baffle step on just midrange or whole system

[mazurek]

I am still working out my next active crossover. I was curious whether the phase effects of the baffle step correction should be applied to the whole speaker, or just the midrange. I guess the essential question is are the phase effects equalizing the midrange phase in a good way to make the phase plot more straight, or not?

For everyone's amusement here is a plot of the gain and phase effects of a typical baffle step correction circuit.

By the way, I will continue to make divergent pcbs from the main diyaudio pcb until they start including muting relays (I'm sure after the active crossover building boom, others will wish they were included).

Lee

[lndm]

I can't say I believe the phase changes from the compensation circuitry are anything but a side effect. It is the level that is the reason for the baffle step compensation. As far as that, it will effect all frequencies below a certain point, no matter what driver they are coming from. Each driver may be a little different due to individual mounting conditions but not hugely.

Imagine (only), compensating each driver individually for the baffle step first, then implementing a text book crossover. In theory this would work.

Do you have a way of measuring to be sure? If not you may find a baffle simulator a helpful (necessary) toy to play with.

[mazurek]

I finally produced good measurements from my bookshelf for the first time. I had long been having trouble getting good in room measurements, but usingn the technique of pull out the system into the middle of the room and measure from 12-24 inches away gave really good results.

The whole baffle is covered in felt to deal with no flush mounting. Baffle is parts express 0.25 cubic feet, 7.5" wide.

Here is measurement of the midwoofer with no crossover.

[mazurek]

Here is a simulation of the woofer

[mazurek]

The woofer exhibits almost perfect baffle step - -3db down from enclosure at 85, + -6db down from baffle step. Its response is slightly smoothed, but there is some noise.

Interesting comments: I used my audiophile roomate to a large part to tune with and without various options such as baffle step on the behringer deq, as I was having a lot of difficulty getting reliable measurements in my room. He liked it better without a baffle step, said it sounded more detailed. Evidently he liked midrange hump, also I think the drivers have broken in slightly since then. I myself cannot usually hear anything concrete A/B style, but have to listen over long periods of time, a nagging feeling of missing warmth led me to try to measure again. It is sort of amazing actually that both with and without bsc sound plausible, I would have expected 6db difference to be ridiculously large. Before the recent correction experiments, my roomate suggested my speakers sound like high end b&w.

[lndm]

Glad to see you are getting the results you want. Be aware that the baffle step does not show its strength up close. Recommended measuring distance is at least 2' but I notice that in anechoic chambers, it is more common to measure from 10'.

Your measured results look promising. You'll notice there is a steady rise from the lower midrange right to the upper limits of your midwoofer. If you use an RC across your driver to flatten the impedance and then apply an inductor in series with a rolloff centred in the lower midrange, it will (probably) flatten the woofers response, compensating the baffle step in one fell swoop.

PS. done right, this will give you the warmth you seek.