Tweeter Mounting Offset?

[Bob Scott]

Hi Guys.

Could someone please tell me why popular speaker cabinets have the tweeter offset to one side of the front baffle?

This may be elementary, but I have not seen the reason mentioned in this forum; neither in John Krutke's website. To me, the tweeter in the middle looks nice and symmetrical and sounds fine.

Thanks,
Bob

[billfitzmaurice]

So that diffraction effects (when the wavefront wraps around the baffle) will occur at different frequencies to either side of the cab, minimizing their impact.

[oldloder]

Is it beneficial to offset the tweeter in any case or does it's effect need to be considered in the crossover design? I'm about to construct baffles for a small TM and was wondering if it mattered whether I did so or not - there was no mention of tweeter placement by the crossover designer.

[dlneubec]

The benefit of offset drivers is much debated. It improves on axis response, but as soon as you move off axis the benefits dissappear and in fact, a centered driver often has better off axis response.

The other issue to consider is that an off center tweeter creates an asymmetrical or unbalanced soundstage around the speaker, due to one side of the speaker having more low frequency energy as you move off axis than the other side (as longer waves wrap around the cabinet sooner on the short side).

Since so much of the energy we hear in the room is off axis, and many times we listen off axis anyway, I've been personally favoring centered drivers of late.

[PunkSweeper]

The benefit of offset drivers is much debated. It improves on axis response, but as soon as you move off axis the benefits dissappear and in fact, a centered driver often has better off axis response.

The other issue to consider is that an off center tweeter creates an asymmetrical or unbalanced soundstage around the speaker, due to one side of the speaker having more low frequency energy as you move off axis than the other side (as longer waves wrap around the cabinet sooner on the short side).

Since so much of the energy we hear in the room is off axis, and many times we listen off axis anyway, I've been personally favoring centered drivers of late.

Than why do just about all studio monitors have centered tweeters? Is there an alternative.


For those of you detecting a pattern, I'm designing a pair of budget near fielders, and I really don't like how off-center tweeters look ;)

[dlneubec]

Than why do just about all studio monitors have centered tweeters? Is there an alternative.


For those of you detecting a pattern, I'm designing a pair of budget near fielders, and I really don't like how off-center tweeters look ;)

Perhaps because they favor a balanced rather than an asymmetrical soundfield to the left and right of each speaker?

For a centered tweeter it is even more important to use a large roundover or chamfer and vary the distances to the center of the tweeter from the top and bottom of the baffle (the golden mean is a good place to start) so that you don't concentrate the diffraction at the same frequency as the baffle sides. I have found that a tapered box/baffle will somewhat reduce diffraction from a centered tweter as well. You can model all this in BDS or The Edge quite easily.

[romanbednarek]

Bill hit the nail on the head with his answer.

I tried an offset tweeter configuration in one of my designs (the Extremis) and I have off axis response plots on the project page for that design that show how the diffraction effects change at different horizontal listening axes. In theory, an offset tweeter is intended for speakers that are toed inward so that the tweeter is pointed directly at you. Speakers that have the tweeter centered on the baffle are, in theory, less optimal when aimed at the listener because the diffraction effects along the horizontal axis of the baffle are doubled (due to equal distance between the center of the tweeter to each side of the baffle)... but if the speakers face forward, you are listening slightly off axis from the tweeter which actually disperses the diffraction effects slightly since your ears are closer to one edge of the cabinet than the other.

Regardless of this theory, I prefer to have the tweeter centered on the baffle. I also often toe the speakers inward towards me but not fully so that the tweeters are still pointed slightly to the outside of my ears (I usually line my ears up to the inside edge of each cabinet so that when looking at the speakers from my listening position my line of site follows that edge from the front the the rear of the cabinet). With this speaker configuration I am still slightly off axis from the tweeter which helps disperse the diffraction effects slightly but I am not off axis so much that I lose the top end response of the tweeter which tends to beam at higher frequencies especially for larger diameter tweeters.

