Poor directivity with good on axis response is really bad?

[dogmamann]

HI everyone, this is my first post here. I had been following this forum for long, but I think its time that I pitch in for asking something which had been bothering me for some time. Attached is an example of a speaker with a decent frequency response, and distortion figures. But when I look at the directivity from this graph, it is all around the place. Such a speaker, will it sound good anywhere?

To me everything here looks good , until 60 degree and 90 degree respsonse. Those have a huge hump in upper mids, unlike the state of the art speakers. Or am I reading this graph wrong?

 

[fpitas]

Depends on the room, really. In a highly reflective room all that off-axis weirdness makes it back to your ears. If you use them in a highly damped room, they might be fine.

 

[dogmamann]

Depends on the room, really. In a highly reflective room all that off-axis weirdness makes it back to your ears. If you use them in a highly damped room, they might be fine.

That means, at 60 degree to 90 degrees this would need some heavy damping which absorbs the extra energy from these frequencies alone, right? Is it even possible to find materials that absorb only certain frequencies?

 

[TurtlePaul]

Speakers with good on-axis but worse off axis can sound ok if you listen in nearfield on-axis but sounds worse the further back you get from the speaker or further off-axis.

 

[oivavoi]

Speakers with good on-axis but worse off axis can sound ok if you listen in nearfield on-axis but sounds worse the further back you get from the speaker or further off-axis.

This is my experience too. If placed well away from room boundaries and listened to in the near-field directivity doesn't matter that much. The farther away one sits from the speakers, the more directivity tends to matter.

 

[Absolute]

Do you know how they have measured the speaker? Klippel/anechoic non-smoothed measurements would reveal whether or not those bumps are purely directivity errors, diffractions/resonances smoothed or a combination. Directivity error only doesn't need to be a major drawback at those angles.
if so I would change my M2's that have rather similar 60 degree dispersion;

 

[ROOSKIE]

Speakers with good on-axis but worse off axis can sound ok if you listen in nearfield on-axis but sounds worse the further back you get from the speaker or further off-axis.

This is my experience too. If placed well away from room boundaries and listened to in the near-field directivity doesn't matter that much. The farther away one sits from the speakers, the more directivity tends to matter.

The closer you are the more extreme off axis reflections reach you as 1st refections. So, I am dubious of your statement.
Further away they don't reach as 1st reflections but the extra energy is additive and affects the in room steady state equilibrium/over all tonality.

I think in both cases they will not sound as good from an objective/accuracy based lens as a otherwise comparable set that does not have this issue. Unless the room has very minimal 1st reflections.

Unfortunately, blind testing would be the only way to really know and most of us simply can't. Most solid published testing is farfield. Nearfield is next??

Subjectively someone may like them, especially if one is already used to that sound sig or loves the brand or similar.

Personally, for now, I don't want them based off the graph. What speakers are they?

 

[ROOSKIE]

Do you know how they have measured the speaker? Klippel/anechoic non-smoothed measurements would reveal whether or not those bumps are purely directivity errors, diffractions/resonances smoothed or a combination. Directivity error only doesn't need to be a major drawback at those angles.
if so I would change my M2's that have rather similar 60 degree dispersion;

View attachment 250455View attachment 250456

But the M2s have really decent directivity. The on and off axis are similar and quite smoothly rolling away, so easy to eq.

The OPs example does not. It will difficult to eq well.

M2 is below...very decent easy to eq

 

[TurtlePaul]

The closer you are the more extreme off axis reflections reach you as 1st refections. So, I am dubious of your statement.
Further away they don't reach as 1st reflections but the extra energy is additive and affects the in room steady state equilibrium/over all tonality.

It is true that the closer you are the further off-axis the first reflections are. However, this is overwhelmed by the direct/reflected mix going way way up the closer you get to the speakers. If your normal listening position is four meters and you move closer to the speaker to one meter distance, the on-axis SPL will increase by 12 dB while first reflection decreases by a couple of dB and the off-axis sound power will stay relatively the same, so the soundfield will become more dominated by the direct sound.

