Just how severe are the effects of baffle step?

[The Paulinator]

I am about to make some sattelites using a pair of Tang Band 871's in each speaker, and I though about crossing over one of them so it only covers the lower frequencies, but I can't help but think about the center channel I made using 2 other TB drivers (the neodymium 3" ones) and how excellent it sounded having the drivers only running in series with no crossover. I mean, are the effect really obvious enough to warrant using a crossover at all? Because I sure havn't heard them.

Plus all the speakers, including the center channel, will have the drivers arranged vertically, so horizontal dispersion will not be a problem either. Is it really worth the trouble? Or is it another "I think I might hear a slight difference" thing?

[planet10]

Originally posted by The Paulinator
Is it really worth the trouble?

Most of the time, Yes.

dave

[The Paulinator]

But how audible are the effects? Or is it one of those things that you can only hear with an RTA? Isn't that the only difference, a slight change in frequency response?

[Volenti]

baffle step shows up quite clearly on an rta, but you will hear it as well, it tends to make the midrange sound "thin".

This is one of the reasons that dipole speakers sound so good, the large baffle lowers the baffle step to a point that the room boost (depending on speaker position) starts to compensate for the loss in output below the baffle step frequency.

[Zozo]

It depends on the baffle size. The larger the baffle the audible the effect is. In my opinion wo compensation the speaker starts to shout. In theory it causes a 3-6dB/octav rise in the repsonse, but the compensation room and position dependent.
Some software like Loudspeaker Lab can calculate the necessary coil and resistor from the woofer and box properties. Try and see ...I mean hear !

Zozo

[SY]

Even the most hard-core so-called objectivists will tell you that swings of 3-6 dB in midrange frequency response are clearly audible. These absolutely swamp any of the more subtle (or sometimes nonexistent) effects that audiophiles agonze over.

The simple 6dB response step nearly never occurs- it's generally more complex than that, and a function of baffle geometry and driver position, as well as size. But once you measure (or do a good model) to see what the baffle effect is for your particular geometry, it's pretty easy to compensate for.

[Timn8ter]

I used BDSC networks in my Tangband speakers and it does make a huge difference. I also noticed a difference wiring the resistor and inductor parallel rather than series. When it came to the surrounds however, I decided to skip it. I think with the configuration of drivers, the type of drivers and the bottom firing port I managed to minimize the effect.
The center driver is a W3-879 which doesn't have the top end the two W3-871s do.

[pooge]

Originally posted by SY
The simple 6dB response step nearly never occurs- it's generally more complex than that, and a function of baffle geometry and driver position, as well as size. But once you measure (or do a good model) to see what the baffle effect is for your particular geometry, it's pretty easy to compensate for.

It is also a function of driver placement with respect to room boundaries, i.e., the "Allison effect", and/or relative postion to other drivers operating in the same range. As the baffle step is simply a change in radiation impedance due to the loss of the baffle boundary (with corresponding loss of speaker power output), which is wavelength dependent, other room boundaries next to the speaker may come in to play at appropriate wavelengths to cancel out the baffle loss. Similarly, two woofers, for example, can reinforce one another to increase power output more than the two standing alone. Again, this is wavelength dependent, and dependent upon distance between the drivers. If all of these effects are taken into account, they can be played off with one another for smooth power output into the room. Roy Allison's patent is good reading on this situation.

[SY]

pooge, that's exactly right. And that's exactly why so many smart designers put the woofer near the floor- the superb NHT 3.3, for example. Allison has written quite a few clear and pertinent papers on the subject.

[Salas]

Allison wanted to avoid the response through caused by first cancellation due to path length from a high positioned woofer (normally on a stand mount cabinet) and floor. There is destructive interference between axis propagation and reflected off the floor midbass wavelengths. With common setups is occurs between 150-200 Hz. You get the thin midbass associated with good floorstanding speakers.
The baffle step is broader, roughly from 200Hz to 1kHz.