So that's my summary but for some extra, somewhat related, info you can check out one of the "Audiofiles" that I put up on my site at this link...

http://www.rjbaudio.com/Audiofiles/tweeter%20diffraction.html

[PunkSweeper]

Hmm, interesting. If they don't have an alternate wave negation solution, all those High-Dollar studio monitors are operating compromised.

[billfitzmaurice]

Bill hit the nail on the head with his answer.

I tried an offset tweeter configuration in one of my designs (the Extremis) and I have off axis response plots on the project page for that design that show how the diffraction effects change at different horizontal listening axes. Sound waves don't care for symmetry, either in the speakers that they emanate from or in the rooms that they waft around in. More often than not asymmetry in both the speakers and the room just works better.

[tedwilt]

Centered tweeter speakers are cheaper to construct and distribute because only need one product instead of two.

[romanbednarek]

Centered tweeter speakers are cheaper to construct and distribute because only need one product instead of two.

I have a funny story that I have yet to figure out related to your comment. Back when the Dynaco A25 speakers came out my uncle bought two pairs at the same time, one pair for him and one pair for my dad. I ended up getting both pairs in the late 80's. I was able to distinguish each pair by the binding posts because one pair had binding posts that had very narrow grooves in them and the other pair had binding posts were smoother with wide contoured grooves. The odd thing was that the one pair had both tweeters offset to the left while the second pair had both tweeters offset to the right. I really can't explain this and I'm not familiar enough with that speaker, despite its popularity, to know what the original intended pair was supposed to be (symmetrically mirrored or both tweeters offset to the same side).

http://home.indy.net/~gregdunn/dynaco/components/speakers/myA25.jpg

[Face]

Correct tweeter placement. ;)
http://img143.imageshack.us/img143/9880/tannoysystem600gr.jpg (http://img143.imageshack.us/i/tannoysystem600gr.jpg/)

[dlr]

I have a funny story that I have yet to figure out related to your comment. Back when the Dynaco A25 speakers came out my uncle bought two pairs at the same time, one pair for him and one pair for my dad. I ended up getting both pairs in the late 80's. I was able to distinguish each pair by the binding posts because one pair had binding posts that had very narrow grooves in them and the other pair had binding posts were smoother with wide contoured grooves. The odd thing was that the one pair had both tweeters offset to the left while the second pair had both tweeters offset to the right. I really can't explain this and I'm not familiar enough with that speaker, despite its popularity, to know what the original intended pair was supposed to be (symmetrically mirrored or both tweeters offset to the same side).
Surprisingly to me, I found that if both tweeters are offset in the same direction, it's audible if your listening distance is a bit on the short side and set up for better soundstage (I like equilateral triangle). I re-used an old set of EPI boxes and kept thinking that I'd hooked up something wrong in a crossover. Measurements verified they were correct. I never thought I'd notice it.

I have to disagree with some of the points about diffraction and the off-axis. For one, the variation in diffraction to the near side is less than the far side and the overall diffraction delta (integrated) of that short side off-axis is less than the the delta for a centered driver. There may be specific angles at which one or the other may be better than the other, but move another few degrees and the results may switch. Choose a third offset and the third one may be better than both previous ones at any specific angle. Or worse. It's all case specific on that account, but the integrated off-axis in all likelihood is better for an offset on the short side.

There's another benefit, small but it can help, for those who have a bit less experience with design. Too many still work to get flat on-axis. With an offset tweeter, the on-axis is flatter and the result of a flat response will be closer to achieving a flat power response as well.

I also have to take issue with the idea that an asymmetric response due to small tweeter offsets will have an audible effect on perceived room response due to asymmetry in 4-pi. Given that every room has such a profound influence and that the tweeter off-axis is very small vis-a-vis the power response in 4-pi since they become direction very quickly I don't see it. 2-pi, yeah. I just do not believe that any change in 4-pi response of the tweeter offset vs. centered is audible. It's swamped by the room effects and is too narrow in its influence.

dlr

[Jeff B.]

