Originally Posted by lovingdvd
Can you please elaborate on the strategy and reasoning behind setting your curve that way? Thanks.
Let me describe how I approach this with Dirac Live.
I'm trying to get a common crossover point for all speakers. This will lead to the most predictable results. So what I do first is to measure all speakers, check how low they can go and determine a common crossover point that is achievable with EQ. Ideally this would be 80Hz.
Then group all satellite speakers in DLCT, load your preferred target curve into that set and set the lowest breakpoint one octave below your desired crossover frequency. Move the "curtain" to the left (or edit your target curve file and set LOWLIMITHZ to 10), add another breakpoint below and drag it down. The resulting curve should look similar to the blue curve in the response graph above.
Do the same for your sub, i.e. set the highest breakpoint one octave above your desired crossover frequency. Move the curtain to the right (or edit HIGHLIMITHZ in your target curve file), add another breakpoint above and drag it down. The resulting curve should look similar to the red curve in the response graph above.
Click optimize for sub and satellite speaker set and load filters into the 88A.
Now set the crossover in your AVR to the common crossover frequency you've determined in the beginning. Also make sure the crossover slopes in your AVR are symmetrical and add constructively.
Often different filters with different slopes are used. This will screw up the response in the crossover region. Some bass management filters will even create a deep notch at the crossover frequency because sub and satellite add destructively (2nd order Linkwitz-Riley for example - you would need to invert your sub connection). Even when filters add correctly you might get a bump. For example standard 12dB Butterworth filter create a 3dB bump around the crossover frequency.