Ok I think I see, kind of what described subterFUSE then. But mixing nearfield and listening position.
But first, question for confirmation: bandpass = between XO points, stop band = below & Above XO points?
In fact there's a lot of similarities with what I was trying to do with Dirac recently (and APL before).
Meaning focusing PEQ+XO to get to the acoustic slope, on the stop & transitions band, and leave the FIR device focusing on the bandpass.
Well with these actual words it seems much simpler.
But I never used near-field measurements for anything else than just checking the driver.
I'll have to give it a go, it will simplify the process, less steps, less time.
(yesterday with all these sweeps on such a large area I got lost at one time, what to EQ what not)
How close do you measure?
And how do you do for the HF drivers? (for flat)
Also I must say, I'm getting confused.
- after all these readings from Andy on how we should EQ the whole system (drivers+car reflections).
Yet here you're removing some of these effects of the equation.
- but in the same time I can see some value, maybe just trade-offs or for better balance like you said.
- and I'm close to think that these points were the ones blocking me before.
To take what I said and add some common sence in your own way and your own system. Your gain structure will be different than another and you will know how your amps and equipment like to be gained. So use good judgment and steer clear of large amounts of boost when there's already a strong signal going in/out. There's barely any digital headroom, in most recordings post y2k.
So in general when tuning your system for good balance there is a way with all the various in/out eqs.
My method works in my system and makes sence acoustically for my system and previous renditions on how I like things to sound. 1st off I undergain my amps. I think it sounds better and I like all analog signals going in to be there max without clipping. Just how I roll.
But as far as what I meant , if you have let's say a cdsp and you have input eq and output eq.
If your tuning the system as a whole you don't want to turn down any frequency if it doesn't affect all of the speakers on that channel. We're 2ch people so tuning needs be done in 2ch to be balanced. If you turn down let's say 500hz that will affect harmonic structure of the magnitude all the way up and down the spectrum. Haven't you ever turned up or down 20khz? Ever notice it will affect the other frequencies like 10khz and 5khz (I think it effects all the frequencies go but I'll use harmonics to make the point) if you change the harmonic structure of the signal only on part of the signal the harmonics won't be in sync so to speak. So any eq I make to a single Speaker I apply it to fix crossover slopes. So when measuring for abnormalities in crossovers I want to eliminate the roomout of my measurements so it more faithfully represents the electrical signal. right in front of the Speaker will dictate the speakers closest resemblance to the electrical signal so that's where you want to make measurements for making acoustic and electrical signals match for crossover slopes. If I want a linear crossover I need the crossover acoustic behavior to match the linear slope that is electric so that way it's doing what it says and not something else.
So stop band is everywhere below/above the crossover point. When I make these close measurements I want to flatten the areas below/above my crossover points. Except I will have the crossovers off for this. Now the common sence. If the close measurement isn't manageable within a few db of cut boost on it I won't radically alter the signal. It's like we just want a bit of smoothing so the response is more flat. As long as it sounds good it will work. Meaning again I wouldn't go boost a frequency 10db or something absurd. If the speakers close responce isn't already close to flat I fix the problem to make it more flat . At crossover and the first oactave I try to make pretty flat. Remember this will bean area that will be attenuated , so the louder levels will be heard more so its more important closer up the slope and less important down slope. Depending on how shallow your slope is would also bag for some common sense on how much to use . You just want the responce to be somewhat flat so when you do engage the crossover it will do what you see on your dsp screen and not something else like have a big peak half oactave into the crossover.
Once crossovers are measured and linearized than I apply gloabal eq to actually flatten the frequency response at the listening position.
So why is it okay to add eq at a specific speaker at crossover and not anywhere else? Because that will be an attenuated spot in the spectrum the summed signals from the Speakers overlap areas will be "in tune" first. Then it's time to address the room and the system as a whole . There are some people who are dead set against anything but global EQ . Saying there's no need for it if your crossovers are good and placement is good, and I am onboard with that a single 30band eq on a system an offense more deadly than the most exotic processor in the wrong hands. And I think that that sentiment is valid, to keep the fidelity of the original signal. Only because environmental issues are so complicated to deal with . That's why I am really interested in Dirac and apl. It's a approach that effects the entire signal at once. It does it just fix phase per speaker he just fixes the phase of the entire signal which would be better on a more grand scale two include timing/phase moreover FR.