Okay. I have a measuring mic, but no anechoic chamber
I don't have much experience in measuring loudspeakers, but I could possibly make some in-bookshelf measurements of the speakers when they have been built. But how would you go about doing this? Simply measure each driver on its own and then design the xo based on these measurements? If so, where should the mic be placed when measuring? And won't I get too many room reflections resulting in dips and peaks on the FR making it difficult to design a xo based on these measurements?Thanks for all the help!
Looking at your simulations it seems that a 300-350Hz xo to a 2dB higher midrange could work? And the tweeter sensitivity I guess should be similar to or slightly higher than the woofer?
Yes, I think 300Hz is perfectly reasonable in theory but it can depend on the exact shape of the driver responses. Mid about 2-3dB higher and tweeter the same or a touch more. You loose more high frequencies as you go further off-axis so you may want a little more to play around with there. Or maybe just a little boost with the treble knob will do the trick for that.
You're going to need someone more knowledgeable than me to advise you in terms of measurement. I haven't gone down that road yet. Once you know what to do and get properly set up, it's not really that difficult though. You take short, gated measurements to nullify any room effects for example. But the learning curve can be a little steep to start off with.
This was why I mentioned using a receiver with a built-in EQ function a ways back because of your unique situation and, if you don't have the time or patience for attacking that curve, these receivers do the measurements and corrections automatically. Takes 5 minutes. But don't I know it, it all comes down to $ in the end.......
You're going to need someone more knowledgeable than me to advise you in terms of measurement. I haven't gone down that road yet. Once you know what to do and get properly set up, it's not really that difficult though. You take short, gated measurements to nullify any room effects for example. But the learning curve can be a little steep to start off with.
This was why I mentioned using a receiver with a built-in EQ function a ways back because of your unique situation and, if you don't have the time or patience for attacking that curve, these receivers do the measurements and corrections automatically. Takes 5 minutes. But don't I know it, it all comes down to $ in the end.......
A couple other inexpensive but good SQ 10" woofers to look at are :
Dayton Audio DA270-8 10" Aluminum Cone Woofer | 295-334
Peerless 830668 10" Paper Cone SLS Subwoofer | 264-1110
Both of them however want similarly large vented boxes - too large for you. But either might work sealed, perhaps a little better if you could give a little boost to about 50Hz and below.
This one also looks possible:
Dayton Audio RSS265HF-8 10" Reference HF Subwoofer 8 Ohm | 295-442
It's suggested best for a sealed box but the response actually looks quite good in 50L vented. The FR isn't quite so smooth as others but the lower sensitivity (closer to 85dB in the LF I think) makes SB 8ohm mids much better matches, which keeps the speaker impedance acceptably high. I'd want to sim it before making a decision though.
Dayton Audio DA270-8 10" Aluminum Cone Woofer | 295-334
Peerless 830668 10" Paper Cone SLS Subwoofer | 264-1110
Both of them however want similarly large vented boxes - too large for you. But either might work sealed, perhaps a little better if you could give a little boost to about 50Hz and below.
This one also looks possible:
Dayton Audio RSS265HF-8 10" Reference HF Subwoofer 8 Ohm | 295-442
It's suggested best for a sealed box but the response actually looks quite good in 50L vented. The FR isn't quite so smooth as others but the lower sensitivity (closer to 85dB in the LF I think) makes SB 8ohm mids much better matches, which keeps the speaker impedance acceptably high. I'd want to sim it before making a decision though.
Nice with some driver suggestions. I have also been looking at the Peerless SLS-10. It seems like a good candidate if I go sealed. The Daytons are either hard to find or too expensive here in Europe, but I’ll definitely consider the Peerless.
I have attached a table showing three potential 3-ways. What do you think of these options? The prices are all from Falcon Acoustics in the UK.
I guess the all SB speaker appeals the most to me since I think this will give us the biggest ‘wow-effect’. Also, there is actually a proven design using (almost) these drivers, although it is made with the 10” woofers as subs crossed much lower, and also the XO is not available, but will have to be purchased with the drivers and XO components from Germany
By the way, what do you think about the SB being ported in a 55L box? The port would have to be quite big and long, and this will take away volume.
Attachments
Last edited:
EQ does appeal to me, especially if I end up with a sealed enclosure where a bit of low freq boost would come in handy. But EQ doesn’t solve everything. I will still need to design the XO, but I guess I could somewhat simplify the XO and let the EQ handle the rest?
