So I'm trying to figure out what Woofer to use for my upcoming active 3-way project, and I've been looking in to various Woofers in the 8"-12" range that will work in sealed enclosure and still deliver adequate bass ( about -3dB @ 40hz, tests done in WinISD).
One that seems to get good reviews and is affordable is the Scan-Speak 10" Discovery (26W/8534G00).
However the Xmax is only 6mm and in a box that's at Qtc 0.707 I'll reach the limit at 103dB (at 40hz and without EQ to boost low end, or subsonic).
Looking on the other hand at mechanical limit of 12mm of that woofer the SPL is 108dB at the same 40hz (again with no addition or subtraction of the signal).
So two questions:
What's a resonable headroom that I would have to subtract from these numbers to get nominal listening volume, 5dB, 10dB, 15dB?
How far above the Xmax of 6mm is safe to go on 'party mode' where some distorsion isn't a big deal for the purpose of listening pleasure?
As say -15dB from 103dB for nominal listening will be just 88dB.
And yes, DSP will be used so I can/will use a HP filter (subsonic) as my music taste doesn't use much below 40hz(-ish).
And yes, I have looked in to other drivers as well, but I've only found a handfull that might fit all requirements, and it's few reviews/test to go by.
One that seems to get good reviews and is affordable is the Scan-Speak 10" Discovery (26W/8534G00).
However the Xmax is only 6mm and in a box that's at Qtc 0.707 I'll reach the limit at 103dB (at 40hz and without EQ to boost low end, or subsonic).
Looking on the other hand at mechanical limit of 12mm of that woofer the SPL is 108dB at the same 40hz (again with no addition or subtraction of the signal).
So two questions:
What's a resonable headroom that I would have to subtract from these numbers to get nominal listening volume, 5dB, 10dB, 15dB?
How far above the Xmax of 6mm is safe to go on 'party mode' where some distorsion isn't a big deal for the purpose of listening pleasure?
As say -15dB from 103dB for nominal listening will be just 88dB.
And yes, DSP will be used so I can/will use a HP filter (subsonic) as my music taste doesn't use much below 40hz(-ish).
And yes, I have looked in to other drivers as well, but I've only found a handfull that might fit all requirements, and it's few reviews/test to go by.
Depending on the music you listen to, 20 dB of peak to average ratio is not uncommon. Some music more, some less. I'd plan for 20 dB of headroom though.
If you can roll off at 80hz to a subwoofer, that will greatly increase the amount of clean excursion headroom for the woofers in your main speakers, and the maximum spl capability before either damage or distortion.
If you can roll off at 80hz to a subwoofer, that will greatly increase the amount of clean excursion headroom for the woofers in your main speakers, and the maximum spl capability before either damage or distortion.
20dB?!? 😱
Christ I might have to rethink the whole project.
A sub would be nice, but it's not even remotely in the budget at the moment. 🙂
Christ I might have to rethink the whole project.
A sub would be nice, but it's not even remotely in the budget at the moment. 🙂
Hi,
20db is typical margin for 'high' dynamic range material ( classical, jazz,...). This is not headroom though: it is the 'crest factor' within the recorded material ( it's dynamic range over rms ( average level)).
Headroom is some spare allowed over this dynamic range.
20db is 'worst case' scenario, if you listen to pop it is more 14 or 12 db dynamic range. So if you target pop style listening 20db may be enough.
That said here is a view on it ( i share it) about the situation ( worth a read):
https://www.diyaudio.com/forums/mul...m-spl-low-distortion-2-a-647.html#post6537952
20db is typical margin for 'high' dynamic range material ( classical, jazz,...). This is not headroom though: it is the 'crest factor' within the recorded material ( it's dynamic range over rms ( average level)).
Headroom is some spare allowed over this dynamic range.
20db is 'worst case' scenario, if you listen to pop it is more 14 or 12 db dynamic range. So if you target pop style listening 20db may be enough.
That said here is a view on it ( i share it) about the situation ( worth a read):
https://www.diyaudio.com/forums/mul...m-spl-low-distortion-2-a-647.html#post6537952
Agree that 20dB is often quoted as the headroom needed above the average spl for most music I.e. is average listening is 80dB at the listening position then 100dB peak is worth working towards.
Couple of point to watch out for though, most Sensitivity figures for woofers are stated at 1w at 1m (or 2.83v if 8 ohm nominal).
Please correct me if If am wrong - you will lose 6db for every meter you move away, you will gain 3dB for using two speakers, every 3dB you need to double amp power and every 10dB is a perceived halving of volume. Also note that if you listen at 85dB(A weighted average) for over 8 hours you risk permanent hearing damage and you half the safe listening time for every 3dB increase. As an example most UK music venues restrict their volume at the stage front to 92dB(A) which gives 2.5hrs of time before the daily dose is reached, saves their backsides from litigation.
