Where is the cutoff for "narrow" dispersion speakers and "wide" dispersion speakers?

[a4eaudio]

So this is really asking for a subjective criteria to qualitatively categorize objective data.

We often speak of narrow and wide dispersion speakers. Looking at Amirm's Beamwidth graphs, what level of beamwidth or greater would you generally consider "wide" and what level of beamwidth or less would you generally consider "narrow"? [Yes, I could go through 100-200 reviews, collect the data and come up with an objective interpretation but that is way too much work ]

As examples:
GRIMM Audio LS1c had beamwidth of +/- 90 degrees and was labeled "very" wide
Revel F206 Tower +/- 65 degrees labeled "wide"
Focal Aria 906 +/- 60 degrees labeled "good"
Neumann KH150 +/- 50 degrees labeled a "bit narrow"

 

[eddantes]

60+ Wide | 50- Narrow (In my opinion)

 

[staticV3]

Perhaps @pierre could post a .csv or similar with the computed directivity value of each speaker, then we could run some analysis to see if two "camps" can be observed.

 

[Arindal]

what level of beamwidth or greater would you generally consider "wide" and what level of beamwidth or less would you generally consider "narrow"?

We should take into account that dispersion angle might greatly vary over frequencies, even in the localizable bands. Furthermore the dispersion angle might be not rotationally symmetrical but showing difference between horizontal and vertical dispersion.

Describing dispersion as either wide or narrow based on an angle figure is pretty imprecise, as no information is given how quickly the level drops at which frequency outside the defined window.

I would prefer to describe speakers by directivity index in dB and would consider everything from 2 to 5dB ´wide´, 5-9dB ´medium´ and 9+dB ´narrow´ or ´highly directional´. Ideally looking at the d.i. at 1K, but taking the octave above (and maybe below) that into account.

 

[a4eaudio]

Perhaps @pierre could post a .csv or similar with the computed directivity value of each speaker, then we could run some analysis to see if two "camps" can be observed.
View attachment 454779

Ah, this is really helpful. I have gone into the "Data" area of spinorama.org quite a bit but this was never what I was focusing on.
I can easily get data for a LOT of speakers myself, whereas if I had to search the ASR forum for Amirm's reviews that would take a LOT of time.

 

[a4eaudio]

We should take into account that dispersion angle might greatly vary over frequencies, even in the localizable bands. Furthermore the dispersion angle might be not rotationally symmetrical but showing difference between horizontal and vertical dispersion.

Describing dispersion as either wide or narrow based on an angle figure is pretty imprecise, as not information is given how quickly the level drops at which frequency outside the defined window.

I would prefer to describe speakers by directivity index in dB and would consider everything from 2 to 5dB ´wide´, 5-9dB ´medium´ and 9+dB ´narrow´ or ´highly directional´. Ideally looking at the d.i. at 1K, but taking the octave above (and maybe below) that into account.

I don't disagree with anything you are saying here, but I'm looking for the "shortcut" of just an informal categorization.

But, I have considered what you are saying...consider the image below for the GRIMM Audio speaker which is "called" +/- 90 degrees even though it is less than 90 degrees starting around 3.5kHz. But the data that @staticV3 points to above presents it in a uniform measure 1kHz to 10kHz which is exactly what I am looking for. By the way, the Grimm Audio is in Spinorama.org and is +/- 70 from 1kHz to 10kHz.

 

[Arindal]

consider the image below for the GRIMM Audio speaker which is "called" +/- 90 degrees even though it is less than 90 degrees starting around 3.5kHz.

If I am not mistaken, the calculated ´sound power DI´ in the FR diagram is giving you exactly that information I was mentioning, without having to average sound power from two separated isobaric graphs. In this case you have a DI well under 5dB for most of the localizable bands with broad-banded exceptions around 1.5k (transition interference) and above 4K (tweeter/waveguide dispersion control taking over), reaching a maximum of 5.5dB or 7dB respectively (if we ignore 10kHz+).

In my understanding this still qualifies as a wide-dispersion speaker.

 

[pierre]

Perhaps @pierre could post a .csv or similar with the computed directivity value of each speaker, then we could run some analysis to see if two "camps" can be observed.
View attachment 454779

The spreadsheet is here.

3 graphs:

distribution of SPDI std dev (how smooth the directivity is):

Histogram of horizontal directivity (in deg)

Histogram of vertical directivity (in deg)

 

[Arindal]

I do not have the numeric value above in the spreadsheet but I will add them.

Is the D.I. in dB, and averaged over the whole audible band, or at a specific frequency/in a specific band?

 

[staticV3]

The spreadsheet is here. I do not have the numeric value above in the spreadsheet but I will add them.
Using the horizontal directivity, I find a nicely gaussian histogram:
View attachment 454834

Awesome! Thanks a bunch!

 

[pierre]

Is the D.I. in dB, and averaged over the whole audible band, or at a specific frequency/in a specific band?

It is computed here.
It normalises the data with respect to the on axis spl and compute the angle where the EWM is below -6 across the 1k-10khz range.

EDIT: Fixed the above histograms.

 

[a4eaudio]

...
In my understanding this still qualifies as a wide-dispersion speaker.

