In this post Earl Geddes identified three major frequency ranges and explained his reasoning:
1) Do you agree with Geddes and follow his advice in your designs?
2) Are there additional ranges in the high frequencies that make them less critical to listening?
I place an upper limit on high frequencies at 16,000Hz because I can't hear frequencies above that. But are there additional ranges where performance is less critical? For example, get 1,000Hz to 5,000Hz right but don't worry much about 5,000Hz to 16,000Hz because the wavelengths are so short you can no longer localize them or identify colorations from room reflections. (To be clear, I made that last sentence up as an example not as a fact. I have no idea where high frequencies become less critical other than the physical limit of my hearing.)
1) Do you agree with Geddes and follow his advice in your designs?
2) Are there additional ranges in the high frequencies that make them less critical to listening?
I place an upper limit on high frequencies at 16,000Hz because I can't hear frequencies above that. But are there additional ranges where performance is less critical? For example, get 1,000Hz to 5,000Hz right but don't worry much about 5,000Hz to 16,000Hz because the wavelengths are so short you can no longer localize them or identify colorations from room reflections. (To be clear, I made that last sentence up as an example not as a fact. I have no idea where high frequencies become less critical other than the physical limit of my hearing.)
Last edited:
Forgot to add there is a hard lower limit of 80Hz. Even in a large room or outside in open spaces humans can't localize frequencies below 80Hz. So that would be another limit where you could safely ignore efforts at directivity.
My guess would be a bit lower than 1k; more like 500. A good spot to cross to a different driver. From there I'd go close to 10K before I'd cross again, if at all considering "I cant hear frequencies above that". However there may be something to ultrasonics being present (as defined by frequencies I can no longer perceive) that do effect what I can hear - even though the mechanism is obscure. I'd be open to that possibility - versus chopping off that end of the spectrum based on I cant hear it anyway.
So if I were to break a system into chunks, I'd do 20 - 500, 500 - my upper hearing limit and (optional) my upper hearing limit - 20"+" kHz.
So if I were to break a system into chunks, I'd do 20 - 500, 500 - my upper hearing limit and (optional) my upper hearing limit - 20"+" kHz.
Last edited:
It's an interesting question. The 80Hz figure is often used for how far down the mains need to go, but 120-150Hz f3 would make certain designs a lot more possible.
I want to be clear I'm not trying to argue. If the speaker is a device used to reproduce a recording and *is not* a musical instrument, wouldn't the influence of ultrasonics you can hear be captured in the recording? In other words, the studio recording takes care of that problem for you.
The reason I'd like that to be true is that I can ignore reproduction of ultrasonics at the loudspeaker. It makes the speaker easier to build.
billshurv,
I think a relevant post by John k... might add to design considerations in that range.
I've always been confused about what to do in the ~150Hz region if I am designing speakers according to Earl Geddes' major frequency ranges. I didn't know if you were supposed to go straight to multi-subs or stick with the mains. The paper cited by John k... indicates you should go cardioid through the transition. Unless I misunderstand.
I think a relevant post by John k... might add to design considerations in that range.
I've always been confused about what to do in the ~150Hz region if I am designing speakers according to Earl Geddes' major frequency ranges. I didn't know if you were supposed to go straight to multi-subs or stick with the mains. The paper cited by John k... indicates you should go cardioid through the transition. Unless I misunderstand.
Never bothered to learn anything about HF recording due to most [male] folks I was exposed to during my active education/DIY audio 'career' were happy with a ~ flat 12.5 kHz, but it's my understanding from various postings that there's a hard cut-off for CD around 22 kHz, which I know is the limit the pioneers set for HF horn cut-off, though I assume RTR tapes and vinyl may be higher.
That said, I miss 'feeling' all of the top end 'air' around instruments some recordings had even if it was in reality only vinyl 'hiss' [no clue if true]: There's life above 20 kilohertz! A survey of musical instrument spectra to 102.4 kHz
GM
That said, I miss 'feeling' all of the top end 'air' around instruments some recordings had even if it was in reality only vinyl 'hiss' [no clue if true]: There's life above 20 kilohertz! A survey of musical instrument spectra to 102.4 kHz
GM
With regards to ultrasonics, it's rare to find a mic that'll transduce them accurately. Even if the recording engineer does use such a mic, as soon as the sample rate is brought down to sensible, the UHF is lost.
Interesting article, though. In theory, struck instruments (cymbals, drums, etc) should have a harmonic series that extends virtually to infinity.
Chris
Interesting article, though. In theory, struck instruments (cymbals, drums, etc) should have a harmonic series that extends virtually to infinity.
Chris
Earl says mains should be run full range and not crossed to subs. The more the merrier I suppose?
Yup, something like that.
The thread starter has indicated regions that may be less critical. I'd rather focus on the requirements of fixing a band, because ignore any band at your peril. Even though the amounts vary, and the reasons as well, any significant error can ruin the experience.
The thread starter has indicated regions that may be less critical. I'd rather focus on the requirements of fixing a band, because ignore any band at your peril. Even though the amounts vary, and the reasons as well, any significant error can ruin the experience.
