A finished speaker usually has all the driver sections in parallel and there's no impedance problem.
The point wasn't that they needed to be directly in parallel for the measurement but that they should be played together. This was behind this statement...
Here are some resources that might be helpful. Note, Jeff Bagby's papers refer a lot to his Passive Crossover Design (PCD) excel based model but they are still accurate guides in general. With regards to measuring acoustic offset, I have seen step-by-step instructions for following his procedure in VituixCAD - you can probably find them with Google.
The VituixCAD guide is for Non-USB mics but still addresses some questions, like measuring two woofers. I know there are people on here who disagree on the USB vs Non-USB mics issue, but I can say that moving to a non-USB mic (along with a 2-in/2-out audio interface) and utilizing all of the on- and off-axis (spinorama) data in VituixCAD revolutionized my ability to design a speaker.
The VituixCAD guide is for Non-USB mics but still addresses some questions, like measuring two woofers. I know there are people on here who disagree on the USB vs Non-USB mics issue, but I can say that moving to a non-USB mic (along with a 2-in/2-out audio interface) and utilizing all of the on- and off-axis (spinorama) data in VituixCAD revolutionized my ability to design a speaker.
Hrm,
Coming back to one thing that seems to be an issue.... near field response and correcting it. I read about it in the papers linked above, and I've seen it talked about and addressed in spreadsheets in a few sources. But none of them seem to show the math being used. I'd rather know the math instead of just blindly use a spread sheet. My understanding from reading is that the baffle dimensions and frequency for the near field response need to be corrected as they're excluding baffle diffraction and we need that. Trying to find the math to do this myself.
Anyone?
Very best,
Coming back to one thing that seems to be an issue.... near field response and correcting it. I read about it in the papers linked above, and I've seen it talked about and addressed in spreadsheets in a few sources. But none of them seem to show the math being used. I'd rather know the math instead of just blindly use a spread sheet. My understanding from reading is that the baffle dimensions and frequency for the near field response need to be corrected as they're excluding baffle diffraction and we need that. Trying to find the math to do this myself.
Anyone?
Very best,
I found some info finally.
But I'm a little confused. I'm not sure what the j is in this formula:
Little was defined, so I'm having to fill in gaps. I think this is otherwise straight forward to calculate and to simply do it for each frequency involved in a near field response. The diffraction amplification from half space to full space should be higher as the frequency gets lower, so it would be an easy check. But I just need to figure out what j represents, if it matters. And what the initial f is. f0 is defined, but nothing else is. I'm always floored how incomplete things are in all this.
Or, if anyone has a simplified way to scale half space to full space for correcting near field measurements SPL, that would be great!
Very best,
But I'm a little confused. I'm not sure what the j is in this formula:
Little was defined, so I'm having to fill in gaps. I think this is otherwise straight forward to calculate and to simply do it for each frequency involved in a near field response. The diffraction amplification from half space to full space should be higher as the frequency gets lower, so it would be an easy check. But I just need to figure out what j represents, if it matters. And what the initial f is. f0 is defined, but nothing else is. I'm always floored how incomplete things are in all this.
Or, if anyone has a simplified way to scale half space to full space for correcting near field measurements SPL, that would be great!
Very best,
Last edited:
I found more complete information on this. I'm still confused however with the j & f variables at the end. Any help clarifying this would be appreciated big time. Once I'm past little mental block, I think I can write my own script to then scale near field response measurements from my FRD that I make.
Very best,
Very best,
Anyone have a way to simplify what the values here are for j & f in this formula:
You'd think this is simple, but I'm having a total mental block on being able to use the variables to scale full space and half space here with frequency and averaged dimensions of a baffle. You'd think this would be trivial. I reached out to the author and they literally told me "they forgot."
Very best,
You'd think this is simple, but I'm having a total mental block on being able to use the variables to scale full space and half space here with frequency and averaged dimensions of a baffle. You'd think this would be trivial. I reached out to the author and they literally told me "they forgot."
Very best,
Thanks,
So in that case, it can be simplified out then?
I still don't know what the f is, while f0 is defined. I'm just confused how incomplete it all seems. Or maybe I'm too dense.
Very best,
What are you trying to do.. plot baffle response per frequency, or determine the relative level of your nearfield measurement.. or are you just wanting to know general before and after baffle step levels?
Have you already worked out how low you can measure and where the baffle step is in relation to this? Vituixcad has a procedure for splicing, although I can't remember specifics.
Have you already worked out how low you can measure and where the baffle step is in relation to this? Vituixcad has a procedure for splicing, although I can't remember specifics.
I'm trying to correct near field measurement's SPL, usually bass frequencies.
The methods used to do this all use a basic calculation and then do it automatically and use a spread sheet to do it, etc. Where you input your near field FRD and correct the SPL and then output a new corrected near field FRD. Then splice to far field as normal.
Purpose is to produce more accurate measurements for building crossovers and for producing loud speaker measurements in general.
I know VituixCad has a lot of things that can be done, but I'm just trying to understand the math being used that is documented above to equalize near field measurements between full space and half space as listed. Instead of using someone else's script or software, I'd just like to know the math and how to apply it to my measurements to equalize it to be more accurate and do the calcs myself. I'm stuck on the above formula and how to apply it properly using my baffle size and the full and half space scaling.
Very best,
There are several reasons you don't want to use the nearfield measurement any higher in frequency than necessary, and you want to take the regular measurement as low as possible. If it turns out you can splice below the baffle step then you have a forward (half-space) power change of -3dB and pressure change of -6dB
Do you know if your cross is going to overlap with the baffle step.. it may help you prioritise the need for all this?
