Yes I get that, I try to draw it out of you and reverse engineer how I feel about it lol...So far it seems that the Reverb at nearfield levels are insignificant, in yours and others representation....I will have to create a A/B between my small dome tweeter 2 ways and the Axi+350hz horn and see what it sounds like. I don't know if thats a fair comparison or not.
Last edited:
My understanding is that linear DI is equal to constant directivity, is it not?
Hi Rob,Is it me or should the high and low pass be reversed?? Never used the program. I would expect a 20Hz high pass and 80Hz Low pass to do that. Just looking at a dummy load vs. sim filter response.
Rob
We are using the woofer under it's in-box impedance peak, meaning is it is acting like a Linkwitz Transform. Basically the low pass at 21 Hz is compensating
for the natural roll off of the driver below the peak.
Welcome! There are ways of measuring/estimating excursion that i've never mastered. (Art Welter is good at stuff like this).
Here's one place i do rely on Hornresp, after the build. After all Hornresp filters are in place, raise the voltage until xmax is reached at some frequency in the Diaphragm Displacement Window. Whatever Acoustical Power curve goes with that maximum voltage, represents maximum linear SPL...(as i see it...)
If there is no musical content in the frequency range where xmax is hit or close, then there is of course more headroom available at non challenged frequencies. But, if using limiters to protect against xmax (good practice!), the limiters will take away that non challenged headroom.
As said before, if listening at low volume, all this is big whoop.
But that's what doesn't make sense to me about you design intentions Camplo.....why build such ginormous SPL capability just for low volume...and for close listening at that? Don't mean to be critical here, just saying design seems a bit schizoid.
Yeah, you have plenty of headroom for a small to mid size room, but probably at lower SPL than realized once you move to a more normal listening distance (again, lower than realized if shooting for flat 20Hz response).
Hey, i'm not all that convinced THD matters so much, but i'm thrilled how clean my PPSL sounds so i'm gonna post some measurements after all...just as food for thought for where you decide to take f-3.
Here's a raw measurement outdoors, green is raw, blue is with 3 small PEQs, and a linear phase low pass at 120Hz.
Here's the distortion at almost 126! dB (+6 added to fundamental taken at 2m)
Squeaky clean loud in my book.
I have to believe your 18H+ are capable of very similar....if you ugh, lets go of 20 Hz
Hi wesayso, you mean with the accompanying large boost...correct? If boosting like that, I'd think an attenuating Linkwitz transform would be a much better fit than a low pass, so as not to increasingly/disproportionately attenuate higher frequency response, that is still desired.
But really, i'm out of my element here...i say just get some bigger drivers (or more drivers) for deep down low, and do it right.
Indeed with the boost within the Linkwitz transform idea, that's the gain boost Camplo is using.
If he has the headroom, this "should" work, 24 dB is a lot though.
The lowpass is offsetting the extra gain at higher frequencies. Basically it is a sort of EQ function to use it like this.
Offsetting the natural roll off of the driver(s).
If he uses the subs primarily under the impedance peak, before handing off to the next driver group it "should" work, right?
If he has the headroom, this "should" work, 24 dB is a lot though.
The lowpass is offsetting the extra gain at higher frequencies. Basically it is a sort of EQ function to use it like this.
Offsetting the natural roll off of the driver(s).
If he uses the subs primarily under the impedance peak, before handing off to the next driver group it "should" work, right?
Directivity index is just one number value.
Often given per frequency.
A constant directivity means that the DI is linear over a certain freq range.
So they are related but not the same thing.
This assumes a perfectly linear curve for Kms, Bl, and such with excursion, of course. Which is almost never the case in reality. I would take the displacement calculations from linear models with large pinches of salt for anything other than very low levels of voltage drive, personally.
Also it's interesting to see that simulations would be stopped once a physical speaker is made. I consider the qualification and refining of a model based on real measurements (or, vice versa) a big part of validating a design - and being confident in the process for any future work. I can see why that may be boring once you've got something to listen to, though!
It also seems there might be some confusion about what the 'near field' is. It's a clearly defined transition to the radiated pressure falling off with a 1/r, regardless of angle. In the acoustic near field, this cannot be considered to be true. The transition is also frequency-dependent.
