I just ran Dirac and am going through the different channels and notice the range on my L/R and C channels is 20hz - 20khz. Do I need to change the target and move the curtains to 100hz if I am crossing them over to my sub at 100hz?
Same applies for the sub channel, it has a range from 15hz - 200hz, do I need to adjust the curtains down to 100hz as that's my crossover point to the speakers?
Or if I simply load the default Dirac targets for all channels, is Dirac still applying EQ for my L/C/R from 20hz-100hz even though I have the crossover at 100hz?
Hope this is clear and I appreciate all responses, thanks!
The curtains control the frequency range over which Dirac will try to correct the response of a certain channel. The response "outside the curtains" will be left as it was (the Dirac filter will be flat there), and the target curve inside the curtains will be adjusted. (The frequency response cannot jump suddenly; the target curve will be gradually bent on the ends so it can join up with the uncorrected portions without any kinks or jumps.) By default, Dirac will position the curtains at the points that it thinks are the response limits of the speaker.
However, other than that, Dirac treats each channel as "full range" for purposes of correcting it. Then, when you apply the crossover LATER, the crossover will "use" as much of the corrected/full frequency range as it asks for. Even if you're going to put a crossover at 80 Hz, so you're not going to be sending anything to that channel below 80 Hz (nominally), there's no particular reason NOT to let Dirac correct that channel down to 40 Hz. (This is both because crossovers are NOT perfectly sharp, and because it really doesn't hurt anything to correct portions of the spectrum you won't be using anyway. It also lets you change the crossover settings after you run Dirac, and try different settings, or even large vs small, without having to re-run Dirac.)
The only reason to specifically adjust the curtains is that you specifically DON'T want Dirac to adjust certain frequency ranges. For example, if your speakers sound really just perfect with voice and upper frequencies, but the bass is bad, you might set the top curtain down to 250 Hz. This would allow Dirac to adjust the bass, but specifically prevent it from adjusting anything above 250 Hz (because you think that's just perfect the way it is).
Or, maybe, if your surrounds had really bad tweeters, and, when Dirac tried to make them flat by boosting the treble, it just made them sound worse. In that case, you might set the top curtain down to 12 kHz for those channels (in effect telling Dirac "forget everything above 12 kHz - you can't fix it - so don't try").
Dirac does not use the crossovers you set. It will measure and adjust each speaker according to it's capabilities and completely ignores your set crossovers. The XMC takes care of the bass management. So you can leave the curtains as full range unless you specifically do not want certain frequencies adjusted. An example would be if you don't want Dirac to adjust from 200Hz to 20,000Hz then you would drag the right curtain down to 200Hz or so. If you don't want it to bother with things below 20Hz you could drag the left curtain up to 20Hz. Do that for all speakers that might play those frequencies.
BUT, if you let Dirac correct frequencies above or below the intended range of a given channel, it will use so much of the available output of the DSP trying to boost an un-need area that the output of your preamp will be seriously limited in normal listening use. Example, my subs are crossed over @80 hz "later on" by the XNC-1. Dirac is trying to boost the output of my subs all the way to beyond 1khz. This unwanted and un-needed boost eats up the full 10db of boost above the lp crossover point and works toward limiting total output. It is also trying to add 10db of boost to my surrounds, trying to EQ them into a BOOST all the way down to 20hz. Yes, they are later crossed over @100hz, but the output damage is already done. I realize dragging the curtains around can somewhat control that, but I don't know if everybody is aware that there IS a reason to Not let Dirac do what ever it "thinks" is right