I guess that something like a Behringer DEQ2496 could work well here?
Hi,
This is getting very tedious. A committee can't design a speaker
to save its life and its being reduced to just nonsense about
choosing drivers with no real idea about what will work well.
rgds, sreten.
If you want pragmatism, its already in this post, my last
suggestion, which will work with the clients amplifier.
You won't do better, spending a lot more, unless you
really know what you are doing. FWIW I specialise in
building great cheap systems for people, KISS.
I often build systems way below the persons expected
budget if they think they have to spend a lot, and of
course can't do miracles (but pretty near) on a very
restricted budget, if you can create what they need.
This is getting very tedious. A committee can't design a speaker
to save its life and its being reduced to just nonsense about
choosing drivers with no real idea about what will work well.
rgds, sreten.
If you want pragmatism, its already in this post, my last
suggestion, which will work with the clients amplifier.
You won't do better, spending a lot more, unless you
really know what you are doing. FWIW I specialise in
building great cheap systems for people, KISS.
I often build systems way below the persons expected
budget if they think they have to spend a lot, and of
course can't do miracles (but pretty near) on a very
restricted budget, if you can create what they need.
Last edited:
Nobody asked you to join in on this “tedious” thread, sreten. I’m learning something here, and I think that especially jReave and andy19191 are being very helpful.
Let me have another look at your last post in this thread. Your “pragmatic” advise is actually not so clear to me. You mentioned Paul Carmody’s Tarkus, which is way too big for this setup and is also not designed for near-wall placement. You then say you have sympathy with system7’s suggestion, which was an 8” two-way that I’m not sure will be powerful enough for the room. And lastly you mention the ZD3C again, now modified for the second time. A few people in this thread have already pointed out various problems with using this design for this setup, such as high XO with off-axis listening as well as lobing issues.
First of all, the ZD3C, which I’m guessing you’re favouring among your other suggestions, is actually more expensive than some of the 3-ways discussed recently in this thread. Secondly, the ZD3C will need to be changed so much that to me it seems like it’s actually a new design. And if you were to design something from scratch for this setup (assuming you really knew what you were doing and adopted the KISS principle), would you then start with 2 x 7” woofers crossed at 1kHz to a 2” mid? Or would a 3-way with a 10” crossed to a 5” mid seem more reasonable?
Instead of this degrading attitude, why don’t you enlighten us with your great skills and experience and tell us how we could improve our 3-way design?
Looking at woofers, I like the SB 29NRX as well as this one: Peerless 835026 HDS-205 AL Woofer Speaker
The Dayton RSS265 I mentioned previously doesn't sim well above 100Hz. Forget about it. The Peerless 10" SLS sealed doesn't give you enough in the bottom end even with room gain (with bass boost might be a different story, I'm not sure), and the SS 26W8634 doesn't really work well ported in 50L - it wants a bigger box than that. The HDS Peerless linked above gives almost the same performance as the 10" SB but won't play quite as loudly - it will be xmax limited to about 100dB at 1m which quite frankly should be plenty loud enough. It also is a little lower in sensitivity which is a good thing.
My preference is for this SB tweeter: SB Acoustics SB29RDC-C000-4 Tweeter. Lower distortion than the others and able to cross below 2000Hz which is probably a good idea here so that the off-axis response in the HF of the mid remains as good as possible. But some of the other SB's could probably work too.
Your buddy's amp will actually handle a 4ohm nominal speaker which I wasn't sure of before so choosing a 4ohm mid can work out if the xo doesn't drop the impedance down too much below 3ohm. That can depend on the specific drivers but I don't think it hurts to keep the impedance on the higher side. So to match with the SB 10" woofer, I'd go with the highest sensitivity SB woofer, SB Acoustics SB17NRXC35-4 Midwoofer. If you want to match the Peerless 8" woofer, I'd go with the highest sensitivity 5" SB driver, SB Acoustics SB15NRXC30-4 Midwoofer.
If you want to keep the impedance higher, the 5" Audax is a good choice: Audax HM130Z10 Aerogel 13cm 5.25" woofer. In simulations, I'm finding so far that it's a little easier to work with than the SB mids, which means less components are necessary.
The 10" SB looks good in 50L net tuned to about 24Hz. You may want to find a compromise between max SPL and max port velocity for this one. Port resonance could also be a factor. Try a port diameter of 7.75cm and see if that works for you. The port may need to have a 90* angle in it and will take up about 1.6L or so if you go with 7.75cm. You'll have to work out if that is going to work with your size constraints. Mid and tweeter chamber should be about 3 to 5L probably. It's also possible to use 2 smaller diameter ports and get the same response. That may take up a little less volume.
From my point of view, there is no wow factor and no need in choosing drivers all from the same company. The wow factor is in the sound and different company's drivers can mix and match easily. And all of the above drivers I've listed are high quality, good value drivers.
My approach at this point would be to see which combination of drivers is going to work financially for you, physically space-wise (ie. volumes of drivers, ports, bracing etc) and then xo-wise. I can help you with this if you need it. Then I'd build the best version and see how it sounds, fine-tuning by ear if possible. I think there's a good chance this might actually turn out to be just fine. But if the simulations don't quite make it work in the real world, then at that point I'd be looking at measuring and/or some form of eq.
That Behringer certainly looks like it will do the job but looks like a little overkill for this kind of non-critical listening setup.
The Dayton RSS265 I mentioned previously doesn't sim well above 100Hz. Forget about it. The Peerless 10" SLS sealed doesn't give you enough in the bottom end even with room gain (with bass boost might be a different story, I'm not sure), and the SS 26W8634 doesn't really work well ported in 50L - it wants a bigger box than that. The HDS Peerless linked above gives almost the same performance as the 10" SB but won't play quite as loudly - it will be xmax limited to about 100dB at 1m which quite frankly should be plenty loud enough. It also is a little lower in sensitivity which is a good thing.
My preference is for this SB tweeter: SB Acoustics SB29RDC-C000-4 Tweeter. Lower distortion than the others and able to cross below 2000Hz which is probably a good idea here so that the off-axis response in the HF of the mid remains as good as possible. But some of the other SB's could probably work too.
Your buddy's amp will actually handle a 4ohm nominal speaker which I wasn't sure of before so choosing a 4ohm mid can work out if the xo doesn't drop the impedance down too much below 3ohm. That can depend on the specific drivers but I don't think it hurts to keep the impedance on the higher side. So to match with the SB 10" woofer, I'd go with the highest sensitivity SB woofer, SB Acoustics SB17NRXC35-4 Midwoofer. If you want to match the Peerless 8" woofer, I'd go with the highest sensitivity 5" SB driver, SB Acoustics SB15NRXC30-4 Midwoofer.
If you want to keep the impedance higher, the 5" Audax is a good choice: Audax HM130Z10 Aerogel 13cm 5.25" woofer. In simulations, I'm finding so far that it's a little easier to work with than the SB mids, which means less components are necessary.
The 10" SB looks good in 50L net tuned to about 24Hz. You may want to find a compromise between max SPL and max port velocity for this one. Port resonance could also be a factor. Try a port diameter of 7.75cm and see if that works for you. The port may need to have a 90* angle in it and will take up about 1.6L or so if you go with 7.75cm. You'll have to work out if that is going to work with your size constraints. Mid and tweeter chamber should be about 3 to 5L probably. It's also possible to use 2 smaller diameter ports and get the same response. That may take up a little less volume.
From my point of view, there is no wow factor and no need in choosing drivers all from the same company. The wow factor is in the sound and different company's drivers can mix and match easily. And all of the above drivers I've listed are high quality, good value drivers.
My approach at this point would be to see which combination of drivers is going to work financially for you, physically space-wise (ie. volumes of drivers, ports, bracing etc) and then xo-wise. I can help you with this if you need it. Then I'd build the best version and see how it sounds, fine-tuning by ear if possible. I think there's a good chance this might actually turn out to be just fine. But if the simulations don't quite make it work in the real world, then at that point I'd be looking at measuring and/or some form of eq.
That Behringer certainly looks like it will do the job but looks like a little overkill for this kind of non-critical listening setup.
I know this tweeter from Jeff B’s Piccolos and it’s very nice, but also quite a lot more expensive than the SB26STCN, which also should be capable of a relatively low XO, say just above 2kHz? (Fs=960Hz)
But why do I need the 1dB extra of the 6.5”? In a previous post you said that ideally the mid should be 2-3dB higher than the woofer, but the 6.5” midwoofer is 4dB higher than the 10” SB woofer. Wouldn’t the 5” at +3dB be sufficient? I’m afraid that a 6.5” will need more space, which will take even more volume out of the already limited cabinet space.
