Amplitude at Crossover Frequency

[Otto]

OK, I've been looking around for info about the measured level of speakers at the crossover. I saw a graph that Wayne posted here that showed the signal level being down 3 dB at the crossover frequency. I've always thought of the crossover point as the -3 dB point as well. However, I measure my mains rolloff as 6 dB down at the crossover point. I've also read somewhere that speakers should be 6 dB down for even-order filters, and 3 dB down for odd order filters. I'm not sure if I buy that...

Anyway, discussion on the above is welcomed.

Tonight, while playing with making an LFP out of my BFD, I was playing with the rolloffs for the sub, as applied to target level. There is no house curve loaded. By playing with the crossover slope, and cutoff buttons in REW, I can get the target level to appear as either of these; 3 db down at the crossover point (40 Hz) or 6 dB down at the crossover point. The target is 75 dB, no house curve, nothing but the target as calculated by REW.

???


3dbdownat40hz.jpg

Graph 1: Target=75 dB, Crossover=40 Hz, Response Shown 3 dB Down at fc

6dbdownat40hz.jpg

Graph 2: Target=75 dB, Crossover=40 Hz, Response Shown 6 dB Down at fc

[brucek]

Not sure what your question is, but the REW targets are a standard 6dB down at the crossover frequency. As you know, if you combine two signals at the same level and frequency you'll get a 6dB gain as a resultant.

So in a perfect world, the red line (shown below in the pic) would be the resultant of a set of mains and sub when combined and given the standard crossovers. A perfect flat response...

80crossover sub plus main.jpg

brucek

[Otto]

Yeah, makes sense to have it be 6 dB down. I edited my post to include descriptions of the graphs. Note that REW has calculated two different responses for a fourth-order 40 Hz LPF.

I think it's a bug in REW. You can cause it to happen by playing with the "Cutoff" and what I'll call "slope" controls in REW. Change slope, change cutoff, change slope... Something like that. You'll see it change the x dB down point for the target.

Wonder if anyone (myself included) has taken it as the target, and later had problems integrating with mains...

[brucek]

I think it's a bug in REW. You can cause it to happen by playing with the "Cutoff" and what I'll call "slope" controls in REW

That bug has already been noted before. See here.

John has fixed it and it will be repaired in the next rev....

brucek

[Otto]

Thanks/oops. Missed that, or forgot about it.