If you use a bypass cap in the attenuation portion. This will tend to equal things out. The lower of the tweeter frequencies are attenuated the full amount and the bypass cap allows (it's values worth) the upper frequencies to pass through the attenuation circuit.

If you do not have a bypass cap in the attenuation circuit. then the whole scale will get pushed down.

When I use an active crossover, and the tweeter need say 12db attenuation. I still use a compensation circuit. It cuts a lot of the hiss from the HF. So for 12 db total, I have a 6db compensation circuit and use the active crossover for the other 6db. This probably isnt as "pure" as using the active xo for the whole 12db as it is designed, but right now this is giving me a good sound.
This is some ball park info and Wayne can give you better details.

Bill

But there are two things to consider. One is "EQ on purpose" and the other is "EQ on accident." The part I'm talking about when I say "EQ on purpose" is what Bill alluded to; When you put a bypass capacitor across the attenuator, it serves to remove attenuation as frequency goes up. That then also augments the highest frequencies by contrast, and that's something you want to do on pretty much every compression horn tweeter. Some need it more than others, and some don't need it much at all. But by and large, all of our HF horns could use some boost in the top octave.

The other thing that happens is "on accident." Since the tweeter is a reactive device, a resistive attenuator changes the response. It changes the damping of the circuit, and can cause the tweeter to develop a peak in response. That's the reason that the R1/R2 ratio was carefull chosen, to set the damping of the system.

The best way to understand this deal is with pictures. See pages 27-43 of the π Speaker Crossover document, which shows these kinds of circuits and includes response charts to illustrate their behavior.

bill