Thanks for all the help. It seems that I could boil this down to two options, a less expensive 8” 3-way (Peerless 835026 + SB15NRXC30-4 + SB26STCN(?)), which would be £280 for the drivers, and then a more pricey 10” all-SB (SB29NRX75-6 + SB17NRXC35-4 + SB29RDNC), which would cost £370. I will have to discuss this with my friend.
By the way, how much should I except to spend on XO parts for either of these 3-ways? I’m looking for cheap / good-value caps and coils.
Yup, if budget is a concern, the SB 26STCN should work fine. If Jeff is happy with it then that more than works for me.
Concerning the SB 10" woofer: SPL=88dB at 80HZ. Baffle diffraction and room boundary effects (using the power response without the rear wall reinforcement - 1st graph below) adds another 1.5dB below 100Hz. Typically, when you cross a woofer in a 3-way the xo can also have a tendency to add a little bit of peaking below the xo point. So the 2nd graph is what the SB29NRX and SB17NRXC-4 look like with a xo in place. Now also notice that although the SB17 is called 92dB, it's really closer to 90/91dB in the middle of its usable FR for our application (see the 3rd graph), so this is the level we'll be adjusting to in the xo.
So this combination can work together if the simulations are correct. But as I've been mentioning, your situation is a bit unique and I'm sort of best guessing with the diffraction and boundary modelling. So usually what I like to do in a situation where I'm modelling a speaker for someone else, is to choose drivers that allow a measure of flexibility so you can make adjustments up or down depending on what actually happens in your particular room and placement and for personal preferences.
So for example, if you set this up and find that there isn't enough bass (so if the room reinforcement is less than we think), you can pad down the mid and the tweeter so that the levels match to your liking. But if you think that there is too much bass (so we have underestimated the amount of boundary and room reinforcement), we are a little out of luck - in this combination I have no added resistance on the mid so this is as loud as we can get it without resorting to eq. It's a general no no to ever pad down a woofer btw.
And just to compare, 4th graph is the SB15NRCX-4 rated at 91dB. You can see that its FR is very similar to its bigger brother and might just work with the 10" SB but it's going to be the same kind of situation but about 1dB even lower.
You are correct the 17 needs more room than the 15. With heavy fill though, the 17NRXC will work fine in 5L. The 15NRXC can work in 3L or so. Basically, you just have to sit down and do the math to see if everything will work together within your size constraints.
I'm thinking your shelving cabinet will probably be 20mm material. If you put a speaker box inside that, all walls except the front baffle could be made from thinner stock, say 13mm maybe even 10mm. Even better if you add a thin layer (3 or 4mm) of softer material between them and the cabinet, like cork or rubber or similar. The front baffle should be solid, at least 20mm, better at 25.4mm. But if you have 60L internal to play with, the mid and tweeter chamber may take 5 to 7L. Port, bracing, xo and woofer driver volume may take another 3 or 4L. That's 49 to 52L left for the woofer. And that works fine for both the woofers I've mentioned.
I would guess xo total cost could be $100-200, but it totally depends on the kind of parts you go with. Big inductors and capacitors in particular can get expensive depending on what kind they are (and you will need some large values for some of these SB drivers). Iron cores tend to cost less and it looks like Falcon caries a good supply. I know Europe Audio does too. There are raging debates about capacitor types but at least for all the larger values, you can probably get away with non-polarized electrolitics in your situation, although I would prefer to see at least 1 step up in cap quality with these drivers.
Also, a smart idea in your kind of situation is to buy some extra resistors and capacitors so that you can do any possible fine tuning required. This will add a tiny bit more to the cost. Inductors may be fine tuned by unwinding them but a proper meter helps in this respect.
Concerning the SB 10" woofer: SPL=88dB at 80HZ. Baffle diffraction and room boundary effects (using the power response without the rear wall reinforcement - 1st graph below) adds another 1.5dB below 100Hz. Typically, when you cross a woofer in a 3-way the xo can also have a tendency to add a little bit of peaking below the xo point. So the 2nd graph is what the SB29NRX and SB17NRXC-4 look like with a xo in place. Now also notice that although the SB17 is called 92dB, it's really closer to 90/91dB in the middle of its usable FR for our application (see the 3rd graph), so this is the level we'll be adjusting to in the xo.
So this combination can work together if the simulations are correct. But as I've been mentioning, your situation is a bit unique and I'm sort of best guessing with the diffraction and boundary modelling. So usually what I like to do in a situation where I'm modelling a speaker for someone else, is to choose drivers that allow a measure of flexibility so you can make adjustments up or down depending on what actually happens in your particular room and placement and for personal preferences.
