With an understanding of the nature of these "flaws" it can be seen that normalised is not quite ideal.. it's more of a simplistic approach to the problem.
I just said that?
Don't follow the word "simplistic" here?
A graph relative to a reference shows just different data.
Therefore used for completely different things.
That doesn't have anything to do with simplicity?
In that case we could also call non-normalized graph "simplistic" since they show "flaws" in the directivity (relative data) not quite ideal.
Anyway, I already said that I think defending one graph over another is far from being a constructive conversation.
We need all of them for different reasons.
In reference to the DI of different cone materials....is the difference between the two after crossing over maybe more due to cone profile than cone material?
Cone material and vibration gets a lot of discussion but I think a lot is also determined by cone shape which defines the starting points for sound across the cone surface. Deep cones like guitar speakers and the venerated RS225 seem to get a broad dip somewhere in the response.
Thanks for posting these interesting data. I’m seeing more people work on directivity than used to be the case.
I’m trying to go in the opposite direction. To make a speaker with very narrow directivity. From your plots, I think that would look like a thin horizontal line of intense red and then lots of black above and below. (Below, say, 150 Hz it will be widely dispersed, regardless.) Any suggestions for how to get there?
I started a thread here:
Thanks.
I’m trying to go in the opposite direction. To make a speaker with very narrow directivity. From your plots, I think that would look like a thin horizontal line of intense red and then lots of black above and below. (Below, say, 150 Hz it will be widely dispersed, regardless.) Any suggestions for how to get there?
I started a thread here:
Thanks.
Tktran was trying to show that these two, otherwise identical, drivers have different off axis performance when they've got the same target acoustic filter applied.
One has a metal cone and one has a paper cone.
Most people would assume that the difference in sound, between the two, has to be because of the cone material. When in fact it could have absolutely nothing to do with that and everything to do with the slight differences in off axis performance.
Obviously it's not a surprise that there's a difference in the off axis, given the differences in dust cap geometries, but people have been making claims to the sound of cone materials for years. Perhaps the differences are more related to the cone profiles, typically used with specific materials, rather than the materials themselves.
Yes, that is the point of this thread, and it is a very valid point. Like any thread which is meant to be educational, the target audience is those who do not yet know the information or concept being presented.
It would be interesting how much material choice affects directivity/off axis performance - but of course it's rather a combination of material and geometry.
When I've taken MLS measurements of a paper cone driver and then an aluminum cone driver they sound very different. That's just listening to the phhhssst sound during the measurement. The paper sounds more smooth and pleasant, the metal harsher. Just as you might imagine. Because the measurements send out the full frequency range, so you hear everything. Including, for example, the breakup of the metal cone. It's not a valid comparison of paper versus aluminum because, by design, the paper and metal cones will sound different if no filters are applied. Nonetheless, I have wondered if it is measurement experience like this that makes people feel as though one material sounds different from another in a properly designed speaker. But this conversation is getting into more subtle differences like off axis response. Interesting! 😀
Hello,
Thank you for this thread.
It is interesting to see more measurements and exchanges around directivity. The directivity is a key point in the design as it is intrinsic to components and architecture choices. I started to make my own some weeks ago. I thought it have been more difficult but not. REW does most of the job. The second part is to export the directivity plot. I haven't tested Virtuix CAD but I wrote a small Python script to get the directivity plots. Of course there is a limitation in the result in the 200/400Hz area. I found the raw and normalized plot very informative, they are complementary. I have even a third one which is a kind of simulation of the listening window equalization, in the idea to limit the gain applied on the normalized view. This third plot gives an idea of a possible improvement by EQ. With that I would have probably design in a better way my previous loudspeakers. For now, I apply it to DML. Ok maybe a bit strange, this is an other story.
I have a question... DML are usually with open back. We see mainly directivity plots of front firing loudspeakers, some large planars at spinorama.org (other good sources of those plots). Are there examples of 360° plots of well appreciated open baffles?
Christian
Thank you for this thread.
It is interesting to see more measurements and exchanges around directivity. The directivity is a key point in the design as it is intrinsic to components and architecture choices. I started to make my own some weeks ago. I thought it have been more difficult but not. REW does most of the job. The second part is to export the directivity plot. I haven't tested Virtuix CAD but I wrote a small Python script to get the directivity plots. Of course there is a limitation in the result in the 200/400Hz area. I found the raw and normalized plot very informative, they are complementary. I have even a third one which is a kind of simulation of the listening window equalization, in the idea to limit the gain applied on the normalized view. This third plot gives an idea of a possible improvement by EQ. With that I would have probably design in a better way my previous loudspeakers. For now, I apply it to DML. Ok maybe a bit strange, this is an other story.
I have a question... DML are usually with open back. We see mainly directivity plots of front firing loudspeakers, some large planars at spinorama.org (other good sources of those plots). Are there examples of 360° plots of well appreciated open baffles?
Christian
I don't see the interest in that?
Mostly because that question can be answered with a bit of mechanical and acoustic knowledge.
If we assume the same cone (as in shape, size and form), same dustcap and same all other parts, PLUS we forget all kinds of other non-material issues like comb filtering and other off-axis interference issues, we know that ANY driver works like a nice moving piston up till the first brake-up modes.
That also gives you the answer.
I've blind tested an awful lot of drivers in my life, and I don't share the general idea that they seem to sound so wildly different at all.
Finding two that are very close to identical (incl dustcap), is pretty difficult though.
What's even more tricky is to find identical ones that don't have some very clear and obvious issues, especially in the important frequency region.
Think about very obvious harmonics around 800-3000Hz.
The fast majority of "quick and dirty" test are even not done (double) blind, or compare drivers that measure very obviously different.
Like for example playing full-range with a proper demodulation ring, vs no ring.
Or drivers with a very different BL(x) curve.
Or just claim to hear difference because of materials, yet the entire system + filter is completely different.
This is what he concluded:
So in other words, concluding that the difference in CONE material.
Yet, the have entirely different dustcaps? (and maybe more)
What I am trying to point out here, is NOT the fact IF there is a difference.
But the fact that these conclusions can't be made from this simple experiment, since there are other variables and parameters that also have an effect on off-axis behavior.
So no, the drivers are NOT "otherwise identical".
I have even pretty poor eyesight, and I think it's pretty obvious to spot the difference here?
Blind test using EQ on a Dayton Reference aluminum driver vs paper. Same size drivers, cone profile and phase plug.
Once again: only if these two are equalized to exactly the same transfer function.
Wanna make a bet under that condition the differences disapppear like snow before the sun? Even in the plots shown by tktran303 the differences are i.m.o. relatively small.
Ah I see that's what you were trying to say. Being pedantic about the use of the word cone. I took that to mean the entire radiating surface which would be inclusive of material, geometry, dust cap, surround etc
Yes, we do.....
But for the blind and deaf these matters cannot be repeated 'nuff: time and again a number of members here- occasionally not the least unfortunately- keep on making far reaching statements about driver performance (and applicability) on the basis of raw and totally uncorrected SPL characteristics (or some magic diaphragm material..).
But for the blind and deaf these matters cannot be repeated 'nuff: time and again a number of members here- occasionally not the least unfortunately- keep on making far reaching statements about driver performance (and applicability) on the basis of raw and totally uncorrected SPL characteristics (or some magic diaphragm material..).
It's clearly not being pedantic if:
1 - it's the subject of the discussion
2 - people clearly put another meaning to it (as shown as in this exact response and example)
These drivers have been tested by many people and magazines.
Unfortunately it doesn't immediately proof much either.
Because of acoustic mechanical impedances, different material can (and often will) behave differently to the rest of the mechanical design.
Resulting in some kind of (bad) resonances for example.
These resonances have nothing to do with the cone material itself, but the interaction between everything.
I very much in doubt that Dayton put in the effort to optimize all of this.
So alas, still doesn't say much at all.