Anyway, for me 75dB at the listening position is pretty loud which equates to 81dB at the 1m data measured point giving minus the 3dB for two speakers, therefore a max spl at Xmax of 98dB would do me fine and us in the realms of possibility for most designs, 105-110dB at 1m for a single speaker is not uncommon, for active studio monitors 115dB is possible.
Peak SPL Calculator
This is a useful calculator to check.
Not sure if this helps but a few odd comments.
Couple of point to watch out for though, most Sensitivity figures for woofers are stated at 1w at 1m (or 2.83v if 8 ohm nominal).
Please correct me if If am wrong - you will lose 6db for every meter you move away, you will gain 3dB for using two speakers, every 3dB you need to double amp power and every 10dB is a perceived halving of volume. Also note that if you listen at 85dB(A weighted average) for over 8 hours you risk permanent hearing damage and you half the safe listening time for every 3dB increase. As an example most UK music venues restrict their volume at the stage front to 92dB(A) which gives 2.5hrs of time before the daily dose is reached, saves their backsides from litigation.
Anyway, for me 75dB at the listening position is pretty loud which equates to 81dB at the 1m data measured point giving minus the 3dB for two speakers, therefore a max spl at Xmax of 98dB would do me fine and us in the realms of possibility for most designs, 105-110dB at 1m for a single speaker is not uncommon, for active studio monitors 115dB is possible.
Peak SPL Calculator
This is a useful calculator to check.
Not sure if this helps but a few odd comments.
No, it is 6db each time you double the distance: at one meter 100db, at 2meter 94db, 4 meter 88db,.....
Yes.
There is more than that: 85db is ok for 8hours each days. You have to take into account resting time between exposition. I doubt about the 92db it is usually more 102db spl in France ( integrated over 15mn period, it is somewhat harmonized over European Country's so my doubt) 92db is low ( a drummer already break it even playing softly).
Take a look at link given it is all explained by E.Geddes.
You are mixing things: the 75dbspl is for one loudspeaker only, and it is a reference level.
Xmax is in no way the limit for headroom: you are well at the limit of distortion hence zero headroom left.
From what you describe i suppose your listening point is at 2 meters. So you loose more or less 6bd: 75+6+3 = 84db ( rms spl). So for 20db headroom you need at least 104dbspl CLEAN ( so well under xmax).
The calcul made by E. Geddes and Salas in the link i provided is the way to calcul it ( all pro does it that way for decades...).
Some great answers here!
I guess I should make a measurement of how loud I usually listen, to get an idea of what's right for me. 8hrs a day is more than common though.
But I also reflect on that it can't be that many speakers that reaches these levels as there's a lot of them with smaller woofers like a 2-way with 6,5", and even larger 3-way kits have woofers that would be bellow that.
I guess I should make a measurement of how loud I usually listen, to get an idea of what's right for me. 8hrs a day is more than common though.
But I also reflect on that it can't be that many speakers that reaches these levels as there's a lot of them with smaller woofers like a 2-way with 6,5", and even larger 3-way kits have woofers that would be bellow that.
Yes take measurement of your own habits, it make much more sense than to rely on pro habits ( however keep in mind that if it is done that way there is good reasons for it.).
In the link provided there is interesting exchange between Graaf and Dr.Geddes about smaller drivers.
I'm biased ( and agree with E.Geddes/Salas) but yes, expecting small drivers to reproduce things without distortion is utopia. That said distortion is not always something we dislike and in some case it makes things 'nicer'.
In the link provided there is interesting exchange between Graaf and Dr.Geddes about smaller drivers.
I'm biased ( and agree with E.Geddes/Salas) but yes, expecting small drivers to reproduce things without distortion is utopia. That said distortion is not always something we dislike and in some case it makes things 'nicer'.
If you build your speakers to do home cinema THX standards then I doubt you'd be disappointed with the output.
From memory 105dB max for each speaker at listening position.
Cheers,
Rob.
From memory 105dB max for each speaker at listening position.
Cheers,
Rob.
Does anyone have suggestions for a woofer in the say 100Euro range that would do well?
Beyma 12BR70 looks better on paper (sim) for instance, but it seems like there's been two versions and conflicting data, so without the tools to do T/S measurements it seems risky.
Also the Peerless SLS 12" (P830669) looks good on paper (sim), but again I'm not finding much info.
Beyma 12BR70 looks better on paper (sim) for instance, but it seems like there's been two versions and conflicting data, so without the tools to do T/S measurements it seems risky.
Also the Peerless SLS 12" (P830669) looks good on paper (sim), but again I'm not finding much info.
Rob this is primarily for music, but as I use my PC as a mediacenter I'm also using the speakers when I watch a film, so I do keep that in mind, it's just not main priority.
Keep in mind that you have to take spectral distribution into account as well. There is usually less demand for SPL at 40 Hz than at 60 to 80 Hz. Depending on your listening habits the Scan-Speak could be sufficient. In the 70ies we did mostly listen to closed-box speakers (at least here in centrral Europe) whose woofers had less x-max and still it was sufficient for most.
