My understanding of the open baffle design trade offs

[maxime.levesque]

I tried to understand the merits of open baffle design (OB), and I would like to verify if I got it right or if I missed anything.

As far as I understand, there are two advantages to OB:

1. The elimination of "box" sound

In a boxed design, lining the box surface with damping material will only help to a certain extent, some "ugly box sound" will come out through the driver cones given the thinness of driver membranes.

2. The drivers get less back pressure resistance (and hence less distortion from back pressure resistance)

Drivers in an OB can move more freely, much more so than in a sealed enclosure. The difference is lesser when compared to ported enclosures, and the wider the port, the lesser the pressure resistance.

Question: How much distortion does back pressure resistance cause ? Can it (and has it been) measured ?

The advantages of OB come at the cost of having the front sound waved cancelled by the sound comming from the back.

Strategies for cancelling the cancelation effect are based on the following realities:

i) The cancelation penalty is worst at the bottom of the sound spectrum, and is lesser as you go up towards the top.

ii) the isolation distance required between the back and front is the longest at the bottom of the spectrum, and shortens as you go up. Around mid range, a resonably small wall (inches, not feet) is sufficient to cancel the canceling effect.

For the upper "tweeter" range, the design is most often not even OB, since most tweeters are sealed in the back.

In the mid range, the cancelation effect is solved with very reasonably sized walls, measured in inches, not feet.

The main challenge of an OB design is to solve cancelation where it hurts the most: in the bass.

# The Hybrid approach

With this solution, you give up the advantage (along with the inconvenients) of OB for the lower end, and use a sealed or ported enclosure, but still enjoy OB's benefits in the mid and high range.

# The "brute force" approach

By using more powerful woofers, and/or using a greater number of them, you eventualy get strong enough SPL in the lower range to overcome the cancellation effect.

# Distributing power proportionaly according to where it's needed along the spectrum with a DSP

A DSP can be used to adjust the response curve to augment power according to the needs along the spectrum.

# Tweaking the cancelation path

The floor solves half of the cancelation problem given that sound can't travel through it, this is why most OB designs place the woofer close to the ground. Other tweaks can further cancel cancelation, for ex: by tilting the woffer upwards you get more "cancelation cancelation" from the ground.

# Mysterious Tweaks

I am curious of the theory behind design decisions that I don't undersand, there seems to be many of them, for example the placement of wooders on V shaped plates in the Linkwitz LX521.

I would love to hear about the theory that I may have missed or goten wrong.

And mostly, I would love to hear an OB with my own ears, but I might just have to build an OB to find out, given that they have not had a commercial success and hence not many living rooms have them. Is there a there a do it yourselfers in my area (Sherbrooke Québec, Canada) that built one and is willing to invite me for a listening session ?

 

[MrPeabody]

The advantages you mention are potential, hypothetical advantages, but not necessarily real advantages. As for "box sound", both internal standing waves and baffle edge diffraction will potentially color the frequency response, however there are a great many sealed and ported speakers where neither of these effects is apparent, especially the internal standing wave effect. The other advantage you mention, avoidance of distortion caused by the high pressure on the back side of the diaphragm, is even more a potential effect that likely doesn't materialize. Very many sealed and ported speakers exhibit very low distortion. The effect you mention isn't included in any of the analyses I've read of caused of distortion, and while I can't say for certain that there isn't any empirical evidence that open baffle speakers exhibit less distortion, I'm not aware of that evidence.

The reason that some people prefer open baffle speakers is most likely that they like the effect of a lot of reverberant sound coming from the reflective surfaces behind the speaker. Another related effect that may contribute to the preference is that the reflections from the side walls near the speaker can be weaker than with a sealed or ported speaker, owing to the dipole radiation pattern, but depending also on the directivity of the sealed or ported speaker. All it all it probably comes down to the reverberant sound. The open baffle speaker shares this effect with other dipole radiators, however the open baffle speaker constructed of conventional drivers with voice coils has an advantage not shared with the panel type dipole radiators. This advantage is that by virtue of the high-frequency radiators being small, the directivity is more constant than it is with most any large panel speaker. What constant directivity means is that the tonal balance does not change as you move the listening position to the side of the speaker and to the rear of the speaker. With large panel speakers, the loss of treble as you move off-axis is very noticeable.

One of the risks of making a big investment in a dipole speaker is that the reverberant effect that appeals to you at first might become distracting after a period of time. One way you can find out whether you will like this effect over the long haul is to rig up a simple arrangement where you have a second speaker adjacent to each of the main speakers but facing to the rear. You'll want to wire them in parallel so depending on the impedance, you may have to watch the volume in order not to overdrive your amplifier. You may also need to use an L-pad to get the levels matched (one L-pad for the louder of the two speakers on the right channel, another for the left channel). If you find yourself preferring to use the L-pad to set the volume of the rear-facing speaker lower than the volume of the front-facing speaker, then you have your answer. In this case if you like the effect but only with the volume of the rear-facing speaker lower than the volume of the front-facing speaker, you might prefer something like the bipolar speakers made by Definitive Technology, which you allow you set the volume of the rear-facing speaker to suit your preference. I did some experimenting of this nature back around thirty years ago, and what I discovered is that after several weeks I had grown tired of reverberant sound effect. I did not like anything that sounded like echo or reverb, or in general, anything that created a time smearing effect. I concluded that the main reason that I like the sound of good headphones is the absence of time-smearing effect. This is very much an individual thing. Some like it, some don't. If you have either a strong like or strong dislike for the sound of headphones, this may be a clue as to whether, over the long haul, you will like the reverberant sound field effect.