This is partially why I asked the question. There is no generally accepted definition of "wide dispersion" speaker. If we look at the data, clearly +/-70 or +/-90, is well above normal and could be considered wide disperstion.

 

[a4eaudio]

Last night, before Pierre posted, I downloaded the data I could find available as a spreadsheet from Spinorama.org. This allowed me to have the speaker name and preference rating filled in without me typing so much. I filtered the data as follows:

• Only preference score 5.0 or above
• Only reviews by ASR or Erin's Audio Corner
• I had to remove duplicates (where both reviewed) or some where I couldn't access the data ("403 Forbidden" error)

This gave me 102 speakers. I then went to the data sheets and found the Beamwidth cutoff, which as Pierre mentions above is the average angle where the SPL is down 6dB limited to 1kHz to 10kHz. The Spinorama numbers are in even increments of 10, which makes sense to me as there is no "real" difference between a Beamwidth of 51% and 54%.

In some cases, they are not symmetric, so the numbers below in red are the average of the positive and negative angle. (-80/+70 is 75 degrees, -60/+50 is 55 degrees, etc.)

To simplify things, here is the breakdown with 65+ degree consolidated as the top category and 40- degree as the bottom category. Note here that I have included the -60/+50=55 degrees in with 60 degree group rather than the 50 degree group.

Is the difference between +/-60 degrees and +/-50 degrees an important threshold? If not, then 83% the speakers with a Preference Rating of 5.0 are in the same "normal" category of +/-50 to +/-60 degrees Beamwidth.

 

[a4eaudio]

By the way, the outliers in the data above are:
Selah Integrity -80/+70 degrees
Meyer Sound Amie +/- 30 degrees

 

[BenB]

I don't particularly like beamwidth as a measurement of vertical dispersion. If there's a narrow mainlobe, along with numerous grating lobes, I'd consider that a wider dispersion speaker compared to one with a wider mainlobe but less spreading or lower sidelobe levels. The latter likely has a higher DI, too.

 

[dualazmak]

So this is really asking for a subjective criteria to qualitatively categorize objective data.

We often speak of narrow and wide dispersion speakers.
....
....

I think it would be difficult establishing "subjective criteria to qualitatively categorize objective data"...

In DSP-DIY tuning of existing multiple SP drivers, sometimes your point would be dependent on each of the SP drivers and/or Fq-covering ranges, and it would be "the pinnacle of the perfection" and a kind of "one of the best practices" in DSP-DIY tuning, I assume.

Spoiler: Just some example cases regarding super-tweeters on my project thread...

- A new series of audio experiments on reflective wide-3D dispersion of super-tweeter sound using random-surface hard-heavy material:
Part-1_ Background, experimental settings, initial preliminary listening tests: #912
Part-2_ Comparison of catalogue specifications of metal horn super-tweeter (ST) FOSTEX T925A and YAMAHA Beryllium dome tweeter (TW) JA-0513; start of intensive listening sessions with wide-3D reflective dispersion of ST sound: #921
Part-3_ Listening evaluation of sound stage (sound image) using excellent-recording-quality lute duet tracks: #926
Part-3.1_ Listening evaluation of sound stage (sound image) using excellent-recording-quality jazz trio album: #927
Part-4_Provisional conclusion to use Case-2 reverse reflective dispersion setting in default daily music listening: #929

 

[cavedriver]

Histogram of horizontal directivity (in deg)
View attachment 454862

Lol - I was going to jokingly reply "52.5 degrees! It's been empirically determined". Then I saw this graph, which ironically puts the peak ("mode") at right around that number.

 

[Mort]

Last night, before Pierre posted, I downloaded the data I could find available as a spreadsheet from Spinorama.org. This allowed me to have the speaker name and preference rating filled in without me typing so much. I filtered the data as follows:

• Only preference score 5.0 or above
• Only reviews by ASR or Erin's Audio Corner
• I had to remove duplicates (where both reviewed) or some where I couldn't access the data ("403 Forbidden" error)

This gave me 102 speakers. I then went to the data sheets and found the Beamwidth cutoff, which as Pierre mentions above is the average angle where the SPL is down 6dB limited to 1kHz to 10kHz. The Spinorama numbers are in even increments of 10, which makes sense to me as there is no "real" difference between a Beamwidth of 51% and 54%.

In some cases, they are not symmetric, so the numbers below in red are the average of the positive and negative angle. (-80/+70 is 75 degrees, -60/+50 is 55 degrees, etc.)

View attachment 454895

To simplify things, here is the breakdown with 65+ degree consolidated as the top category and 40- degree as the bottom category. Note here that I have included the -60/+50=55 degrees in with 60 degree group rather than the 50 degree group.
View attachment 454896

Is the difference between +/-60 degrees and +/-50 degrees an important threshold? If not, then 83% the speakers with a Preference Rating of 5.0 are in the same "normal" category of +/-50 to +/-60 degrees Beamwidth.

View attachment 454898

I'm wondering if the Philharmonic Audio BMR monitor was in your mix, it should have been, I think. It's quite wide.

 

[a4eaudio]

I'm wondering if the Philharmonic Audio BMR monitor was in your mix, it should have been, I think. It's quite wide.

It was not in the list I had to work with.
I went to Spinorama to manually find the data on the BMR and strangely I cannot find the dispersion data that I was looking at on Saturday.