My room is small, the sparse modal range starts around 300Hz. Which is why I am intrigued how hard and fast the 80Hz rule is. I could ponder some interesting implementations if I could get away with the mains rolling off earlier and letting the Sub pick up the rest.
Earl Geddes says to let your mains run into the modal region. However, Floyd Toole and Todd Welti crossed low frequencies out of Toole's Revel Salon 2s because it interfered with Welti's low frequency Sound Field Management.
That means you can get away with mains that don't reach into the modal region.
Earl Geddes says you should excite as many modes as possible, which is why he says to allow the mains to run down into the modal region. That's not the same as saying you should design your mains so they play into the modal region. He also says low frequency sources should not be spatially correlated. Mains are spatially correlated. Finally, he says the marginal benefit of adding more low frequency sources loses value quickly after you've added three multi-subs. Although, Welti uses four subs with his sound field management.
So, if you have a small room that runs into the modal region at 300Hz and you think it's to your advantage to design mains at 300Hz and above, do it. Then add three -- or more -- separate and spatially uncorrelated sources for 300Hz and below.
You are safe to ignore the 80Hz rule.
That means you can get away with mains that don't reach into the modal region.
Earl Geddes says you should excite as many modes as possible, which is why he says to allow the mains to run down into the modal region. That's not the same as saying you should design your mains so they play into the modal region. He also says low frequency sources should not be spatially correlated. Mains are spatially correlated. Finally, he says the marginal benefit of adding more low frequency sources loses value quickly after you've added three multi-subs. Although, Welti uses four subs with his sound field management.
So, if you have a small room that runs into the modal region at 300Hz and you think it's to your advantage to design mains at 300Hz and above, do it. Then add three -- or more -- separate and spatially uncorrelated sources for 300Hz and below.
You are safe to ignore the 80Hz rule.
uh... if you have a small room then one sub with a dual coil is enough. You can play with subs placment but most of the time you still will hear the room here, not the sub(s).
For myself around 100 hz to 150 hz is often a good limit. My main has an acoustical cut-off at 90 hz. Upper limit is 2600 if I remember.
I can play with one bass unit without too much pain in small space without indeed waste spatial behavior. Now let fire the two bass unit and switch one tweeter only... you may experiment a subjective lack of bass on the loudspeaker where it occurs. Miracle ... you have localised the bass lack because not hearing the harmonics I assume !
A very good old loudspeakers that plays well for soundstage is the Kef 104/2 ref for illustration : XO : 125 hz & 3000 hz... bass has a 55 hz -f3 !
imho that's the presurizatio nthat is important related to the room size... not only the -f3 but indeed how it is presurised... hence cardioid, multisubs, big Sd or not endless talkings. To much of it kills the musical experience. Makes me think of the head of PS Audio that said one thing iirc about this and sub numbers in a room ! I understand it as too much Sd too low and you waste the rest cause you will hear the room too much.
I readed that most of the energy in the recordings and where the music was in acoustical events and venues are in the 40 to 800 hz, then spl level drops dramaticaly bellow and above that... even with pipe organ low note you will not see as much in most of the reccordings ?!
I would like to know indeed till what limit the first class area - so above 1 k hz- is important ? ... most of acoustic instruments live below 1 k and exhibit harmonics to 8/9 k hz max often ! Voices are exceptions having not really harmonics if I understood from 100/200 for most voices to .... analogic phones were cutings below 8 k hz iirc? Do I need an acurate phase above 8/10 k hz ?
For tweeters after 15k even the harder wall with white paint are absorbing the waves, so... I trust what Weltersys is saying about the 12.5 k hz flat !
For myself around 100 hz to 150 hz is often a good limit. My main has an acoustical cut-off at 90 hz. Upper limit is 2600 if I remember.
I can play with one bass unit without too much pain in small space without indeed waste spatial behavior. Now let fire the two bass unit and switch one tweeter only... you may experiment a subjective lack of bass on the loudspeaker where it occurs. Miracle ... you have localised the bass lack because not hearing the harmonics I assume !
A very good old loudspeakers that plays well for soundstage is the Kef 104/2 ref for illustration : XO : 125 hz & 3000 hz... bass has a 55 hz -f3 !
imho that's the presurizatio nthat is important related to the room size... not only the -f3 but indeed how it is presurised... hence cardioid, multisubs, big Sd or not endless talkings. To much of it kills the musical experience. Makes me think of the head of PS Audio that said one thing iirc about this and sub numbers in a room ! I understand it as too much Sd too low and you waste the rest cause you will hear the room too much.
I readed that most of the energy in the recordings and where the music was in acoustical events and venues are in the 40 to 800 hz, then spl level drops dramaticaly bellow and above that... even with pipe organ low note you will not see as much in most of the reccordings ?!
I would like to know indeed till what limit the first class area - so above 1 k hz- is important ? ... most of acoustic instruments live below 1 k and exhibit harmonics to 8/9 k hz max often ! Voices are exceptions having not really harmonics if I understood from 100/200 for most voices to .... analogic phones were cutings below 8 k hz iirc? Do I need an acurate phase above 8/10 k hz ?