 

[Jim Matthews]

Here's an early foray into the design and its limitations:
https://www.tweekgeek.com/blog-tweek-geek-speek/the-open-source-open-baffle-speaker-project/

Practically speaking, for an open baffle design to produce bass, the baffle must be very wide. My experience with the Bastanis wideband driver in an open baffle was mostly positive - it had no crossover attached to the main driver and I suspect this aspect is what most of us like best about these - they're dead simple.

The late, great Siegfried Linkwitz employed open baffle designs to minimize room effects where placed near the walls. His designs used complex, active crossover networks.

My favorite open baffle designs do not attempt to cover full range playback and employ standard bass enclosures below 200 Hz. A "helper" tweeter can be added with a simple inline cap for more HF energy.

 

[maxime.levesque]

Ok, If I understand, the benefits of open baffle have more to do with taste than exactness of reproduction, like so many things in sound reproduction...

I guess it's these things aspects that make it more of an art, and less of a science !

That being said I have can very much hear "boxyness" in low quality speakers, and wether or not there is less of it in OB than boxed enclusures, I dislike it very much, especially in accoustic and ac capela music.

 

[Jim Matthews]

If you have access to the Magneplanar loudspeaker, you might try them yourself. While they have many limitations (they are prone to compression) the Magnepan is free from any box coloration.

 

[DVDdoug]

There a thousand ways to make a good speaker and a million ways to make a bad speaker.

Generally (perhaps over-generalizing) I'd say open baffle is the second worst, after no box/baffle.

If you have an old speaker to "play with" it's pretty easy to to try it with the back off, and then try mounting the drivers in a panel. And if you have a pair you can A/B.

I'd expect an "infinite baffle" (very large closed box) to perform better than an open baffle, but some woofers are not designed for this application and they'll work better in a sealed or ported box.

There are lots of design choices and compromises and manufacturer's will tout their particular choices as the "best". That also goes for magnet & cone materials, etc. Everything is touted as a feature or benefit. You should judge a speaker by it's performance, not by the design choices. Woofer size and cabinet size can be a clue about the bass (you can't get realistic bass from a 4-inch "woofer") but just because a speaker is bigger doesn't mean it has better/deeper bass.

I've never heard of any open baffle studio monitors. Almost all (good) speakers & monitors are ported or sealed and that's a clue to what usually works best. Most small monitors seem to be ported but I'm not sure about bigger studio monitors, there seems to be a mix.

Generally, I "feel" a ported speaker is a better overall compromise and my home-built subwoofers are ported. But if size (or efficiency) is not a factor, a sealed speaker can usually go lower (and bass roll-off is smoother and not as steep).

There is a free speaker design tool called WinISD* that you can use to model various cabinet designs and get a frequency response graph after entering (or selecting) the Thiele-Small parameters for a particular driver (it also comes with a database). It doesn't have an open baffle option but you can make a large infinite baffle. There probably is software for open baffle but you'd have to search yourself.

I ASSUME people who like an open baffle design like the reflected sounds coming from the back of the speaker and the software won't simulate that and it certainly can't tell you if you'll like that effect in your particular room.



* Malwarebytes gave me a warning when I downloaded it, but I've used it before and I'm 99% sure it's safe. After downloading & installing it was not flagged by Malwarebytes or Microsoft Defender.

 

[puppet]

A misconception regarding "box sound" emanating through the drivers cone. I don't believe there is any data to support the notion that sound produced by the driver inside the box passes through the drivers cone. The enclosure itself, yes. I believe this is what some find objectionable. Additionally, the enclosures environment adds support to the drivers cone. Something that isn't mentioned much. Outside an enclosure the driver is left to run "wild" and I'd expect distortion products to be unreasonably high in most cases.

As was pointed out by MrPeabody .... reverberation is a byproduct of an OB design. Realistically, they need a med/large treated room to work well. A room that will allow them to be moved out away from boundaries .. like 1-2 meters. Lacking that, the eventual "smearing" that will occur with elevated SPLs will drive you nuts. At medium SPLs OB are nice enough.

GL on your journey.

 

[maxime.levesque]

@puppet :regarding the internal encosure sound comming back through the driver membrane, I'm surprized that there hasn't been any experiments to measure it.

It seems that it could be done by simultaneously recording two equivalent drivers playing the same sound,
one in a box, and the other one mounted on a wall with an anechoic chamber on the other side, and then a comparison of the captured sound.

I wonder if it hasn't been done because there is strong enough evevidence that the phenomenon doesn't introduce distortion...

 

[Plcamp]

I have both dual 15” open baffle with 8” fullrange and tower ported speakers. Both had original list price about $5k per pair.

Both benefit quite a bit from being at least a metre from the walls. That’s no differentiator IMO.

The OB’s can reasonably reach to about 50 hz. I supplement them with a single ported 12”sub from there down. It’s very thrilling.

The OB’s image is different…more of a bubble and less exactly discerned, as one might expect when comparing an 8” fullrange to a 1” dome.

I would not say that overall I prefer one over the other speaker system.