Some speaker driver measurements...

Keep in mind that directivity itself should also be taken a bit on a practical way.
A "perfect" directivity up to 20kHz looks nice, but I really doubt if that's very useful?
I have no idea what a "perfect" directivity is or means in this context. I can only say that I have a very different view of directivity compared to most (at least I think this is the case). I want the DI to be constant and low (broad, even radiation pattern that is indepedent of frequency) and then I WANT the room to be relatively live with especially the front wall a very hard, planar surface that will generate a strong reflection. The floor and to a lesser extent the walls can also be like this, so that they reflect back into the listening space something similar to that which reached them from the loudspeaker. I find this to be very different from the large majority of rooms, which often have damping and diffusing surfaces and materials. That might be fine for a directional speaker, but it does not complement dipolar loudspeakers.

Local absorption above around 8kHz is extremely high (this can be actually very easily seen with difference with in-room measurements for example).

I slight tilt is preferred by most I think, although it also really depends on you hearing (which sometimes also means age :( )
If by saying "local adsorption" you mean the room surfaces, yes, this is often the case in domestic rooms especially in the USA. But one's hearing sensitivity is dropping steeply starting around 5-6kHz, and the ability to differentiate the exact level of sound "up there" is also decreasing, so it is not as big a problem as it would appear.
 
SL did a similar thing with his dipoles, a -3.3dB shelving EQ centred at about 1740Hz, Orion and LX521 were way too bright for me without it engaged. I still use a variation of that same EQ now. Having a tilt at that point does seem to do something right to the sound.

https://www.linkwitzlab.com/orion-rev3.htm
I'm not familiar with that EQ profile, but it doesn't surprise me. SL may have left this as a user-adjustable parameter to suit the various room responses that a user might have, e.g. more or less overstuffed furniture, carpeting, drapes, etc.

I currently use about 4dB to 5dB of down-tilt between 200 and 2kHz or so, plus some other tweaks that I have developed for my dipole speakers that improve the tonal balance and that are based in published audio research findings, etc. The sound is probably a bit on the "rich" side of things because of this, but each time I begin playback (for the day, etc.) and I first hear the sound I feel content that it is accurate.
 
I have no idea what a "perfect" directivity is or means in this context. I can only say that I have a very different view of directivity compared to most (at least I think this is the case). I want the DI to be constant and low (broad, even radiation pattern that is indepedent of frequency) and then I WANT the room to be relatively live with especially the front wall a very hard, planar surface that will generate a strong reflection. The floor and to a lesser extent the walls can also be like this, so that they reflect back into the listening space something similar to that which reached them from the loudspeaker. I find this to be very different from the large majority of rooms, which often have damping and diffusing surfaces and materials. That might be fine for a directional speaker, but it does not complement dipolar loudspeakers.


If by saying "local adsorption" you mean the room surfaces, yes, this is often the case in domestic rooms especially in the USA. But one's hearing sensitivity is dropping steeply starting around 5-6kHz, and the ability to differentiate the exact level of sound "up there" is also decreasing, so it is not as big a problem as it would appear.
Perfect in a sense of a a nice smooth constant directivity.
Or in other words, a straight line in the DI vs freq graph.
Not what one's preference is.

Local absorption, can be room surfaces yes.
Even in a pretty empty room, you will see quite a big of spread above the 8kHz.
My point was, a lot of people really care so much on getting a very straight DI vs freq curve (or eye ball it in the off-axis response). Even above 8kHz.

More reflections means a more lively sound but because of the smearing effect, also less details.
By definition you can't have both.
 

diyiggy

Member
2019-01-16 12:22 am
The discussion is turning into where people like to have dips in the power response as the progressive db loss from circa 200 hz to the high end... non saying off axis preference, use of third party wave guides.

It diserves a thread as here it is about raw drivers measurements. The actual discussion has more to see whith curve shapping and if or not one has a DSP or not for fine tunning. If for instance a dip at 1750 hz is not knew about subjective liking -I remember Hiraga to like slihthy recession in some speakers reviews between 1k to 2k hz- it often involves filter knowledge as it can be in the pass band of a filter cut off.