 

[Tim Link]

I've been wondering about this too. Life is easier in general if the off-axis is smooth. But even if it is, should the total sound power have the same downward tilt for all speakers? I'm starting to think that the more directional the speaker is the less tilt it should have. If a speakers was somehow able to project a laser beam of sound all the way down to 500 Hz my hunch is the sound power should be flat down to 500 Hz. If it's a pure omni all the way up to 20kHz in a small highly reflective room it may need a very steep slope. I got to thinking about this because I finally equalized my SSCS-5 speakers to my liking and noticed the EQ curve has a downward tilt, even though the SSCS-5 total room response of that speaker without EQ already has a downward tilt - with some bumps and dips. It's a small baffle speaker with two tweeters, a tiny one and a normal sized one so I assume it's pretty high dispersion up into higher frequencies, so maybe that's why Amir found it so bright even though it seemed like it should be decent according to the preference score. The slope needs to be steeper for a typical home environment - which is a fairly variable thing so the speaker sounds better out of the box for some than others.

 

[ROOSKIE]

It is true that the closer you are the further off-axis the first reflections are. However, this is overwhelmed by the direct/reflected mix going way way up the closer you get to the speakers. If your normal listening position is four meters and you move closer to the speaker to one meter distance, the on-axis SPL will increase by 12 dB while first reflection decreases by a couple of dB and the off-axis sound power will stay relatively the same, so the soundfield will become more dominated by the direct sound.

Good points and I thought about that as well. That drop in 1st reflection assumes a similar sized space though. When I think of near field I often picture a smallish space - especially in a private home studio or office. Many nearfield set-ups are very close to boundaries including the desk itself and 1st reflections could be powerful.
Obviously some nearfield arrangements are actually in a large space, spaces with good breathing room.

I would still rather buy a speaker with good directivity and smooth even dispersion and fall off. For peace of mind and also as one moves around in the nearfield the relative sound field to ear angle changes rapidly.

Also of course the speaker must be designed to sum properly in the near and many do not (and many are just fine even if originally tailored for distance), that would another big factor to involve in testing for near field listening.

--------
One thing that often doesn't get mentioned that relates is when there is non-linear absorption or poorly designed diffusion that corrupts even/smooth off axis response. That could turn a well measuring speaker into effectively a poor one very quickly.

 

[dogmamann]

From what I can understand, the first reflections matter if there are walls on either side of the speakers which can reflect enough energy back to the listening spot. That means this particular speaker would sound only good if its kept in a room without side walls near to them. This should have been a good speaker if this was not a problem but unfortunately this again means the brands like Revel, Genelec , Neumann are the best.

 

[dogmamann]

Do you know how they have measured the speaker? Klippel/anechoic non-smoothed measurements would reveal whether or not those bumps are purely directivity errors, diffractions/resonances smoothed or a combination. Directivity error only doesn't need to be a major drawback at those angles.
if so I would change my M2's that have rather similar 60 degree dispersion;

View attachment 250455View attachment 250456

This is from Stereoplay magazine. Don't know how they measure it.

 

[fineMen]

HI everyone, this is my first post here. I had been following this forum for long, but I think its time that I pitch in for asking something which had been bothering me for some time. Attached is an example of a speaker with a decent frequency response, and distortion figures. But when I look at the directivity from this graph, it is all around the place. Such a speaker, will it sound good anywhere?

To me everything here looks good , until 60 degree and 90 degree respsonse. Those have a huge hump in upper mids, unlike the state of the art speakers. Or am I reading this graph wrong?

One might assume that off-axis directivity starts to depart from acceptable at about 45°. That would leave about one third (measured by surface area) of the of full frontal half-sphere left intact, but the other two third of the frontal half-sphere would be affected. Let alone the backward half-sphere.

One could say, that the good frontal portion is only 30% of the whole story. And only so if one is graceful enough to ignore the backward sphere altogether.