The direct-to-reverberant ratio is affected by the receiver distance, source size and directivity, and the room's acoustical parameters derived from the reverberation time, early decay time, and energy time curve. Likewise for the critical distance, which is the point where the direct sound and reverberant sound fields are of equal level.
Here are some relevant slides from one of the Klippel Live webinars, which can be found here:
https://www.klippel.de/know-how/education/klippel-live-web-seminars/iec-60268-21.htmlParts 1-3 are particularly relevant, with part 2 being about making useful acoustic measurements in ‘too small’ spaces
And going back to the recommendation of books, these topics are well covered with simple analogies in Acoustics & Psychoacoustics, and Toole's Sound Reproduction.
I understand what you are saying, but at first, I thought that it was wrong. DI is a function of both angle and frequency. It is an average of the polar response at any frequency and thus cannot show precisely what the directivity is, only its average value. Hence is not the same as "directivity. That said, "on average" a flat DI is constant directivity in the way that most of us think about it.
yes, what I wanted to point out is that directivity index is just for ONE frequency.
Hence the name "index", it's very similar like refractive index in optics.
Which can also be different per color (read: wavelength = frequency).
See: https://en.wikipedia.org/wiki/Refractive_index
(or it's just my physics brain, but I see many similarities)
Here you can see the usual way to work with an directivity index, figure 18.20 (left side)
https://www.sciencedirect.com/topics/engineering/directivity-index
So in other words each frequency will have it's own directivity index.
Aka, the average polar response. (aka, how a certain amount of energy (or any other value) is distributed around a flat circle)
A constant directivity will give here a straight (linear) line.
Looking at the example from science direct, from 100Hz till 10kHz, the line is pretty constant.
Some deviations around 1kHz, 1.6kHz and 8kHz.
And to correct myself, this actually means that the DI is NOT constant, but it's a rising linear slope (ax+b function)
With a log function if I am not mistaken (because of the log freq scale), but you're probably much faster to answer that question than me
I have done this once in the past, but to get a sense of how constant the directivity is, one only has to take the standard deviation per frequency per DI from this line.
The higher the STDEV is, the less constant the directivity is.
For those who are less familiar with the standard deviation, it represents the spread around a certain average.
Mathematically it's a bit tricky because of the rising line, there are a couple of variables to keep track of.
Also the reason why I haven't done it much, because it's quite some work to do this manually (or in excel).
But in programs like VituixCAD that would be quite easy to do.
An experienced person also doesn't really needs it when the relative off-axis response is being used.
Although just a quick peak at one number is also easier of course.
Last edited:
Oh yes, and the slope of the DI vs freq graph says something about how strong the speaker is beaming as the freq goes up.
Again, can be written in a simple ax+b function.
So a flat line will represent a linear directivity (DI for every freq is the same).
The higher this straight flat line is, the higher the directivity of the speaker.
In other words, a flat line with a very low DI value will be a pure omni directional speaker.
In this sense the term "constant directivity" is kind of confusing.
Because a linear directivity is certainly constant as well.
Again, can be written in a simple ax+b function.
So a flat line will represent a linear directivity (DI for every freq is the same).
The higher this straight flat line is, the higher the directivity of the speaker.
In other words, a flat line with a very low DI value will be a pure omni directional speaker.
In this sense the term "constant directivity" is kind of confusing.
Because a linear directivity is certainly constant as well.
Yes, I totally consider large signal extrapolations of sims as best guesses....
So i just use Hornresp, etc, displacement as a best guess for setting peal limiting.....while not assuming I can maintain linear SPL anywhere close to said drive level.
It's not really about being bored vs listening, that turns me from sims once something is built. Once built, i see the measurements i think could be improved on, and i hear stuff i think could be better, so i build a new round. For instance, i started with sims on my first few MEH prototypes. But after maybe version 3, I've followed with 6-10 subsequent builds all based off measurements and no sims. Just my style, that's all.Absolutely. I've posted links to a few of the Klippel slides you put up a bunch of times, trying to help show the inverse square law / far field kick-in distance.
I think a big part of the confusion with the term "near field", is how the studio guys co-opted it to describe being within critical listening distance.
it's a shame really, the confusion between the near-field / far-field transition line (region), and the critical distance line (region)
I came for the criticism, just well meaning criticism, and you are on the well meaning criticism team lol. Metal sharpens Metal they say.