Regards
Charles
Regards
Charles
The room right now doesn't allow the speakers to be very far from the wall, I might rearrange the room to listen lengthwise, sort of speak. But I'm guessing that they will still be close(-ish) to the wall, and I've tried to account for that when trying simulations.
And though it's not something one would like to mention on this forum, we usually only know what's in front of us, we usually don't have a set of other speakers at the side to switch between.
I guess I'm afraid to end up with obnoxious "party speakers", I'd rather have 'too little' than ending up with something that's 'too much'. And with a limited budget I can't expect both SPL and SQ in spades.
And though it's not something one would like to mention on this forum, we usually only know what's in front of us, we usually don't have a set of other speakers at the side to switch between.
I guess I'm afraid to end up with obnoxious "party speakers", I'd rather have 'too little' than ending up with something that's 'too much'. And with a limited budget I can't expect both SPL and SQ in spades.
The Inverse Square Law: every doubling of distance = -6db, when the expansion of the sound wave is unconstricted, so for eg., outdoors. In a real room, a certain percentage of the sound waves are contained by the room boundaries and the real SPL loss will be less than the theoretical. Construction materials and room absorption coefficient will have an effect too.
Doubling of the same sound source = +6dB, when the 2 sources are fed the same voltage and when the frequencies of the 2 sources are located within 1/4 wavelength of each other (or is it 1 wave length, that one I'm a little fuzzy on). There seems to be a fair amount of confusion on this topic that pops up now and then. But we see the truth of this every time 2 drivers are wired in parallel and every time we use a xo program to sum 2 drivers in the xo region when the responses at the xo frequency are down 6dB and the summed response is flat.
Here is a link that states it clearly and provides the math to make the calculation: Total dB level adding of coherent correlated sound sources combining decibels or SPL sound pressure level audio logarithmic decibel scale loudspeaker add signal noise levels incoherent noncoherentt - sengpielaudio Sengpiel Berlin. In this site's vernacular, 2 speakers are a "coherent" sound source and therefore sum to +6dB. "Incoherent" sound sources on the other hand, only sum to +3dB.
I think Ugg10 was referring to two speakers as in stereo, therefore incoherent and therefore 3dB gain.
Well, if the 2 stereo speakers are playing the same thing, not always the case of course, but if they are then that would be 2 coherent sources and a 6db increase.
Yes a fair amount of confusion: you are talking about acoustic coupling within a loudspeaker or a stack of loudspeakers, ( so about double the emmissive surface (+3db) and lowering of impedance (+3db) for a given voltage) but ... we are not talking about that in there!
Stal is right, +3db for incoherent sound source and we talk about radiated power from a loudspeaker so 10log ( = +3db the source is considered as omni so you radiate into a sphere).
If we were talking about voltage yes doubling is +6db (20log) or acoustic coupling, but i repeat this is not the case here (if it was we would have induced frequency of interest into the discussion).
Last edited:
Did you see the way Dutch & Dutch handled low frequencies near the wall?
1) Passive cardioid 8 inch
2) Two 8 inch placed near the wall to boost their effective power through boundary reinforcement.
You don't need to match their speaker because you only need to build for your room and tastes.
That's some really cool speakers, great idea!
I'm not sure if I would be able to pull it off though, might be a little bit over my abilities.
Lots of interesting things being said here as well, I thank you for your help.
I'm not sure if I would be able to pull it off though, might be a little bit over my abilities.
Lots of interesting things being said here as well, I thank you for your help.
The cardioid portion is pretty easy to make because you aren't trying to sell the speaker to the public. No need for the perfect to become the enemy of the good.
However, I don't like the idea of the boundary reinforcement below 100Hz built into the two D&D main channels. That's a waste of drivers because you're in the room's modal region. But what's valuable about boundary reinforcement is that you can separate the low frequency drivers from your main channels and place them in the room corners. Not just lower cost but better performance as well.
And in practice you'd do it above 100Hz, up to your room's transition frequency, which saves even more money in your mains. If you try to specifically draw sub-transition frequencies out of your main channels you kill yourself in both performance and cost. In turn, that helps you with your cardioid low-mid performance because they don't have to work all the way down to 100Hz.
However, I don't like the idea of the boundary reinforcement below 100Hz built into the two D&D main channels. That's a waste of drivers because you're in the room's modal region. But what's valuable about boundary reinforcement is that you can separate the low frequency drivers from your main channels and place them in the room corners. Not just lower cost but better performance as well.
And in practice you'd do it above 100Hz, up to your room's transition frequency, which saves even more money in your mains. If you try to specifically draw sub-transition frequencies out of your main channels you kill yourself in both performance and cost. In turn, that helps you with your cardioid low-mid performance because they don't have to work all the way down to 100Hz.
