Directiva r2 monitor prototype build

[CDR&D]

This is how this looks in practice. By using cut-to-size acoustic foam sample-to-sample variance should be absolutely minimal. There will be some more acoustic material, but the material covering the slots is by far the most crucial.

Many thanks for sharing Tim - have you already done measurements for this revised prototype?

 

[TimVG]

Many thanks for sharing Tim - have you already done measurements for this revised prototype?

Alas not, I haven't gotten around to making any further progress. Realistically I'm looking at the end of next month for finishing the build and probably April for measurements.
I know this seems long, but I don't have much spare time these days.

 

[TasTinker]

This is how this looks in practice. By using cut-to-size acoustic foam sample-to-sample variance should be absolutely minimal. There will be some more acoustic material, but the material covering the slots is by far the most crucial.

That was very much my experience, type, density and thickness of acoustic material in or over the slots had the biggest affect on the response patter (I ended up with material in the slots)

 

[Wizard509]

Alas not, I haven't gotten around to making any further progress. Realistically I'm looking at the end of next month for finishing the build and probably April for measurements.
I know this seems long, but I don't have much spare time these days.

Any update on this project? Thank you.

 

[TimVG]

I am close to finishing up the cabinets. After that I'll make some initial measurements to verify that the system performs as intended.
When a satisfactory version has been produced I've made arrangements to have it measured on a Klippel NFS, and will subsequently release all files in terms of measurements, and also the cabinet for either CNC milling, and at a later stage, 3D printing. This should allow people to develop a crossover of their choosing, to match a chosen bass cabinet. The only thing that will need to be altered is the baffle step when the top module is placed on top of another cabinet, but this can be simulated quite easily and accurately.

 

[TimVG]

I've now finished the cabinets (minus paint). I will have to check when I have a big room available for initial measurements.

Just out of curiosity: Would you, if given a choice, put more weight into achieving the most even polar response down to the modal range (constant), or towards a gradually sloping curve yet still controlled without obvious steps in directivity.

 

[olieb]

Just out of curiosity: Would you, if given a choice, put more weight into achieving the most even polar response down to the modal range (constant), or towards a gradually sloping curve yet still controlled without obvious steps in directivity.

Good question!
I assume, that a "perfectly" even polar response is not possible and there will always be some increase in DI/ERDI from 300 Hz to 5000Hz.
But if a truly constant directivity would be achievable, the on axis would probably have to be tilted down a little in most rooms as otherwise the "usual" recording will sound a bit too bright.
Where lies the optimum of flat on axis (+slowly increasing DI) and flat directivity (+gently tilted response on axis)?
We have very little experience with this, as I know only one speaker with DI at 400-800Hz having (about) the same value as at 2-4kHz (Palmer Orbit 11) and that one just arrived on the market.
So, my answer: I don't know. Probably the difference is small if the polar response is smooth.
My best guess: Aim for flat (and smooth) FR of the horizontal side reflections.

 

[TimVG]

Good question!
I assume, that a "perfectly" even polar response is not possible and there will always be some increase in DI/ERDI from 300 Hz to 5000Hz.

I've not yet seen a perfect one myself!

But if a truly constant directivity would be achievable, the on axis would probably have to be tilted down a little in most rooms as otherwise the "usual" recording will sound a bit too bright.

It was exactly this idea that made me ask. As in my mind the direct sound needs to be "flat" (0dB regression ON, maybe -0,3dB for the LW)

Where lies the optimum of flat on axis (+slowly increasing DI) and flat directivity (+gently tilted response on axis)?
We have very little experience with this, as I know only one speaker with DI at 400-800Hz having (about) the same value as at 2-4kHz (Palmer Orbit 11) and that one just arrived on the market.

I bought a pair shortly after they launched. In my room, which is small and well damped, there is nothing to complain about tonally (and in general). If I weren't short on time I'd love to put them in a large, lively room along with a few others.

So, my answer: I don't know. Probably the difference is small if the polar response is smooth.
My best guess: Aim for flat (and smooth) FR of the horizontal side reflections.