So for example, if you set this up and find that there isn't enough bass (so if the room reinforcement is less than we think), you can pad down the mid and the tweeter so that the levels match to your liking. But if you think that there is too much bass (so we have underestimated the amount of boundary and room reinforcement), we are a little out of luck - in this combination I have no added resistance on the mid so this is as loud as we can get it without resorting to eq. It's a general no no to ever pad down a woofer btw.
And just to compare, 4th graph is the SB15NRCX-4 rated at 91dB. You can see that its FR is very similar to its bigger brother and might just work with the 10" SB but it's going to be the same kind of situation but about 1dB even lower.
You are correct the 17 needs more room than the 15. With heavy fill though, the 17NRXC will work fine in 5L. The 15NRXC can work in 3L or so. Basically, you just have to sit down and do the math to see if everything will work together within your size constraints.
I'm thinking your shelving cabinet will probably be 20mm material. If you put a speaker box inside that, all walls except the front baffle could be made from thinner stock, say 13mm maybe even 10mm. Even better if you add a thin layer (3 or 4mm) of softer material between them and the cabinet, like cork or rubber or similar. The front baffle should be solid, at least 20mm, better at 25.4mm. But if you have 60L internal to play with, the mid and tweeter chamber may take 5 to 7L. Port, bracing, xo and woofer driver volume may take another 3 or 4L. That's 49 to 52L left for the woofer. And that works fine for both the woofers I've mentioned.
I would guess xo total cost could be $100-200, but it totally depends on the kind of parts you go with. Big inductors and capacitors in particular can get expensive depending on what kind they are (and you will need some large values for some of these SB drivers). Iron cores tend to cost less and it looks like Falcon caries a good supply. I know Europe Audio does too. There are raging debates about capacitor types but at least for all the larger values, you can probably get away with non-polarized electrolitics in your situation, although I would prefer to see at least 1 step up in cap quality with these drivers.
Also, a smart idea in your kind of situation is to buy some extra resistors and capacitors so that you can do any possible fine tuning required. This will add a tiny bit more to the cost. Inductors may be fine tuned by unwinding them but a proper meter helps in this respect.
Attachments
Last edited:
I don't see how you can choose a midrange or a woofer until you know how you are going to handle the bass boost the location creates. This requires you to estimate the frequency and magnitude of the bass boost and if it is feasible to work a lower sensitivity woofer into the crossover. Without this information you don't know what you would like the woofer and midrange to do in terms of sensitivity and frequency response.
Hi,
To all intents and purposes the the approximation is simply a speaker
with no BSC and a tapered bass roll, rather than maximally flat.
Room modes have hardly been discussed.
The manipulation of the speakers off axis response to suit purpose
isn't getting much truck here, compared to standard 3 way design,
which is then very difficult to manipulate much, usually no problem.
rgds, sreten.
To all intents and purposes the the approximation is simply a speaker
with no BSC and a tapered bass roll, rather than maximally flat.
Room modes have hardly been discussed.
The manipulation of the speakers off axis response to suit purpose
isn't getting much truck here, compared to standard 3 way design,
which is then very difficult to manipulate much, usually no problem.
rgds, sreten.
andy, I have presented what I think is a reasonable estimation of the location's bass boost (which sreten has very succinctly summarized), although I have also noted that I don't think it is by any means perfect.
This is why I would prefer a driver combination that gives the mid both positive and negative wiggle room. That would be the criteria by which I am making selections.
Perhaps the best thing to do in this situation is buy the woofers, set them up and then measure them and then source the mids accordingly. But it strikes me that the OP doesn't really want to go down that road right now. Although I suppose that just using an SPL meter and some test tones could also give a reasonable measure of the in room bass output and that is something to possibly consider I guess.
On the other hand, if there is too much bass (or other peaking) from any simulation that is used, I've also been saying that eq may need to be the option in the end. Too little bass can be corrected with the xo though.
This is why I would prefer a driver combination that gives the mid both positive and negative wiggle room. That would be the criteria by which I am making selections.
Perhaps the best thing to do in this situation is buy the woofers, set them up and then measure them and then source the mids accordingly. But it strikes me that the OP doesn't really want to go down that road right now. Although I suppose that just using an SPL meter and some test tones could also give a reasonable measure of the in room bass output and that is something to possibly consider I guess.