I am sure some would like to know more about off axis datasheet reading about tweeters when it comes to personal preference like soundstage, brigthness, mattier sound, etc... before the use of EQ. I do not like either some zones like the 600 hz area, the 4k hz to 7 k hz, also like a whole 5 db or more dive from the beginning to the end... but not sure how one does to chose raw drivers related to that.

Does it deserves a split thread asked by the op to a mod ?
 
Perfect in a sense of a a nice smooth constant directivity.
Or in other words, a straight line in the DI vs freq graph.
Not what one's preference is.

Local absorption, can be room surfaces yes.
Even in a pretty empty room, you will see quite a big of spread above the 8kHz.
My point was, a lot of people really care so much on getting a very straight DI vs freq curve (or eye ball it in the off-axis response). Even above 8kHz.

More reflections means a more lively sound but because of the smearing effect, also less details.
By definition you can't have both.
My understanding of the current body of research is that after a certain threshold from the direct sound, any reflected sound can add spaciousness without negatively affecting the soundstage or imaging. So you can have both, as long as your reflections are sufficiently delayed. Seems like the magic number is somewhere around 15-25 ms after direct incidence.

EDIT: This has nothing to do with speaker measurements.
 

diyiggy

Member
2019-01-16 12:22 am
Yep I saw,it was to push your idea. How can we talk about wall reflections delay in such a thread as we have all different rooms and speaker distance to wall and of course triangle between speakers and sweet spot ratio?
The op has normalized his measurements and now it is starting off axis in relation to the front baffle, wave guide or not... hard to follow. Discussion is very interesting but disearves a thread for going deeper...imho
 

Brett

Member
2002-01-07 6:02 pm
There was a well known series of listening tests of the Revel Salon 2 vs JBL M2 put on by a Harman dealer. Toole participated in one. The Revel was preferred at all events. Not by huge margin, but preferred nonetheless.
Drawing a bit of a long bow to assume the preferences were based solely on directivity. The M2 is also a wide 120* so isn't necessarily representative of other highly controlled directivity designs such as a Synergy. Preference of one to the other, say Revel to Synergy would also be highly room dependent, so the mentioned test, whilst interesting, doesn't actually mean much.
I am sure some would like to know more about off axis datasheet reading about tweeters
You mean like actual useful measurements? Yeah, I want to see them. On axis is pretty much worthless on it's own.
 
I was certainly inspired to write about which drivers to use, based on their dispersion, and upper and lower limits in terms of distortion, to select a suitable crossover point. Which is clearly related to directivity. So I don't think we're completely off topic.

Say a 3" cone and dome vs a traditional 4" cone, what usefulness does the Lavoce FAN030.71 and Bliesma M74, that Yevgeniy measured, offer, in relationship to the competition.

My interpretation is that the Lavoce FAN030.71 looks like very high value midrange that would fit many different applications. Is it the best of it's size? Probably not but it doesn't do many things wrong, not for US$30...

The M74 is boutique product that I'd use if I needed the high sensitivity. But it doesn't look like it can play as low as the ATC or Volt 3" dome. And the dispersion at the top end doesn't beat, say, a Vifa NE123W-08. Maybe I'm looking at it from the wrong angle?... ;-)
 
Last edited:

Brett

Member
2002-01-07 6:02 pm
I was certainly inspired to write about which drivers to use, based on their dispersion, and upper and lower limits in terms of distortion, to select a suitable crossover point. Which is clearly related to directivity. So I don't think we're completely off topic.
I agree 100% and need to know all of that before I decide if a driver is useful for an application, and how I'm potentially going to integrate it.
Say a 3" cone and dome vs a traditional 4" cone, what usefulness does the Lavoce FAN030.71 and Bliesma M74, that Yevgeniy measured, offer, in relationship to the competition.
Based on the LV 3" measurements, I'm going to try some of the 6" as the ATC is still larger physically.
 
If you're aiming for a more omnidirectional approach, I would suggest a 3/4 (or smaller, but you will loose maxSPL very quickly)

There is much more than just raw performance, there is also the practicality of it all.
Something that seems to be very difficult for a lot of people to get their head around.

If you're adding a 10 or 12 inch on the lower end, you might as well just use a 2-way system with a 4 inch from 150-200Hz and up. As soon as you make a symmetrical setup with those woofer, one can actually even cross pretty high.
Something like a MTM or similar if you wanna go a little louder.
 