I have theories, I am able to come up with things that others don't think of so I have to run the course and find my own conclusions since I tend to try things others have not.
Part of my theory for this design is a maximization of direct energy. I want to see it affects my mix/mastering in the long term.
How to increase Direct Energy;
Radiation Size
Directivity
Proximity to source
That has been my main ingredients. So far it seems like things are going according to plan....Listening to the one sub, sitting as close as I intend, The effects of Direct energy seem to be what I theorized. For example; Listening at low level, I can still feel the bass. This is also the reported experience of others who migrated to a large system. They reported not wanting to turn up the volume as loud, and that a full spectrum of sound (linear down to 30hz at least, I would think) seems to require less volume to be satisfied with volume.
I've been using a 65inch tv for a monitor for the last 5 years? Its within arm length. They said this wouldn't work.
I have had my my current system within close proximity for the last several years. The measurements look horrible on the top end, proof that close proximity isn't enough to cure reverb energy.
If reflections were as tame as you imply, simply from close proximity listening....then room treatment would be a mute point if listening so close.
(repeat for impact)
If reflections were as tame as you imply, simply from close proximity listening....then room treatment would be a mute point if listening so close.
Maybe now we can see how I can up with the inverse; make as big of a system I can that I think could be used nearfield.
Last edited:
Room treatment (done right) is not just about reducing the overall single number reverberation time but producing a smoothly decaying energy without distinct, specular echoes.
While there are targets for the single number metric, depending on the room’s intended use, there’s a lot more to a good-sounding room than that.
Also, trying to dominate the reverberant field with more SPL by simply turning the existing system up is a fool’s errand, since that energy is still put into the room.
The integration time for the early-order reflections is a function of the room size, and the absorptive or scattering nature of the surfaces.
In a small room, even highly directional sources, placed at the corners and cross-fired or toed-in heavily will still bounce off the rear and back end of the sidewall surfaces. It may be more controlled, but it’s still there.
What I'm getting at, is an acoustically small basement cuboid room with untreated surfaces is still going to have a certain character, even if you're sitting somewhere with a high direct-to-reverberant ratio. You might push the majority of the reverberant energy outside of the first 30-80 milliseconds, but if it's specular in nature, then it can still be improved with careful placement of absorption and diffusion.
If you have some space to spare, then you really should look into the 'live end, dead end' approach to small studio acoustic design.
It's early in the year, but I'm voting this for best camploism of the thread.
While there are targets for the single number metric, depending on the room’s intended use, there’s a lot more to a good-sounding room than that.
Also, trying to dominate the reverberant field with more SPL by simply turning the existing system up is a fool’s errand, since that energy is still put into the room.
The integration time for the early-order reflections is a function of the room size, and the absorptive or scattering nature of the surfaces.
In a small room, even highly directional sources, placed at the corners and cross-fired or toed-in heavily will still bounce off the rear and back end of the sidewall surfaces. It may be more controlled, but it’s still there.
What I'm getting at, is an acoustically small basement cuboid room with untreated surfaces is still going to have a certain character, even if you're sitting somewhere with a high direct-to-reverberant ratio. You might push the majority of the reverberant energy outside of the first 30-80 milliseconds, but if it's specular in nature, then it can still be improved with careful placement of absorption and diffusion.
If you have some space to spare, then you really should look into the 'live end, dead end' approach to small studio acoustic design.
It's early in the year, but I'm voting this for best camploism of the thread.
This reminds me of the time I asked if any body had every used side by side 15"s and got crickets. I really appreciate the theoretically correct answers though! Have you experimented with such equipment in such a way as I intend, and as best as you could came to your own personal conclusions?
Omg I almost didn't catch the straw man....No one suggested turning up the stereo lol.
Real life experience, call that a Camploism lol
vs
Logically these two statements are opposed. Which one are we going to subscribe too...
At close proximity, reflections are insignificance
or
Dominating the reverb field with more direct energy (not spl kind sir your statement should of said " dominating the reverb field, by increasing direct energy, if you are to have the same discussion everyone else is) is a fools errand.