On the other hand, if there is too much bass (or other peaking) from any simulation that is used, I've also been saying that eq may need to be the option in the end. Too little bass can be corrected with the xo though.
The woofers are placed on the floor, against a back wall and near a side wall. The omni-directional, long wavelength, low frequencies are therefore radiating into effectively 1/8th space but your later plots show low frequency radiation into something more like 1/2 space.
What do you suggest if it is unknown whether the woofer can be a lot less sensitive than the midrange or needs to be the same sensitivity as the midrange? A high sensitivity midrange or a low sensitivity midrange? Xmax for the midrange? Perhaps a design that takes advantage of the bass boost is impractical but a design with equal sensitivities will result in a relatively insensitive speaker that requires substantial equalisation to get a flat response.
On a different topic, the imaging will be better if the speakers are symmetric with respect to the side walls. This will come at the cost of a pair of irregularly spaced bookcase sections rather than just one but it may be judged worthwhile.
Thanks for the input guys, it’s very appreciated!
As I mentioned earlier in the thread I have measurement equipment and have some experience with REW. So far I have only made outdoor ground plane measurements, but I’ll have a look at making gated measurements as this would probably be the way to go here. However, aren’t gated measurements only relevant at higher frequencies, like above a few hundred Hz? If so, how should I go about measuring the woofers “in-bookshelf” response?
Despite the increased shipping cost I could possibly buy the woofers first and then measure them before deciding on the midrange, tweeters and XO components. But Is this really the best way to go? Are there other options?
Also, if I decide to measure the woofers first, it wouldn’t be so easy since I would have to build the cabinets, install the woofers and port(s), and then put something inside the cabinet (books?) to compensate for the volume that eventually will be taken up by the midrange enclosure.
As I mentioned earlier in the thread I have measurement equipment and have some experience with REW. So far I have only made outdoor ground plane measurements, but I’ll have a look at making gated measurements as this would probably be the way to go here. However, aren’t gated measurements only relevant at higher frequencies, like above a few hundred Hz? If so, how should I go about measuring the woofers “in-bookshelf” response?
Despite the increased shipping cost I could possibly buy the woofers first and then measure them before deciding on the midrange, tweeters and XO components. But Is this really the best way to go? Are there other options?
Also, if I decide to measure the woofers first, it wouldn’t be so easy since I would have to build the cabinets, install the woofers and port(s), and then put something inside the cabinet (books?) to compensate for the volume that eventually will be taken up by the midrange enclosure.
Two weeks ago at post #10 you had roughed out a set of drivers and were ready to take Step 1 in designing the speakers by simulating the bass enhancement due to the non-standard location, simulating a cabinet and roughing out a crossover using the published driver curves. Depending on what you found you might change a driver or two before firming up on the design. Step 2 would be to buy the drivers and check the design using measured curves for the drivers in the cabinet. This might tweak the odd value in the crossover. You might check the crossover using DSP on you PC if you had suitable hardware. Step 3 would be to buy the crossover components and finish the speakers. Without the information in Step 1 how do you know which woofer to buy and test?
Ok, let me try this 1 last time.
For the purposes of modelling in a typical situation, we are targeting a flat anechoic response. This means that we are usually not including the boundary reinforcement effects despite the fact that they are affecting the real in-room response. This is a fundamental concept that is at the basis of what I am doing here: the real, in-room response is not going to be the same thing as the simulated response (this is true primarily below 100Hz).
So let's look at a typical speaker in a typical room. Let's say the woofer is 24" above the floor, 36" from the side wall and 30" from the back wall. The first graph below shows what these boundary reinforcement effects using the axial response look like and yet they are typically not included in the simulation of the speaker. That's about a 15dB in-room increase at 40Hz, 13.5dB at 50Hz and 11.5dB at 60Hz, and so on using the axial response over the flat modelled response (the SPL values are halved if we look at the power response instead - 2nd graph), and this is the correct way of doing things. What we do though typically to reconcile this disparity between simulated and real is to voice the speaker in the end, maybe reducing the BSC by 2 or 3dB, maybe 4dB but that's usually about it. We're not going anywhere near 10 or more dB in compensation.
But this is for a typical speaker. A big woofer close to the floor probably should have its floor gain effects included into the simulation. But the question becomes - do we use the axial or the power response or some portion thereof?
Let's look at a well regarded DIY example - the tarkus. Ten inch woofer, 18" off the ground. Here's what the builder has to say:
"The Peerless SLS 10" has a healthy sensitivity around 89 dB (which equates to a net 85-86 dB after baffle step losses and floor-bounce gain are factored in)".
So that's 89dB - 6dB (BSL) = 83dB + 3dB (floor gain) = 86dB sensitivity. So for the purposes of simulation it would appear that the power response of 3dB gain is the right one to use here (3rd graph below - power response of floor gain only for a driver 18" above the floor).
So now we can take this information and apply it to the OP's situation noting in particular that the standing or sitting off-axis listening positions further support the use of the power response graphs. Let's start with baffle step:
13' wide baffle x 30" high = 6dB loss to baffle step
This is probably too much since the program assumes that there is space on the sides of the baffle that the waves have to wrap around when they in reality don't have to. But the baffle is only 30" tall, it doesn't go to the ceiling so some portion of the waves do wrap around the top.
So add in the side wall gain = +3dB
10" woofer 12" above the floor, add in the floor gain = +3dB
And I'm choosing to stop there since the side walls are 29" away and the rear wall is 17" away and this is not too unusual, albeit not ideal, for a typical room placement which, to say it again, would not be normally included in the sims. That gets me a net increase of about 1.5dB when you end up plugging all the numbers into the program as I've posted previously. And if you look at the 4th graph, modelling the all SB driver combination with this diffraction and room gain simulation gets me about a 90dB speaker, which is a significant increase of about 8dB over what would normally be modelled with 6dB of BSC (88dB - 6dB = 82dB) with that SB 10" woofer. (Btw, the sim is for a listening position .25m above the tweeter axis.) But perhaps having another couple of dB to for the mid (and the tweeter) wouldn't be that bad of an idea.
So a logical course of action, based on simulated precedent. The fact is there is no perfect program (that I'm aware of) that is going work here so I think the thing to do is to make a reasonable estimation, design a xo that works properly for the selected drivers, allow for driver level flexibility (+/- dB) and maybe have a backup plan (eq) and then proceed from there.
For the purposes of modelling in a typical situation, we are targeting a flat anechoic response. This means that we are usually not including the boundary reinforcement effects despite the fact that they are affecting the real in-room response. This is a fundamental concept that is at the basis of what I am doing here: the real, in-room response is not going to be the same thing as the simulated response (this is true primarily below 100Hz).
So let's look at a typical speaker in a typical room. Let's say the woofer is 24" above the floor, 36" from the side wall and 30" from the back wall. The first graph below shows what these boundary reinforcement effects using the axial response look like and yet they are typically not included in the simulation of the speaker. That's about a 15dB in-room increase at 40Hz, 13.5dB at 50Hz and 11.5dB at 60Hz, and so on using the axial response over the flat modelled response (the SPL values are halved if we look at the power response instead - 2nd graph), and this is the correct way of doing things. What we do though typically to reconcile this disparity between simulated and real is to voice the speaker in the end, maybe reducing the BSC by 2 or 3dB, maybe 4dB but that's usually about it. We're not going anywhere near 10 or more dB in compensation.
But this is for a typical speaker. A big woofer close to the floor probably should have its floor gain effects included into the simulation. But the question becomes - do we use the axial or the power response or some portion thereof?
Let's look at a well regarded DIY example - the tarkus. Ten inch woofer, 18" off the ground. Here's what the builder has to say:
"The Peerless SLS 10" has a healthy sensitivity around 89 dB (which equates to a net 85-86 dB after baffle step losses and floor-bounce gain are factored in)".
So that's 89dB - 6dB (BSL) = 83dB + 3dB (floor gain) = 86dB sensitivity. So for the purposes of simulation it would appear that the power response of 3dB gain is the right one to use here (3rd graph below - power response of floor gain only for a driver 18" above the floor).
So now we can take this information and apply it to the OP's situation noting in particular that the standing or sitting off-axis listening positions further support the use of the power response graphs. Let's start with baffle step:
13' wide baffle x 30" high = 6dB loss to baffle step
This is probably too much since the program assumes that there is space on the sides of the baffle that the waves have to wrap around when they in reality don't have to. But the baffle is only 30" tall, it doesn't go to the ceiling so some portion of the waves do wrap around the top.
