[JohnR]

When building OB speaker I find that Edge simulation of OB is not very accurate.

I find it quite accurate, you need to take into account the limitations of the model. With an H-baffle, or woofers close to the floor, Edge will not account for these effects.

[Ant_222]

I find it quite accurate, you need to take into account the limitations of the model. With an H-baffle, or woofers close to the floor, Edge will not account for these effects.

If the speaker's axis is parallel to the floor, the floor reflection is easy to take into account in EDGE: just add a phantom source with its phantom baffle at a distance from the real one equal to twice the distance from driver to floor — essentially simulating a mirror-like reflection.

Anton

[JohnR]

I haven't tried that - I thought Edge only modeled one source... oops... But at any rate you are still dealing with and have to account for the limitations of the model i.e. at some distance/frequency the phantom driver is not behaving the same.

[Ant_222]

But at any rate you are still dealing with and have to account for the limitations of the model i.e. at some distance/frequency the phantom driver is not behaving the same.

Yes, of course. But in this case the only problem could be with very highs, while in the bass region, which is most interesting for us, the modeling of the floor as a phantom source gives perfect fidelity, because you must work really hard to make your floor absorb low-frequency radiation.

Anton

[JohnR]

Not just absorb, but transmit. Different materials/constructions will vary a lot, I'd be interested to see any studies on this as it is a topic of interest to many I'm sure!

[Ant_222]

Not just absorb, but transmit. Different materials/constructions will vary a lot, I'd be interested to see any studies on this as it is a topic of interest to many I'm sure!

When it comes to walls and ceiling, reflections are usually more important than resonances (if it's them you refer to by "transmitting"). Additionally, resonances are easier to control.

Low-frequency sound radiation may be absorbed by Bekeshi panels — heavily damped large-area oscillators (after George Bekesy, I think) often used in listening rooms.

As for resonances in loudspeaker enclosures:

James Moir. Structural Resonances in Loudspeaker Cabinets. Journal of the British Sound Recording Association. 1961, vol. 6, pp. 183-187

James K. Iverson. The Theory of Loudspeaker Cabinet Resonances. JAES, April 1973, Vol. 21, #3, pp. 177-180.

(thanks to Anatoliy Linknitsky aka AML for the links)

Anton

[JohnR]

We were talking about reflections off the floor and the effect on accuracy of Edge simulations.

[Ant_222]

We were talking about reflections off the floor and the effect on accuracy of Edge simulations.

What did you mean by "transmitting" then? A panel can transmit sound to the extent to which it will move/oscillate (resonate). If it is absolutely rigid and fixed, it will not transmit anything.

Anton

[JohnR]

Hm, well, movement doesn't necessarily mean resonance. But anyway, some constructions will transmit more than others e.g. wooden floors vs concrete. My understanding is that constructions such as wood/gyprock transmit more as frequencies drop - somewhat of the opposite effect desired for modeling a woofer near the floor with a phantom image. This is just speculation though - it would be interesting to read specific work done on this.

[gooberdude]

I don't mean to take the discussion down a difft route, but i recently had an interesting OB experience with absorption panels & baffle edge defraction.

An acoustics guru was by my new place & began talking about baffle edge defraction, and how its difficult to get it 100% correct.  What he did was take down (2) gik 242 panels from my walls, and simply laid them up against the far side of each baffle...essentially doubling the width of my baffles.

The change this brought was ridiculous.  Deeper & taughter bass was the 1st thing I noticed.  Rear Waves are being absorbed & interfering less with the front waves (i think).

Its a cool trick to experiment with.  At some point i'll build new baffles & incorporate an absorption panel into the baffle itself.   My new place has many hard reflective surfaces, so YMMV of course.


matt

[Tyson]

I hear that wide baffles interfere with imaging, but that is generally with wood based baffles, which are reflective.  It might be interesting to see what effect wide baffles that are absorptive, seems like it might be the best of both worlds.

[mcgsxr]

Interesting point.

I know that some have covered their baffle surfaces with cork to alleviate to some extent that reflectivity.

Wider baffles equal lower tigher bass for sure in my experience, but tougher to place in the room sometimes too etc.

[gooberdude]

Guys,

the baffle width only looks double wide.  Its a gik absorption panel, with really no reflective surfaces.  Sound travels through the panel...

Just making a wide baffle is a bad idea IMO.   

Srsly, if anyone has 2 acoustic panels at home, pull them down from the wall and snugly set them next to your baffles...try not to have any gaps.  My panels are 4" thick, so of course they stick out from the baffles 4".  this isn't optimal obviously, but works great for experimentation.

Another location for the panels is to place them on the inside edge of your baffles, rather than on the outside edge.  I'd bet this tweak is highly room dependant. 

matt

[Tyson]

How about on BOTH edges

[gooberdude]

i've been thinking about the same thing.  i only have 2 panels to mess with unfortunately.

a diy panel built into the ob baffle would be 1/2 the size of these 2' x 4' panels, or smaller.

[moij]

I made simulation of my test U frame on ABC-dipole and get nearly same results as in Edge. ABC dipole calculated circular baffle diameter equivalent to my U-frame is 1400 mm. Seems Xlbaffle.xls had some error and my measurement had room resonance on 250 Hz what I think was first null.

[moij]

If the speaker's axis is parallel to the floor, the floor reflection is easy to take into account in EDGE: just add a phantom source with its phantom baffle at a distance from the real one equal to twice the distance from driver to floor — essentially simulating a mirror-like reflection.

Can you explane this little more?
What must be phantome baffle size? Actually I can make only one baffle in Egde - this mean: what is size I must add to original baffle?

[Rudolf]

What must be phantome baffle size? Actually I can make only one baffle in Egde - this mean: what is size I must add to original baffle?

Mirror the baffle (including driver) at the floor line - doubling its height. Hopefully the picture helps to explain it: