OB in room response

[valleyman]

Hi

I recently had a read of MJKs OB articles and got hold of his worksheets as a recent £7 OB "tester" project of mine came out remarkably well and has made me want some proper ones.

Having got hold of the worksheets though, I have some concerns.

MJKs Alpha 15 and FE103 OB sims the following nice looking response:



But put a wall 1m behind it and you get this:




I sit at a higher listening position than this generally, so I decided to sim a taller OB using the FE87 instead as I felt it may work better. I also prefer a narrower baffle and while I realise bigger is better with OBs I decided to give it a go. I increased the height to 48in and reduced the width to 16in and came up with this sim:



But this time when you put a wall 1m behind it gets even worse:



This causes me to ask a couple of questions...

Firstly, is it really that bad? Obviously in room there will be a lot more room interaction going on than just a rear wall. Does it seem reasonable that all these other interactions might help smooth things. It would certainly appear so, as since my £7 OBs that were thrown together based on the size of offcuts I found in my friends garden don't seem to have any 15dB trough that I can hear, and while my ears are far from golden, I imagine I would notice a 15dB trough

Second, the sim seems to indicate that just as deep bass can be obtained from the 16in baffle as from the 20in baffle with the loss being in efficiency rather than bass extension. The problems with the rear wall added, however, are more severe. Could this mean that narrow baffles will also suffer more problems with placement that wider ones.

Finally on a more general note, would I be right in thinking OBs are more sensitive to placement than most enclosures?

Many thanks

[LineArray]

Hi valleyman,

seems you are seriously infected with OB virus !
I feel sorry for you, since experience has shown
there is no way back ...

I'd like to try some answers from personal experience,
not from a theoretical point of view. So please add an
"IMO" as a prefix to every statement made ...

The effects caused by a rear wall look evil, but will
in practice have no such diminishing effect on sound
quality as one may suppose.

As you said before, there is more than one wall involved.
There are room modes. You can rotate the OB, so that
the main radiation axis does not point to the rear wall at an
angle of 90 degrees. You can use one or the other null to
reduce excitation of a prominent room mode.

All in all, if there is space enough, you will find a position
where bass response is well balanced subjectively.

Distances to the rear wall <1m are to be avoided,
since the early reflections cannot be compensated well
perceptively.

A narrow baffle, around 2xdiameter of the driver is
to be preferred to avoid lobing in the horizontal plane.

For good low bass response, a dedicated dipole sub should
be used which is not fixed to the same position and orientation
as the "fullrange" baffle. The position near the rear wall is not
very suitable for low bass extension and efficiency ...

A separated <fill in your prefered> frame woofer, placed
according to your specific room, with its radiation axis
deviating from your fullrange baffle will give better
performance and leave you free to optimize room excitation
in the low bass range and the "rest" indepentently.
e.g. you can use a room edge (not corner) to improve efficiency
of your sub ... which can give 6db sound pressure for free
at LF.

Is the sonic result of an OB more room dependent than
that of a conventional design ?

No, not at all. There may be harder limits in placement which
will cause an OB to be dysfunctional at all
( e.g. when placed 10cm away from a rear wall ...).
When limits are obeyed, the dipole's higher directivity
will produce a result more independent from the living room
and its arrangement than a conventional monopolar design.

Kind Regards

[valleyman]

Thanks. Unfortunately, I'm a student, move around a lot and don't always have lots of space.
This really means that most of the time I don't have the luxury of ideal placement and sometimes have to make do with terrible placement. It's unfortunate but unavoidable.
It also makes the 16" width far preferable to the 20" width and from the sims the tradeoff seems to be efficiency rather than extension, which actually turns out to be a good thing as it then mates better to the FE87.
Another tradeoff as you kindly inform will be horizontal lobing but there is no way I can accomodate a 30" wide baffle.

My last question was really more asking that since I will often have the speakers badly placed, is that likely to cause more trouble for an OB than for most other box type speakers

Edit: sorry, I think your last paragraph actually answered that. Thanks

[LineArray]

It is not forbidden to combine a fullrange OB with
a monopolar sub ...

An OB design covered with a resistive wadding on the rear
side can be turned into a cardioid radiator ...

[valleyman]

I have never liked subs for music listening, and I am definitely a stereo bass fan too. Add to that the fact that its 1 more large heavy object to have in my room and it doesn't really present an attractive solution.

Cardioid radiator sounds interesting, I'll do some reading.

Has anyone investigated placing a diffuser behind the driver? I imagine that could have interesting effects

[valleyman]

I have never liked subs for music listening, and I am definitely a stereo bass fan too. Add to that the fact that its 1 more large heavy object to have in my room and it doesn't really present an attractive solution.

Cardioid radiator sounds interesting, I'll do some reading.

Has anyone investigated placing a diffuser behind the driver? I imagine that could have interesting effects

[LineArray]

For a try you could close the back of your current OB
along the wedge shaped wings with a perforated board
or a number of bars and stuff the semi enclosed volume ...

[MJK]

I have run a few tests and have worked on some newer versions of the worksheets and have concluded that the currently calculated nulls are much deeper then reality. They are deep nulls because everything is lining up perfectly, which is never the case. The room will interact with the speaker and screw up that nice flat design we all struggle to achieve but in this case I think the real results will be better then the calculated results.

First, the geometry being analyzed is with the OB parallel to the rear wall and the response calculated on axis. The summed response sees the direct sound from the front of the baffle and a reflected sound from the back of a mirrored OB speaker behind the rear wall twice the distance away. These are both on the axis of the driver, by default, so they are the maximum amplitude SPL’s and will cause the maximum amount of cancellation. That is what is presented in the worksheet.

In reality, people place the OB speaker some distance from the rear wall rotated an amount to point towards the listening position. The direct sound coming from the front of the OB is close to being on the driver axis. The reflected sound from the back of the mirrored OB speaker behind the rear wall will be well off axis and may approach the null that exists in the plane of the baffle. Obviously this would lead to less cancellations and smaller dips as apposed to deep nulls.

I have my pair of fairly large OB speakers about 1 m from the rear wall and pointed towards the listening position. I am sure that the placement is not optimum because the room is not that large. They are at the best position that could be used in the room under the constraints of the other things in the room. I do not hear a ragged response consistent with the worst-case plots you have shown. I am sure that the resulting SPL response is not perfectly flat, and could be better in a larger room, but it still sounds really good to me and other people who have heard the set-up.

[valleyman]

Thanks, a great answer that seems to reflect reality and a decent explanation of why!

[REC1]

Thanks, a great answer that seems to reflect reality and a decent explanation of why!


That is what the world needs more of.

ron