I was certainly inspired to write about which drivers to use, based on their dispersion, and upper and lower limits in terms of distortion, to select a suitable crossover point. Which is clearly related to directivity. So I don't think we're completely off topic.

Say a 3" cone and dome vs a traditional 4" cone, what usefulness does the Lavoce FAN030.71 and Bliesma M74, that Yevgeniy measured, offer, in relationship to the competition.

My interpretation is that the Lavoce FAN030.71 looks like very high value midrange that would fit many different applications. Is it the best of it's size? Probably not but it doesn't do many things wrong, not for US$30...

The M74 is boutique product that I'd use if I needed the high sensitivity. But it doesn't look like it can play as low as the ATC or Volt 3" dome. And the dispersion at the top end doesn't beat, say, a Vifa NE123W-08. Maybe I'm looking at it from the wrong angle?... ;-)
IMO the wrong angle is thinking any one driver is perfect. They all have flaws. The Be M74 has a double hump in the impulse. The Fan 30.71 has a slow decay in the midrange. (edit: i should say though... a damn good value)
 

fluid

Member
2009-01-24 2:20 pm
I think he means the step response as there was no Impulse response graphs in the review.

M74B-6_step.png

There is also this useful paragraph included which is very relevant

"Of course, the transient response of the silk version looks the most beautiful of all, against which the rest of the curves bear a strong imprint of resonant frequency response bursts. With proper filtering in the crossover, all these ripples disappear and the curves will be very difficult to distinguish, so you don’t need to give it an increased value . After all, the transient response is only a direct mirroring of the frequency response."
 
I think he means the step response as there was no Impulse response graphs in the review.

View attachment 1019815

There is also this useful paragraph included which is very relevant

"Of course, the transient response of the silk version looks the most beautiful of all, against which the rest of the curves bear a strong imprint of resonant frequency response bursts. With proper filtering in the crossover, all these ripples disappear and the curves will be very difficult to distinguish, so you don’t need to give it an increased value . After all, the transient response is only a direct mirroring of the frequency response."
Yep, basically anything with a rigid cone resonance is going to have a jaggy step response. I never look at them anymore.
 
Does anyone know what it means when an IMD measurement is captioned "(4to1)"? I'm guessing it's where the one tone is +/- 6 dB vs the other... but this isn't always the case from what I can see.

Also, I don't understand why the output level during the IMD measurements seems to change for different drivers. When comparing drivers with the same Sd, you'd expect the same output for a given Xmax, but it doesn't seem to be the case.

https://hificompass.com/en/speakers/measurements/sbacoustics/sb-acoustics-sb17nbac35-8https://hificompass.com/en/speakers/measurements/purifi/purifi-ptt65w08-01b-ptt65x08-nfa-01
With the "3mm 30 Hz + 255 Hz" test on those, there's nearly a 6 dB difference in the 255 Hz signal between the two (while the 30 Hz seems to be about the same). Could this just be a case where the Purifi measurement is just mislabeled and it's not "4 to 1"?
 
Does anyone know what it means when an IMD measurement is captioned "(4to1)"? I'm guessing it's where the one tone is +/- 6 dB vs the other... but this isn't always the case from what I can see.

Also, I don't understand why the output level during the IMD measurements seems to change for different drivers. When comparing drivers with the same Sd, you'd expect the same output for a given Xmax, but it doesn't seem to be the case.

https://hificompass.com/en/speakers/measurements/sbacoustics/sb-acoustics-sb17nbac35-8https://hificompass.com/en/speakers/measurements/purifi/purifi-ptt65w08-01b-ptt65x08-nfa-01
With the "3mm 30 Hz + 255 Hz" test on those, there's nearly a 6 dB difference in the 255 Hz signal between the two (while the 30 Hz seems to be about the same). Could this just be a case where the Purifi measurement is just mislabeled and it's not "4 to 1"?
Don't take in account the absolute level of 30 and 255 Hz components in IMD plots, they mean just nothing. All you need to know is:
  • the peak membrane displacement at 30 Hz
  • the level difference between 255 Hz and IMA2 components
4:1 - it is the voltage ratio between 30 and 255 Hz, not for you but for me only need