So add in the side wall gain = +3dB
10" woofer 12" above the floor, add in the floor gain = +3dB
And I'm choosing to stop there since the side walls are 29" away and the rear wall is 17" away and this is not too unusual, albeit not ideal, for a typical room placement which, to say it again, would not be normally included in the sims. That gets me a net increase of about 1.5dB when you end up plugging all the numbers into the program as I've posted previously. And if you look at the 4th graph, modelling the all SB driver combination with this diffraction and room gain simulation gets me about a 90dB speaker, which is a significant increase of about 8dB over what would normally be modelled with 6dB of BSC (88dB - 6dB = 82dB) with that SB 10" woofer. (Btw, the sim is for a listening position .25m above the tweeter axis.) But perhaps having another couple of dB to for the mid (and the tweeter) wouldn't be that bad of an idea.
So a logical course of action, based on simulated precedent. The fact is there is no perfect program (that I'm aware of) that is going work here so I think the thing to do is to make a reasonable estimation, design a xo that works properly for the selected drivers, allow for driver level flexibility (+/- dB) and maybe have a backup plan (eq) and then proceed from there.
Attachments
I would suggest that this is not particularly sensible given the speakers are in a single fixed location but it does explain why your later plots asymptote towards the "wrong" value at low frequencies. I am not sure I fully followed your explanation of the assumptions you are making but I am not familiar with the input parameters of the simulation program you are using.
Programs that solve the acoustic wave equation in 3D will be able to simulate how sound is radiated from the cabinet and reflected by the boundaries you wish to include. These are common in engineering but rarely used by DIY speaker folk.
Active XO?
Most of the last posts have discussed the different opinions on the woofers' behaviour in the unusual speaker placement. Although we can simulate and get an idea of what low frequency response can be expected, the reality might differ from these simulations.
Here is a thought. I have been thinking about the possibility of going active. I’m aware that this would increase the cost substantially, but I have discussed it with my friend and he is open to the idea. If doing this I thought about using 2 x miniDSP PWR-ICE 125 plate amps and make a 2-way active xo between woofer and mid+tweeter. Here are a few reasons why I think this would be a good idea.
Most of the last posts have discussed the different opinions on the woofers' behaviour in the unusual speaker placement. Although we can simulate and get an idea of what low frequency response can be expected, the reality might differ from these simulations.
Here is a thought. I have been thinking about the possibility of going active. I’m aware that this would increase the cost substantially, but I have discussed it with my friend and he is open to the idea. If doing this I thought about using 2 x miniDSP PWR-ICE 125 plate amps and make a 2-way active xo between woofer and mid+tweeter. Here are a few reasons why I think this would be a good idea.
- It would give me much more flexibility with regards to XO, filtering and EQ. This would come in handy considering the unusual speaker placement.
- No padding down is needed to align the driver sensitivities, so there will be no “wasting power”.
- My friend would like to be able to move his speakers to another location in the future without having to re-design a new passive xo. With a built-in DSP it would be simple to set up the speakers in another room, do in-room measurements, and change the filters accordingly.
- We would save money not having buy expensive parts in the passive xo.
- Although my friend’s current amp will work fine for e.g. the 3-way SB speakers (which have been discussed as a potential build), the speakers have a higher power handling and would be able to play louder with less distortion with a higher powered amp.
- My friend only has one source, which is a Sonos Connect with digital output and volume control. This means that there is not really a need for an integrated stereo amp with input selector. Also, one of his friends may be interested in purchasing his old amp and speakers.
Last edited:
Hi,
JR is correct in that for certain known aspects of speakers,
such as floor (reflection) dip, attempts at correction generally
sound worse than than ignoring the well known phenomena.
You can overegg predictive simulation. The important
bits to get right right are the ones that really matter.
Which here are IMO a wide and even dispersion with
the lobing at the x/o points optimised to the unusual
typical listening axis being used. Of course minimal
lobing is good, manipulating minimal even better.
Careful layout of the drivers can help the broad central.
e.g. for a mid/treble moving the tweeters out from
vertical will tilt up the lobe towards the centre, and
wont really change the general power response.
rgds, sreten.
JR is correct in that for certain known aspects of speakers,
such as floor (reflection) dip, attempts at correction generally
sound worse than than ignoring the well known phenomena.
You can overegg predictive simulation. The important
bits to get right right are the ones that really matter.
Which here are IMO a wide and even dispersion with
the lobing at the x/o points optimised to the unusual
typical listening axis being used. Of course minimal
lobing is good, manipulating minimal even better.
Careful layout of the drivers can help the broad central.
e.g. for a mid/treble moving the tweeters out from
vertical will tilt up the lobe towards the centre, and
wont really change the general power response.
rgds, sreten.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.