Do you know if your cross is going to overlap with the baffle step.. it may help you prioritise the need for all this?
The baffle is 10" wide (250mm) and 20" height (500mm). Drivers are 8" (200mm) each. Can cross anywhere from 700hz to 1khz comfortably from initial measurements and behavior modeling.
Still, it would be nice to be able to do the calculation to scale it. It's just that ratio above, but it's hard to track down what the variables are. Only one was defined (F0). Everyone's software and spreadsheets address this and yet it's really hard to find a simple way to describe it so that anyone can calculate it. Even just for one frequency, to help understand it.
I have a spreadsheet that does it, but it's formula is locked away and can't see what it's doing under the hood. The author of it said they literally forgot and are not helpful on this front. So I have a tool that does it.
I just really would like to be able to do it myself manually by knowing the math and how it works. Figured it would be child's play for a lot of the folk on this board and perhaps was easier than I thought, and I'm just being dense or something.
Very best,
https://www.audiomatica.com/wp/wp-c...Loudspeakers-at-low-Frequencies-with-CLIO.pdf
So, just went through this and perhaps its not the same thing. But basically it just refers to CLIO, whatever that is, and all the functions are there. The other option being to drill holes in boxes and measure inside the box (microphone in box) which is not an option basically. Pr = k (f/f0)^2 Pb showed promise, but.... k is not defined anywhere... just "new value." f is frequency in hz and f0 is normalized to a reference frequency... whatever that means. It's not defined either. And this requires you drill a hole in a box and measure from inside of it just to get Pb. So, that's not happening.
And everything else just refers back to whatever CLIO is and uses it to process stuff.
So... back to where I was. Either I'm too dense for this, or I'm missing something. I'm not sure. It's weird to me that things are not defined and software reference is made. 🙁
From an Atkin's paper referencing Keele:
Also not explaining or showing any way to properly deal with half space to full space. 🙁
This paper by Keele is a lot better I think. It starts with a glossary of variables defined, as I would expect. I'm going to see if I can make sense of some of this and see if any of it is useful for my purpose here.
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1974-04 AES Published) - Nearfield Paper.pdf
Still... I know a few people around here have to know this like its simple child's play. Someone has to. I guess I'm just at the mercy of time to see who will read any of this thread and get to this point and be able to provide the simple response that solves all this.
Very best,
Last edited:
I think you underestimated the complexity.
https://techtalk.parts-express.com/forum/tech-talk-forum/54417-baffle-diffraction-calculations
These two links are from post #3 from above:
https://trueaudio.com/st_diff1.htm
https://www.t-linespeakers.org/tech/bafflestep/bstepcompo.html
Thanks, I truly appreciate it's complex, more so than I can easily just figure myself. Hence all the questions and the thread. I just know there's a lot of folk who know more than me on this and maybe I'll get lucky with someone willing to take a moment to make it easy with a tell-me-like-I'm-5 explanation or simplification of the math. I'm happy to just get this working on a basic square/rectangle baffle. I'm not looking to get too wild with complex enclosures beyond that in terms of diffraction.
That one article is much, much better with things defined.
Here in the 6db step, we see the same proportion scaler at the bottom. Vi/Vo = 1+entity / 2+entity, where w3 = 115*2pi / Wb and Wb is baffle width (in meters).
Though, I'm getting lost on the square root of 2iw. I'm not sure what i and w are there. This is just like the previous proportion, just different variables... same issue, j and f. This is the same equation.
If someone can easily unravel what that is... this would be pretty easy to do otherwise. I'm just stuck on what variables are. The formula is here in my face, lol.
These are describing the same thing. I'm just missing what some of these variables are. This is an easy proportion/equalizer scaler if I just knew what the variables were.
Very best,
So coming back to this, the author of one of these above that I contacted explained the f = frequency, f0 = the entity/d as explained, j is an imaginary part of a complex number.
The only thing I'm stuck on at this point then is: j and do we calculate something with j or do we just simplify it out (in which case, why is it even presented here)? This may be my obvious lack of understanding of how to use j here.
Our goal is to correct our baffle from a rectangle or square to a sphere.
So if f0 = 42.70/d where d = diameter of this spherical box (full space), but I guess I need clarification of how to establish this value.
And if f0 = 34.16/d where d = width of your baffle (half space) (square shape), or d = w(h/w)^1/3 if using a rectangle shape. This is our real baffle.
f is the frequency, should be straight forward.
j is some imaginary part of a complex number, I'm not sure what to do with it, other than perhaps reduce it out for simplicity unless someone can instruct me otherwise? Clarification; author told me that j cannot be simplified out, it's somehow included. So I'm trying to work out how to do that.
Very best,
The only thing I'm stuck on at this point then is: j and do we calculate something with j or do we just simplify it out (in which case, why is it even presented here)? This may be my obvious lack of understanding of how to use j here.
Our goal is to correct our baffle from a rectangle or square to a sphere.
So if f0 = 42.70/d where d = diameter of this spherical box (full space), but I guess I need clarification of how to establish this value.
And if f0 = 34.16/d where d = width of your baffle (half space) (square shape), or d = w(h/w)^1/3 if using a rectangle shape. This is our real baffle.
f is the frequency, should be straight forward.
j is some imaginary part of a complex number, I'm not sure what to do with it, other than perhaps reduce it out for simplicity unless someone can instruct me otherwise? Clarification; author told me that j cannot be simplified out, it's somehow included. So I'm trying to work out how to do that.
Very best,
I'm too old to remember my math class about complex numbers so give a look at this page:
https://www.mathsisfun.com/numbers/complex-numbers.html
https://www.mathsisfun.com/numbers/complex-numbers.html