dumb question: what does a baffle do?

[cuibono]

I know its a dumb question, but I can't find anything here answering what I'm wondering: pretend we have an open baffle three way speaker (xo at about 250 and 10kHz, or anywhere really..) -- for each section of the audio spectrum (bass, mid, highs), what does the baffle do sonically. I'm aware of the baffle step, so ignore that for now. And ignore structural support of the drivers, I'm looking for things audible.

Specifically, I'd like to know what happens across the audio range if you had no baffle at all.

I suppose bass would be diminished, and maybe the tweeter would beam more - but why? Is it only because of the baffle step? Suppose the crossover could handle any equalization necessary. What other trickery is the baffle doing besides helping the hand-over from full space radiation to half space?

Also, ignore some of the full range speaker's trickery, like the frugal horn series: I'm thinking about a three way system here...

[Steve Eddy]

Basically the baffle just keeps the front and rear waves emanating from the speaker from seeing each other and subsequently canceling each other.

se

[5th element]

If we consider a closed box for a start, then the box is there to prevent the front and rear wave from canceling each other out, whilst loading the driver correctly.

Remove the box and you are left with just the simple baffle. As you have mentioned baffle step you are already aware of presumably why this occurs.

In an open baffle loudspeaker, baffle step, as we know it, does not actually happen; rather the front and rear wave now combine destructively where baffle step would normally occur and this causes the 6dB roll off per octave.

Baffle step is a transition period where the width of the baffle becomes negligible in comparison to the wavelength, such that the wave starts to see the baffle as insignificant and starts to go 'around' it. The radiation goes from being hemispherical, to radiating into full space, half the acoustic power is now lost behind the enclosure, hence the 6dB loss.

If you understand why baffle step occurs, then you will also know that the narrower the baffle, the higher up it will occur, this translates directly to an open baffle too. The narrower the baffle, the higher up the 6dB per octave cancellation will begin, this is the important part, especially with open baffles.

I suppose bass would be diminished, and maybe the tweeter would beam more

Bass would be diminished yes, because the baffle width would most probably have diminished too, hence starting the 6dB roll off earlier. The reason people use W, H and U frames, is to extend the apparent width of the enclosure and hence push the roll off lower in frequency. Remember open baffles require (generally) equalisation to compensate for the roll off, the lower you can push this, the less EQ you will require and therefore less amplifier power, less cone excursion etc.

The tweeter will not beam any more then it did before. Beaming is where the off axis response deviates from the on axis response, as a result of the diameter of the cone in comparison to the frequency it is producing.

Suppose the crossover could handle any equalization necessary. What other trickery is the baffle doing besides helping the hand-over from full space radiation to half space?

The baffle isn't doing anything else beneficial, apart from maybe helping with diffraction issues.

[Svante]

The baffle reduces the amount of cancellation that occurs when the wave from the back of the speaker interferes with the wave on the front side. Typically, a larger baffle gives more bass, like in the example sim below. The details in the simulation response around 1 kHz depends on the dimensions and placement of the driver. What I am after to demonstrate is the higher level of the wider baffle at low frequencies.



Even though the effects of an open baffle is not usually called a "baffle step", it has great similarities with one. It can be modelled almost identically as the baffle step, with sources at the edge of the baffle, but with a different amplitude.

PS one could read your post as if you think that the baffle step does not have audible effects; for the record, it does.

[cuibono]

thank you everyone for your response! I really appreciated them.

Svante - thanks for the graph - I'm sorry my reply seemed ambiguous. I know baffle shape has a big sound signature. I meant that besides the baffle step, what other effects does baffle have on the sound. So looking at your graph, would you say a wider open baffle has better low end response because of lower cancelation between front and rear waves (physically blocking the waves from interacting), or because of diffraction at the edges? Or are they the same thing...

Although I've had a bit of phyics in school, I've never really understood how the frequency of a sound wave determines how it interacts with a physical barrier, particularly, as the wavelength is approximately the size or bigger than the curvature it is moving around, that they can 'not see' the barrier...

any help?

[Svante]

Originally posted by cuibono
thank you everyone for your response! I really appreciated them.

Svante - thanks for the graph - I'm sorry my reply seemed ambiguous. I know baffle shape has a big sound signature. I meant that besides the baffle step, what other effects does baffle have on the sound. So looking at your graph, would you say a wider open baffle has better low end response because of lower cancelation between front and rear waves (physically blocking the waves from interacting), or because of diffraction at the edges? Or are they the same thing...

Although I've had a bit of phyics in school, I've never really understood how the frequency of a sound wave determines how it interacts with a physical barrier, particularly, as the wavelength is approximately the size or bigger than the curvature it is moving around, that they can 'not see' the barrier...

any help?

I think there are two ways of thinking of the baffle and diffraction.

One is the way commounly tought in school, ie that low frequencies "goes round" the edges and interferes with the sound on the front side. Low frequencies being defined as when the wavelength "is larger" than the size of the baffle. Unfortunately, this viewpoint does not help much when it comes to calculating the details of the behaviour. Also, how does the wave that hits the edge at a given point know what the other parts of the wave does, at the other parts of the baffle edge? How does it know when to go around the edge?

The other way of looking at it, which resolves the above problems, is to realise that the local wave does not know anything about the other parts of the baffle. Diffraction is actually frequency independent, locally at the baffle edge.

This has lead to the GTD, geometric theory of diffraction. In this, diffraction can be modelled by adding a lot of point sources along the baffle edges. These sources are delayed corresponding to the local distance between the edge and the driver. The sources themselves are frequency independent, but when they are added up with their different phases the sum is frequency dependent. At low frequencies they are all more or less in phase, thus low frequencies "goes around the edge". High frequencies, otoh will almost sum to zero except for special geometries.

That was for the sound from the back to the front. Almost the same theory can be used to describe the baffle step. For this the sound from the front side is reflected at the edge, but with opposite sign, which instead leads to the well known loss of 6 dB at low frequencies.

You can also have a look at the tech docs of Basta! and The Edge in my signature. I wrote a bit about the modelling there.