Surprisingly to me, I found that if both tweeters are offset in the same direction, it's audible if your listening distance is a bit on the short side and set up for better soundstage (I like equilateral triangle). I re-used an old set of EPI boxes and kept thinking that I'd hooked up something wrong in a crossover. Measurements verified they were correct. I never thought I'd notice it.

I have to disagree with some of the points about diffraction and the off-axis. For one, the variation in diffraction to the near side is less than the far side and the overall diffraction delta (integrated) of that short side off-axis is less than the the delta for a centered driver. There may be specific angles at which one or the other may be better than the other, but move another few degrees and the results may switch. Choose a third offset and the third one may be better than both previous ones at any specific angle. Or worse. It's all case specific on that account, but the integrated off-axis in all likelihood is better for an offset on the short side.

There's another benefit, small but it can help, for those who have a bit less experience with design. Too many still work to get flat on-axis. With an offset tweeter, the on-axis is flatter and the result of a flat response will be closer to achieving a flat power response as well.

I also have to take issue with the idea that an asymmetric response due to small tweeter offsets will have an audible effect on perceived room response due to asymmetry in 4-pi. Given that every room has such a profound influence and that the tweeter off-axis is very small vis-a-vis the power response in 4-pi since they become direction very quickly I don't see it. 2-pi, yeah. I just do not believe that any change in 4-pi response of the tweeter offset vs. centered is audible. It's swamped by the room effects and is too narrow in its influence.

dlr


Not to mention the effects of driver induced directivity as you move higher in frequency. Diffraction effects at higher frequencies are significantly reduced off axis if very little energy is being radiated 90 degrees off axis.

[soiset]

Two comments:

An offset tweeter in MTM designs allows closer spacing of the mid-woofers, and can allow the rule of one wavelength of the crossover frequency c-c spacing.

Also, any tweeter diffraction problems can be mitigated or eliminated by using heavy wool felt around the tweeter, placed to a distance of at least a wavelength of the crossover frequency from the center of the tweeter on the baffle. It's just not as pretty as finished wood.

So, unless building a MTM design, I'd center the tweeter and use felt to control diffraction.

Somewhat newbie here, but in all my reading, the above info seems to be very solid.

[DDF]

Too many still work to get flat on-axis. With an offset tweeter, the on-axis is flatter and the result of a flat response will be closer to achieving a flat power response as well.

I just do not believe that any change in 4-pi response of the tweeter offset vs. centered is audible. It's swamped by the room effects and is too narrow in its influence.

dlr

Don't these statements contradict each other?

I'm strongly of the opinion that flat on axis is lazy and often misinformed. Equing out a diffraction signature for a centered tweeter at one point in space directly on axis will unbalance a speakers timbre in room IME.

I feel this way since my experiences are counter to the second statement. I've tweaked too many xovers to sound right non-flat on axis with centered tweeters now to reach any other conclusion. The key is looking at the electrical (not acoustic) response of the end result: no goofy difraction eq.

IME FWIW.

Dave

[dlr]

Don't these statements contradict each other?
As written, yes. I was referring to the statement about "asymmetrical or unbalanced soundstage around the speaker", but didn't make that clear. I don't see this aspect being audible. Actual power response and timbre, yes, that would be affected and could be audible if EQ'd as you state. Sometimes I don't express my thoughts well. This was definitely one of those times.

dlr

[dlneubec]

I'm betting that the soundstage would present itself differently and in that way the speaker would sound different. That's speculation on my part, since I have not done a controlled study. Perhaps when the tweeters are offset to the inside the soundstage sounds more centered and not as wide, for example. I know I've read that the range of 2khz-6khz is where the human ear is most sensitive to locating sounds, and this would fall in that range in most cases.

You also have the question of how important is driver vertical alignment to your off axis response, since the offset tweeter becomes more and more offset from the mid/woofer, in terms of time of flight to your ear, the further you move off axis. At some point, offset becomes detrimental or we would be seeing many more speakers with drivers side by side horizontally. Just where is that point?

That would be an interesting experiment. The same speaker in every way, but with one pair having the tweeters centered and one with them offset. Does anyone know if this comparison has been done and documented? That would be more meaningful than speculation, IMO. Of course, the results would have to based on subjective auditioning, which has its own set of problems.;)

[dlr]

I'm betting that the soundstage would present itself differently and in that way the speaker would sound different. That's speculation on my part, since I have not done a controlled study. Perhaps when the tweeters are offset to the inside the soundstage sounds more centered and not as wide, for example. I know I've read that the range of 2khz-6khz is where the human ear is most sensitive to locating sounds, and this would fall in that range in most cases.
The reason I disagreed is that the soundstage is largely determined by the first arrival. If the off-axis is affecting it, that would say that the system is placed too close to something (say a wall) that is reflecting enough to muddy things. At least that's how I see it. My experience also has been that it's very easy to notice non-symmetric lateral offsets such as tweeters both on the same side than it is to hear a soundstage change from mirror-imaged tweeters. Mono and stereo with centered vocalists shift or sound off-center or unfocused very quickly for anything unbalanced in the system.


You also have the question of how important is driver vertical alignment to your off axis response, since the offset tweeter becomes more and more offset from the mid/woofer, in terms of time of flight to your ear, the further you move off axis. At some point, offset becomes detrimental or we would be seeing many more speakers with drivers side by side horizontally. Just where is that point?
Good question. I wish I knew the answer. I'm one who places less emphasis on absolutes with regard to driver offsets since we generally listen at a fair distance that allows for better driver integration at the listening position. The one thing that can't be changed in that regard is the power response since driver separation affects it.

The issue with driver lateral offset is more of an issue with a 2-way than a 3-way unless the M/T Fc is too high. The problem is that more must be known, since driver directionality, acoustic offset and baffle diffraction must also be factored in. All affect the change off-axis. They may add to the problem or in some situations help the problem to some degree.

Directionality probably complicates it for larger mids or woofers because as the distance to driver centers increases for angles on the short side, the rolloff coupled with its influence on phase may exacerbate the issue. Of course the off-axis on the large side will see improvement because that helps to reduce the relative acoustic offset at angles on that side. If diffraction is controlled, the long side may be the better side for the design axis on not on-axis. The long axis also reduces the relative acoustic offset on that side, so driver ntegration may improve, again with the caveat that the diffraction doesn't counter it.


That would be an interesting experiment. The same speaker in every way, but with one pair having the tweeters centered and one with them offset. Does anyone know if this comparison has been done and documented? That would be more meaningful than speculation, IMO. Of course, the results would have to based on subjective auditioning, which has its own set of problems.;)
The problem would be that it requires different crossovers for each one. That's a point I missed in my earlier posts that confused things. The difference in perception between two systems, one with centered tweeter, the other offset, may be small with each having similar, appropriate crossovers.

That brings in other questions, though. What is a similar crossover? Is it one optimized for a simliar on-axis response? Is it one with a similar power response? As Dave (DDF) pointed out, optimizing a centered driver for flat response affects the power response. That relates to my point earlier that too much emphasis is placed on the on-axis response for flatness that then affects the power response. An offset tweeter, though not demonstrably superior, will benefit those less experienced because they CAN go flatter on-axis and have a better chance of having a smooth power response. It's similar to using a windowed-average measurement for design purposes.

You really can't have any situation with "all else equal". That's probably part of why DIY is enjoyable, there is never a single best way to do any of it.

dlr

[maynardg]

An offset tweeter, though not demonstrably superior, will benefit those less experienced because they CAN go flatter on-axis and have a better chance of having a smooth power response. It's similar to using a windowed-average measurement for design purposes.

Which is essentially why I tend to offset. The modeled response seems easier to work with - more forgiving.

[DDF]

In the "you read it here first category", my next speaker will probably have the geometry such that the on axis response is equivalent to the "power" response in 60 degree cone, below ~ 4 or 5 khz.

I don't know why people don't just go for this out of the gate, its so obvious.

Simple tilted baffles with appropriate toe in (and perhaps bipole) can achieve this without any nutty upward facing drivers and reflection/diffraction causing junk in front.

If anything it should be a lot easier to place, the crossover will be simple etc.

[brianpowers27]

In the "you read it here first category", my next speaker will probably have the geometry such that the on axis response is equivalent to the "power" response in 60 degree cone, below ~ 4 or 5 khz.

I don't know why people don't just go for this out of the gate, its so obvious.

Simple tilted baffles with appropriate toe in (and perhaps bipole) can achieve this without any nutty upward facing drivers and reflection/diffraction causing junk in front.

If anything it should be a lot easier to place, the crossover will be simple etc.
Is there any way to achieve this below 500hz without going dipole?

[dlr]

In the "you read it here first category", my next speaker will probably have the geometry such that the on axis response is equivalent to the "power" response in 60 degree cone, below ~ 4 or 5 khz.

I don't know why people don't just go for this out of the gate, its so obvious.

Simple tilted baffles with appropriate toe in (and perhaps bipole) can achieve this without any nutty upward facing drivers and reflection/diffraction causing junk in front.

If anything it should be a lot easier to place, the crossover will be simple etc.
I've only heard a couple of bipoles, but none have had good timbre. Implementation I suppose, bit I've not had any desire to go that route from what I've heard.

I am following john k's current work with narrow baffle dipoles and how they may provide a better power response while simultaneously reducing diffraction artifacts. That's the direction that I think I may be headed soon.

dlr

[brianpowers27]

I've only heard a couple of bipoles, but none have had good timbre. Implementation I suppose, bit I've not had any desire to go that route from what I've heard.

I am following john k's current work with narrow baffle dipoles and how they may provide a better power response while simultaneously reducing diffraction artifacts. That's the direction that I think I may be headed soon.

dlr
I built mine narrow to insure that the dipole cancellation would reach into the upper frequencies. I believe mine stays almost correct until about 6khz. Above tht, the drivers directionality is not affected by dipole cancellations.

How does a narrow baffle reduce diffraction artifacts?

[dlr]

I built mine narrow to insure that the dipole cancellation would reach into the upper frequencies. I believe mine stays almost correct until about 6khz. Above tht, the drivers directionality is not affected by dipole cancellations.

How does a narrow baffle reduce diffraction artifacts?
Above the dipole peak it doesn't do much, but one of the more difficult parts to work with in terms of diffraction is the step area. My first impression on reading john's paper is that this area will be easier to handle and will be primarily in the tweeter range. This will mean using a midrange driver (no midwoofers need apply as I see it) that extends rather high and/or using a 2nd order L-R. It sounds like you may already be doing something along this line, at least as far as Fc is concerned.

I left out one important point, that being that the power response can be smoother if done right according to John. That's what cinched it for me. If you haven't read his paper, I recommend it:

Dipoles and Open Baffle Design Considerations (http://www.musicanddesign.com/Dipoles_and_open_baffles.html)

What you've mentioned relates to it. Above the dipole peak, for the same reason that the dipole cancellations drop off, so too does diffraction effects. The increasing driver directionality comes into play for both the dipole effect an diffraction simultaneously. This is also the area of prime baffle diffraction control effectiveness for felt.

My intent is to test John's ideas coupled with diffraction control for the 2pi region of a narrow baffle dipole. Well, actually, I expect that to be something of a cardioid because I also intend to damp the back side to some degree. It's mostly just to experiment and learn. My hope is that the result will be better than my current setups.

dlr

[badman]

Which is essentially why I tend to offset. The modeled response seems easier to work with - more forgiving.

yeppers. We can use all the forgiveness we can get in DIY.

[brianpowers27]

Above the dipole peak it doesn't do much, but one of the more difficult parts to work with in terms of diffraction is the step area. My first impression on reading john's paper is that this area will be easier to handle and will be primarily in the tweeter range. This will mean using a midrange driver (no midwoofers need apply as I see it) that extends rather high and/or using a 2nd order L-R. It sounds like you may already be doing something along this line, at least as far as Fc is concerned.

I left out one important point, that being that the power response can be smoother if done right according to John. That's what cinched it for me. If you haven't read his paper, I recommend it:

My intent is to test John's ideas coupled with diffraction control for the 2pi region of a narrow baffle dipole. Well, actually, I expect that to be something of a cardioid because I also intend to damp the back side to some degree. It's mostly just to experiment and learn. My hope is that the result will be better than my current setups.

dlr

You are righ, I am taking the concepts mentioned above into consideration. The driver above is an alesis buyout midwoof/midrange. THe FS is high(100hz?) and the bass extension is not great. The tweeters are ND20s spaced as close as possible both out of phase and opposing each other. It would be hard to imagine having a narrower baffle....

I have read over much of John K.'s work but need to re-read it. I don't recall the area that specifically deals with diffraction control.

Here is one hraph from the power response. I did that one a long time ago and can't remember much of the details other than to say this is a sampling of on/off axis for the tweeter. The midrange and woofer sections nearly had perfect power response.

http://i668.photobucket.com/albums/vv50/brianpowers27/OB1/tweeter.jpg

http://i668.photobucket.com/albums/vv50/brianpowers27/OB1/20090514_1316.jpg
http://i668.photobucket.com/albums/vv50/brianpowers27/OB1/n664151102_2634340_5203192.jpg

[DDF]

Is there any way to achieve this below 500hz without going bipole?

Could go side firing, or follow Dave's line with the dipole (or upward firing, or filler drivers or or ...).

I'd build the test mule with a side firing extra driver and be prepared to tune it's response with a passive xover.

It'd take simulations (and time, as i eat dinner at 10PM) to nail down.

[bring me into the light]

I'm betting that the soundstage would present itself differently and in that way the speaker would sound different. That's speculation on my part, since I have not done a controlled study. Perhaps when the tweeters are offset to the inside the soundstage sounds more centered and not as wide, for example. I know I've read that the range of 2khz-6khz is where the human ear is most sensitive to locating sounds, and this would fall in that range in most cases.


Hello,
I have just registered following stumbling on this thread with a googled question.

I have nothing profound to add but wondered if someone may answer this which would help me :)

Testing and seeing which works best aside, I have seen some speakers sold with an offset tweeter that are not marked left or right and give no indication as to it's intended positioning? i.e. tweeters on the inside or outside by reversing the pair. (quoted this post as it suggested a possibly narrower field if inner most?)

My experiences of designs come purely as a listener and marketing hype :rolleyes:

So it would appear that many manufacturers just don't offset as cost of production is higher and selling as pairs also adds to the cost.

Also on this thread seen a few comments on symetrical being used in studios so this must be correct? Well away from HiFi but M&K being used for movies are nearly always offset and even angled?

So the speakers I have that never made a mention of inner or outer placement are Ruark Solus. Does anyone know which was the INTENDED orientation for this model?

Thanks in advance

:cool:

[Æ]

Hello,
I have just registered following stumbling on this thread with a googled question.

I have nothing profound to add but wondered if someone may answer this which would help me :)

Testing and seeing which works best aside, I have seen some speakers sold with an offset tweeter that are not marked left or right and give no indication as to it's intended positioning? i.e. tweeters on the inside or outside by reversing the pair. (quoted this post as it suggested a possibly narrower field if inner most?)

My experiences of designs come purely as a listener and marketing hype :rolleyes:

So it would appear that many manufacturers just don't offset as cost of production is higher and selling as pairs also adds to the cost.

Also on this thread seen a few comments on symetrical being used in studios so this must be correct? Well away from HiFi but M&K being used for movies are nearly always offset and even angled?

So the speakers I have that never made a mention of inner or outer placement are Ruark Solus. Does anyone know which was the INTENDED orientation for this model?

Thanks in advance
:cool:

I'm only answering you question in general. With offset tweeters it is often subjective and you to make the determination yourself. Place the enclosures the way it sounds the best.
I made a pair of enclosures with offset tweeters, the tweeters offset to the insides sounded best to my ears.