That is the idea, in fact the prototype was already quite close to what I'd consider ideal, but I was never able to test them in a large space.
With the redesign it will be easier to build them consistently but it may have a slight adverse affect with the vent being placed slightly further from the driver, we'll know soon enough.

 

[olieb]

That is the idea, in fact the prototype was already quite close to what I'd consider ideal, but I was never able to test them in a large space.

Well, those polars look pretty awesome!
I might need to build a pair (upon a bit of thinking: make that five). What kind of foam are you using?

They do have somewhat increasing DI, obviously. 60°-polar is going from -5dB/400Hz to -8dB/4000Hz.
How flat do you think you can make the DI?

It was exactly this idea that made me ask. As in my mind the direct sound needs to be "flat" (0dB regression ON, maybe -0,3dB for the LW)

I am not so sure about the need for flat on axis. After all this is all inside the circle of confusion and there are variations from recording to recording (everybody uses EQ). Toole does advocate the use of tone controls after all.
I experimented with tilted sound (mainly with headphones) and my experience is that it mainly changes the perception of distance of the source. So an ensemble gets a bit more depth and distance with a tilt, but often that is a good thing as many recordings sound quite close and up front for my taste (99% classical listener). And even a rather strong tilt might sound darker, but not necessarily unnatural.

 

[Tell]

I've now finished the cabinets (minus paint). I will have to check when I have a big room available for initial measurements.

Just out of curiosity: Would you, if given a choice, put more weight into achieving the most even polar response down to the modal range (constant), or towards a gradually sloping curve yet still controlled without obvious steps in directivity.

From what I understand the "ideal" speaker should have constant directivity as far down as possible, ie on and off axis response should be the exact same just with lower overall amplitude, so I guess that is what I would choose.
(I'll probably get some backlash writing this )

 

[TimVG]

Well, those polars look pretty awesome!
I might need to build a pair. What kind of foam are you using?

That was with a type of rockwool, for the sake of health and consistency I've now switched to Polyether SG25.

They do have somewhat increasing DI, obviously. 60°-polar is going from -5db/400Hz to -8db/4000Hz.
How flat do you think you can make the DI?

With this type of baffle this will be about as flat as we can go without introducing other anomalies. Give it some baffle width and you can get it flatter up to ~4,5kHz but then it'll fall anyway and the transition to the upper treble is noticably worse (imo) like on the Kii3

I am not so sure about the need for flat on axis. After all this is all inside the circle of confusion and there are variations from recording to recording (everybody uses EQ). Toole does advocate the use of tone controls after all.
I experimented with tilted sound (mainly with headphones) and my experience is that it mainly changes the perception of distance of the source. So an ensemble gets a bit more depth and distance with a tilt, but often that is a good thing as many recordings sound quite close and up front for my taste (99% classical listener). And even a rather strong tilt might sound darker, but not necessarily unnatural.

That is a fair point. I admit it is something that is probably mostly in my own mind. The reality is that the measurements will likely dictate the best compromise for this particular version. Either way at this point it is nitpicking of course. I'm going to do my very best to get some measurements in tomorrow, but I have some prepping to do.

 

[olieb]

Give it some baffle width and you can get it flatter up to ~4,5kHz but then it'll fall anyway and the transition to the upper treble is noticably worse (imo) like on the Kii3

The Kii is not flatter. 60°-polar is -4dB/500Hz and -7dB/4000Hz. So it is a bit wider but not flatter.
As I said, as far as I have seen the Orbit is somewhat unique with its balance of the off axis polars.

What you achieved above looks as good as one could wish. Though again I think the "pretty look" of straight curves might be a bit deceptive. My guess is, that the balance of levels (lower mids - upper mids - treble) is more important (audible) than some rather smooth wiggles. All this reaches the ear only after reflection.

At what distance did you measure those polars? The directivity looks quite high, considering a (true) cardioid would have -6dB/90° (-2.5dB/60°).
Even at 300Hz (wavelength >1m) the measurement is ≈-8dB/90° (<-4dB/60°).

 

[TimVG]

The Kii is not flatter. 60°-polar is -4dB/500Hz and -7dB/4000Hz. So it is a bit wider but not flatter.

I meant if you'd use a similar baffle on R2 you'd end up with something flatter up until 4,5kHz, as it would widen radiation like on the Kii3 in that particular region.

