• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Directiva r2 project: market requirements gathering

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Did you not see the pictures I posted before? It's exactly the same principle as used by D&D.

View attachment 160498

In bad weather I measure in an old church, using gated measurements. In good weather I'd like to move outside from some groundplane measurements. The gated measurements are certainly not perfect in terms of resolution, although relative levels still come out well.

View attachment 160499

Since then I've got a better mic stand ..

View attachment 160500
May I be pertinent ask what is in those slots?

Also, at 20Hz wavelength is 17 metres, as large as a church will be measurements of such low frequencies is not reliable. There are very few anechoic chambers in the world that can measure such low frequencies.

Obviously ground plane measurement will not show any polar pattern beyond +/- 90 degrees.

If the box is filled with compressed rockwool we can expect the speed of sound to be as fast as 200m/s. That reduces the wavelength of 20Hz to 10 metres. In other words the sound coming out of the slots are 70cm in front. This difference may well give you the cancellation you need but that is at one frequency. How are you going to spread that linearly to the operating range of the speaker?
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Well, they were in plain sight, so from that sense, no secret ;)
I am sorry for a bit thick. Are they just perforated sheet metal to stop the rockwool escaping?

PS. Thank you for persevering to answer my questions.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
May I be pertinent ask what is in those slots?

Also, at 20Hz wavelength is 17 metres, as large as a church will be measurements of such low frequencies is not reliable. There are very few anechoic chambers in the world that can measure such low frequencies.

Obviously ground plane measurement will not show any polar pattern beyond +/- 90 degrees.

If the box is filled with compressed rockwool we can expect the speed of sound to be as fast as 200m/s. That reduces the wavelength of 20Hz to 10 metres. In other words the sound coming out of the slots are 70cm in front. This difference may well give you the cancellation you need but that is at one frequency. How are you going to spread that linearly to the operating range of the speaker?

Oh, simple racing mesh with some cloth behind it, purely aesthetic.

I understand the resolution is not very good at these low frequencies, and unfortunately I can't provide you with any more data until next spring when I can perform ground plane measurements, outside. Since I cross these over at 80Hz anyhow, it wasn't very important for me. On the other hand, all of my iterations showed different patterns, so I still believe, while the resolution is not optimal, that the differences are real since these are normalized measurements.

PS: I'm quite sure I can do a series of ground plane measurements to get full 360° low frequency data :)
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,227
Likes
17,806
Location
Netherlands
I am sorry for a bit thick. Are they just perforated sheet metal to stop the rockwool escaping?
That’s for Tim to answer, but in general no secrets there either. Indeed the purpose would be to leave the material inside the box. The size of the openings in general is supposed to be 1/3 to 2/3rd of the SD of the woofer if I remember correctly. But you’ll need to tune for the specific design. That also makes this not an easy concept. You’ll need quite a lot of experimentation to get it right.
PS. Thank you for persevering to answer my questions.
No worries, not asking is far more troublesome in the long run.
 
  • Like
Reactions: 617

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,227
Likes
17,806
Location
Netherlands
@TimVG: what was the effect of the extra stuffing compared to less? Did it also affect the upper range of the cardioid effect, and if so, in what way? How high does it still work? With your 15” woofer and big horn you’ll probably only need it to work up-to 650~700 Hz?
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
@TimVG: what was the effect of the extra stuffing compared to less? Did it also affect the upper range of the cardioid effect, and if so, in what way? How high does it still work? With your 15” woofer and big horn you’ll probably only need it to work up-to 650~700 Hz?

In this wide cabinet there was little effect with too little stuffing. What's interesting is that the midrange response, while not different in terms of pattern, did clean up in terms of response. I believe there's an optimal relation between slot size and amount of damping material and cabinet size perhaps even - but I can't say for sure until I can work on some more prototypes.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
With your 15” woofer and big horn you’ll probably only need it to work up-to 650~700 Hz?

I cross over at 800Hz. One could argue even a regular sealed enclosure holds its pattern down low enough to use in a regular room, but where's the fun in that :)
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Since I cross these over at 80Hz anyhow, it wasn't very important for me.

PS: I'm quite sure I can do a series of ground plane measurements to get full 360° low frequency data :)
If you are going to use them only for the bottom two octaves why do you want to go all that trouble to reduce rear emission? What will took over above 80Hz?

I’m afraid you can’t simulate more than +/-90 degree polar diagram on a 2PI ground plane measurement. You need the front emission. Without that there will be no cancellation.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
If you are going to use them only for the bottom two octaves why do you want to go all that trouble to reduce rear emission? What will took over above 80Hz?

I’m afraid you can’t simulate more than +/-90 degree polar diagram on a 2PI ground plane measurement. You need the front emission. Without that there will be no cancellation.