The degree of contamination of the reverberant sound field depends on the reflection coefficient of the rooms's boundaries in the referring frequency range. In this particular case you would need to dampen at 4kHz a lot, while not also dampen the range below and above, which is actually hard o do.

That is, why the spinorama was invented. To avoid specifically that glaring sound from common BBC-like 2-way designs in peoples homes. The BBC coped with it by inventing the British 'vocing' with deliberately subdued treble, and not to forget the infamous BBC dip.

Btw: I can read German--the text says: "perfectly balanced". Not really, but just typically bad, as expected.

 

[Blumlein 88]

I can provide some anecdotal experience. Vandersteen and Thiel speakers of yore had pretty good on axis response and 1st order crossovers. They even could do a good squarewave in just the right spot. Neither were very resonant. However, due to the 1st order crossovers off axis the response was a roller coaster. In damped rooms in limited listening positions they could sound pretty good. They were a big nightmare to make sound good and in some rooms just didn't work.

Once you have used a few of these JBL/Harman/Revel speakers with well controlled directivity and even off axis response, wow what a breeze they are to get sounding pretty good in nearly any room. They sound very good with a little care and much less hassle than most other speakers.

 

[MarkS]

Vandersteen and Thiel had huge legions of fans (and Vandersteen still does). I was always underwhelmed by the ones I heard, but others who heard them with me (so same room and music played) thought they were wonderful.

 

[dogmamann]

One might assume that off-axis directivity starts to depart from acceptable at about 45°. That would leave about one third (measured by surface area) of the of full frontal half-sphere left intact, but the other two third of the frontal half-sphere would be affected. Let alone the backward half-sphere.

One could say, that the good frontal portion is only 30% of the whole story. And only so if one is graceful enough to ignore the backward sphere altogether.

The degree of contamination of the reverberant sound field depends on the reflection coefficient of the rooms's boundaries in the referring frequency range. In this particular case you would need to dampen at 4kHz a lot, while not also dampen the range below and above, which is actually hard o do.

That is, why the spinorama was invented. To avoid specifically that glaring sound from common BBC-like 2-way designs in peoples homes. The BBC coped with it by inventing the British 'vocing' with deliberately subdued treble, and not to forget the infamous BBC dip.

Btw: I can read German--the text says: "perfectly balanced". Not really, but just typically bad, as expected.

I guess they meant perfectly balanced in their listening room, which would have been treated, but graphs shows poor off axis response. I feel then, the companies research of directivity is rather not upto the level of other brands. But the low end distortion and extension are excellent for a speaker costing 5000 euros. Unfortunately we cannot equalize this one if I understood correctly

 

[ROOSKIE]

I guess they meant perfectly balanced in their listening room, which would have been treated, but graphs shows poor off axis response. I feel then, the companies research of directivity is rather not upto the level of other brands. But the low end distortion and extension are excellent for a speaker costing 5000 euros. Unfortunately we cannot equalize this one if I understood correctly

Based on what is published so far, I would not pay 5000 euros. No way. They are not insanely bad but just so many other choices.
There are a lot of great speakers in that $$ range.

...so what speakers are these?

 

[dogmamann]

Based on what is published so far, I would not pay 5000 euros. No way. They are not insanely bad but just so many other choices.
There are a lot of great speakers in that $$ range.

...so what speakers are these?

Quadral aurum Vulkan VIII R Final Edition. I looked at several passive speakers at 5000 range, and none had the low end extension these had, with this low distortion.

 

[xaviescacs]

A speaker with a bad on-axis response can have a perfect directivity? Does that question implies that I don't quite understand the relationship between both? I tend to think that a bad directivity can't be fixed with EQ whereas one can mildly fix a bad on-axis response with EQ, just like amir does in his speaker reviews. I'm sure this changes would also affect the off-axis response, but I'm not sure if the diference in tonality between on-axis and off-axis would stay the same, improve (¿?) or worsen if one uses EQ based on the on-axis response only.