What you achieved above looks as good as one could wish. Though again I think the "pretty look" of straight curves might be a bit deceptive. My guess is, that the balance of levels (lower mids - upper mids - treble) is more important (audible) than some rather smooth wiggles. All this reaches the ear only after reflection.

I was more referring to the 'knee' (widening) in the response rather than the small wiggles which I'd argue are inaudible and could even be some sort of artifact for all we know.
The AsciLab C8C also shows a bit of widening in the treble (unclear in the DI because it uses the LW and not ON to calculate it) due to the baffle weing slightly wider than optimal for the waveguide used.

At what distance did you measure those polars? The directivity looks quite high, considering a (true) cardioid would have -6dB/90° (-2.5dB/60°).
Even at 300Hz (wavelength >1m) the measurement is ≈-8dB/90° (<-4dB/60°).

I want to say 2m, but it could have been 1,5m - it's been some years. I have a small church which I can use and I do have good resolution as I can get around 3m reflection free in any one direction.

One thing is 'true' cardioid was never the end goal here, from the start it was always about "controlling" (it doesn't need to be constant) directivity down to the modal region (aka a non-collapsing polar as you find on most speakers).

 

[olieb]

One thing is 'true' cardioid was never the end goal here, from the start it was always about "controlling" (it doesn't need to be constant) directivity down to the modal region (aka a non-collapsing polar as you find on most speakers).

Ah, my fault. I did not mean "true" in the sense of right or ideal, but as in "mathematical". With a small source (in relation to wavelength) it is just not easily possible to have more directivity than a cardioid, as higher order multipoles just need some size/distance between the sources (or a lot of cancelling).
One could make a hypercardioid but that would result in a different polar map.
That was the reason I was asking. Nothing wrong with higher directivity, presumably the opposite.

I want to say 2m, but it could have been 1,5m - it's been some years.

In any case plenty of distance for reliable results. Interesting.

I meant if you'd use a similar baffle on R2 you'd end up with something flatter up until 4,5kHz, as it would widen radiation like on the Kii3 in that particular region.

I see, good point, so a wider baffle would increase off axis level around 3-4kHz at the price of a knee above.
Still something that might be interesting as the off axis above 5kHz will become less important quickly with increasing frequency.
But again, probably these are very small effects, hard to tell apart and even harder to judge what is (audibly) better.

 

[TimVG]

I may be able to do some basic measurements tomorrow. But I need to go out for a new rotary laser as well, so -if- I get to measurements tomorrow they will have to be taken with a grain of salt, but a basic +/- 180 Horizontal sweep may be in the cards in a smaller room. It should tell us something at least.

 

[Justin Zazzi]

Exciting!

 

[TimVG]

Hello all.

I have done some preliminary measurements, but unfortunately the results are not as great as I had hoped. There is an anomaly due to the new slot shape and location that is not to my liking. I had expected this could be the case, but compared to the initial prototype I feel like we're not there yet. A silver lining is that the new top shape seems to have made a positive impact on tweeter performance.

I will experiment with the vents and damping some more to see if things can be improved upon, otherwise I'm afraid it's back to the drawing board.

 

[olieb]

I had expected this could be the case, but compared to the initial prototype I feel like we're not there yet.

I was wondering whether the slot that far back would work as intended.
The initial prototype is the one on your avatar?

A silver lining is that the new top shape seems to have made a positive impact on tweeter performance.

More spherical is often a good idea. The improvement is significant?

 

[TimVG]

The initial prototype is the one on your avatar?

Yes!

I was wondering whether the slot that far back would work as intended.

The issue was that the driver kept pushing the required damping material out of position. It took me a good while to get two consistent speakers. Hence trying to put the slot behind the magnet assembly.

More spherical is often a good idea. The improvement is significant?

It was already quite good, but it's slightly better now

 

[olieb]

The issue was that the driver kept pushing the required damping material out of position. It took me a good while to get two consistent speakers. Hence trying to put the slot behind the magnet assembly.

I understand. But would not the change to foam already take care of that?
And using a foam block like this (from two layers) could make a slot more in front possible?