on a surface but in a separate box where the side coincides exactly with the diameter of each speaker, basically there are no sides it is the speaker itself that has sides.
In a theoretical situation such as this (but hardly practically realized) I suppose that on range 100-700Hz the response of all lines should be similar becouse there is no energy loss on surface, becouse there is not surface.
It is true ?
Now, suppose that ideally each speaker is not monted on a surface but in a separate box where the side coincides exactly with the diameter of each speaker, basically there are no sides it is the speaker itself that has sides.
In a theoretical situation such as this (but hardly practically realized) I suppose that on range 100-700Hz the response of all lines should be similar becouse there is no energy loss on surface, becouse there is not surface.
It is true ?