For tweeters after 15k even the harder wall with white paint are absorbing the waves, so... I trust what Weltersys is saying about the 12.5 k hz flat !
Last edited:
Momentary room pressurisation with box speakers below the room's fundamental frequency is a "thing" apparently and can be felt. My U-frame subs don't do it, it's not the same as exciting room modes, that still happens with a dipole. I have an IB sub in the wall in the corner and that gives a very smooth response in the room, just exciting my room's one big mode at 40Hz, easily cut.
Just to be very clear, my interest is in making sure that I can't localise the sub. I suspect this is listener and room dependent.
You should be able to tell by listening 😉 Seriously, experiment to see when you can locate it, and report back. 🙂
If you room is small and you use just a coil in serie for the low pass of the bass driver, imho you can climb up to 150 hz cut off without problem. Steeper slope may be ask a different lower XO perhaps... but here it means you certainly active with EQ... so...
in small rooms, monitors with -f3 around 100 hz to 80 hz with a standalone sub in between or below the sofa table in front of the speakers is working very well imho. Slim towers à la Proac D15/d18, etc are too much boomy in small rooms (say below 18/20 square meters) : the 6" units often seen in these design with a BR alignment made to expand the bass are.... boomy ! You hear the room in the bass and unfornatully the chest punch and snap overlapping the room modes or not according the room is not coralated. Idem with two 8" in a small rooms with 40 hz -f3 ; it's often too much ! That's maybe not by chance we see very good two ways monitors working wonderfully in small space😱.... Of course I don't know for the D&D 8C I never heard, my bet is it's working because the dsp should slide away in the low the bass frequencys in a half cardioid way ! My mainn loudspeaker radiates below 90 hz (first slope low pass electric around 120 hz) towards the front wall so at the back at the loudspeaker and it's very ok !
Below Schroeder you listen to your room, above 15 k hz whatever the age of the ears you listen to your walls !😀 Don't bother with the 80 hz rule the room already won and rules the size choice of your the drivers 😱 .... that's what I'm saying to myself to avoid to change speakers every year !*
I a hesitating with slim tower with low SD but low bass till 35 hz (that's what I have already in a circa 40 square meter living room) and a PA made driver but with higher -f3 and -f10 because the increased Sd! Typically:research of the subjective snap behavior and instant presurization... a good bass is light, never boomy à la Qtc =1 or higher if we talk of better reproduction... rooms always winn !
Last edited:
I disagree with this/agree with Geddes. My room is fairly small, my mains run full range (12" woofer but -3 db at 54Hz). I started with two subs, when later they were on sale I bought two more. By measurement I had nice smoothing in the modal region up to 3 subs, the fourth didn't do much as he suggests.
Traditionally it has been thought that localization doesn't happen below 80 Hz. My experience is that it is higher (perhaps throughout the modal region).
So yes, the more the merrier 🙂
Bill
Hi Bill,
What is the Sd of each sub please ? what are the -f3 and -f10 ?
What's the alignment of your two mains with 12" ? Bessel or fast stiff roll off ?
What is a small room in Texas 😉 ?
And have you a sub or more below you firering towards the main ?
You listen mainlyamplified rock (albums & concerts) or also classical events ?
Look at what say for Instance Patrick Bateman here about the big suma (15") without sub in a small space...
What is the Sd of each sub please ? what are the -f3 and -f10 ?
What's the alignment of your two mains with 12" ? Bessel or fast stiff roll off ?
What is a small room in Texas 😉 ?
And have you a sub or more below you firering towards the main ?
You listen mainlyamplified rock (albums & concerts) or also classical events ?
Look at what say for Instance Patrick Bateman here about the big suma (15") without sub in a small space...
Last edited:
I posted a sequence of three in the "multi-subs" thread here with links to speakers/subs and a description of the room.
https://www.diyaudio.com/forums/subwoofers/134568-multiple-subs-geddes-approach-227.html#post6108694
Neither the speakers nor subs go incredibly deep, nice "foundation" to the music, though.
Old jazz and classical, occasional guilty pleasures (mostly older R&B).
We previously had a "Texas-sized" houses but now like our more reasonably sized house in the woods 🙂.
Older picture of the room. Looks nicer now. I used to have spaces that I could control, now it is an ongoing negotiation........
I have some more plans to further smooth the response/optimize the acoustics.
https://www.diyaudio.com/forums/subwoofers/134568-multiple-subs-geddes-approach-227.html#post6108694
Neither the speakers nor subs go incredibly deep, nice "foundation" to the music, though.
Old jazz and classical, occasional guilty pleasures (mostly older R&B).
We previously had a "Texas-sized" houses but now like our more reasonably sized house in the woods 🙂.
Older picture of the room. Looks nicer now. I used to have spaces that I could control, now it is an ongoing negotiation........
I have some more plans to further smooth the response/optimize the acoustics.