Because I'm curious on how these things work - this is my hobby (one of them) and since this is still not very well understood in practice, I'm trying to make sense of it all.

I'm sorry but I've made full 360° measurements before using the ground plane method, I've seen others do it. I've seen these measurements verified.. To be clear I'm not talking about simulating. I'm talking about making 36 measurements in 10° incrementations (or well, 18, for a symmetrical speaker)
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Because I'm curious on how these things work - this is my hobby (one of them) and since this is still not very well understood in practice, I'm trying to make sense of it all.

I'm sorry but I've made full 360° measurements before using the ground plane method, I've seen others do it. I've seen these measurements verified.. To be clear I'm not talking about simulating. I'm talking about making 36 measurements in 10° incrementations (or well, 18, for a symmetrical speaker)

Won’t one of the slits be blocked by the ground?
 

briskly

Active Member
Joined
Jul 10, 2019
Messages
115
Likes
153
"dispersion"
It's the radiation pattern or directivity. Dispersion is the variation of phase velocity with frequency. Dispersion is more an issue for sound waves in solids, whereas phase velocity in free air is more stable. Rising HF phase velocity in solids is responsible for the blaster noises you get when striking a metal coil or wire, among many other sounds.

I’m afraid you can’t simulate more than +/-90 degree polar diagram on a 2PI ground plane measurement. You need the front emission. Without that there will be no cancellation.
I don't think is a measurement with the speaker in a soffit, but one in which the normal speaker base is laid on the ground. The mirror image is very nearly in phase at large wavelengths.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
It's the radiation pattern or directivity. Dispersion is the variation of phase velocity with frequency. Dispersion is more an issue for sound waves in solids, whereas phase velocity in free air is more stable. Rising HF phase velocity in solids is responsible for the blaster noises you get when striking a metal coil or wire, among many other sounds.


I don't think is a measurement with the speaker in a soffit, but one in which the normal speaker base is laid on the ground. The mirror image is very nearly in phase at large wavelengths.
As I apologised in my first post today I had used wrong terminology on my posts yesterday. I should have said radiation pattern. Dispersion is completely wrong.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
so the enclosure would be laid on its back?

For this purpose the enclosure would be on its head, aimed with the center of the driver towards the mic capsule which is on the ground.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
For this purpose the enclosure would be on its head, aimed with the center of the driver towards the mic capsule which is on the ground.
That’s not ground plane. Ground plane means speaker is in 2pi. Nothing from the rear is heard. In your case this is not the case. Speaker is in 4pi.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
That’s not ground plane. Ground plane means speaker is in 2pi. Nothing from the rear is heard. In your case this is not the case. Speaker is in 4pi.

Ground plane is very much what I'm describing. Or at least that is what just about every source available tells me. What you're describing is what I know as half space.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,314
Location
UK
Ground plane is very much what I'm describing. Or at least that is what just about every source available tells me. What you're describing is what I know as half space.
If the speaker has only one sound source you are correct but in your case you have multiple sources. Moreover, in your suggested measurement there is the issue of diffraction; due to the reflected image of the loudspeaker, the baffle appears to be twice larger as it really is, causing a different diffraction response along the edge in contact with the ground. This will skew your measurement.
 
OP
Rick Sykora

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,517
Likes
7,028
Location
Stow, Ohio USA
Using specially made custom wave guide for the tweeter and a narrow baffle it is possible to have controlled directivity above around 400Hz.

It is possible to integrate an extended range "subwoofer" that works up to 400Hz that is made of multiple drive units on that narrow baffle.

That is similar to Revel Performa 3 which I suggest the route we take for r2.

Maybe borrow some aspects, but really not interested in doing an active knockoff.

Want something that has a unique value prop in its own right. r1 had a head start with the Purifi woofer. Unless the bass directivity can be simulated, looks like a non-starter to me. But perhaps we can get good enough in first pass?

If the main benefit is better bass in the room, controlled bass directivity may be worthwhile. There has been no discussion of a dipole bass module ala Linkwitz LX521. From what I have read, some have a major advantage as cabinet resonances are well controlled. See the gainphile speaker results on the Princeton site.
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,281
Location
Oxford, England
I found this on the D&D website, which explains a lot. How I missed it before, God knows. I blame the wine :)



What they are doing is ingenious. They are leaking the rear energy of the midrange driver through two acoustic filters on each sides. As the rear energy is in reverse polarity the output from the ports cancel the emissions from the side, forcing the otherwise omnidirectional pattern of the narrow baffle to become cardioid. This normally cause pronounced FR anomalies due to cancellations and additions, but they corrected them with the DSP.

Very clever indeed.

Now who is there to design a passive acoustic filter? :)

You may find interesting source of information in this topic about the development of the Dutch & Dutch 8c:

https://www.diyaudio.com/forums/multi-way/192737-2-waveguide-cardioid.html
 
Top Bottom