Driver level matching is where we approximately match the amplitude of individual drivers so that when their outputs are combined they measure flat. We could, if we wanted to, allow the room correction step to correct for this but what your actually doing is shifting the target acoustic roll offs that you set when created the crossover. For example you created 2nd order rolloff but since the room correction step works on a unity style approach then when it sees a level mismatch and adjusts the responses for individual drivers so they follow the room correction target it actually not only adjusts the gain of a driver but also the rolloff of the adjacent driver. No longer do you have a correct acoustic rolloff.
Level matching before the room correction step allows for a situation where driver outputs are roughly equal and no wideband shifting of gain will occur. In a nutshell room correction is only intended to correct for the room and isn’t a substitute for proper level matching.
First thing to do is setup your mic at approximately 1m distance on axis with the tweeter. Your loudspeakers will be in position within the listening room. How many measurements you will need to take depends on the number of ‘ways’ your speaker is. For a 3-way you will need 5 measurements per speaker. For a 2-way, 3 measurements.
!IMPORTANT! For this method to work correctly do not move the mic position and any gains through the recording section should be set and then left alone ie. mic preamps, channel level for mic and recording level. We’re relying on relative measurements here so this is very important for best results.
OK I’m going to detail level matching 2 drivers in a 3-way – the midrange and tweeter. If you have more drivers that need matching then simply extend the process. An important note is to work from the top down and by that I mean start with the tweeter to midrange matching and then after those are matched move on to midrange and bass. If you find that you have to adjust gain on the midrange when comparing to bass AFTER you’ve level matched it to the tweeter then that gain reduction needs to be applied to the tweeter as well to maintain correct results.
You will now perform a series of logsweep so create a 50hz-15Khz logsweep within Acourate. Also be sure you’ve loaded your linearised filters for the loudspeaker about to be measured, into the convolution engine along with the time alignment figures we worked out. After this do the following:
Perform logsweep for tweeter(all other drivers off)
Perform logsweep for midrange
Perform logsweep for midrange and tweeter combined (optional but very useful)
Save the results of each.
Now within Acourate do logsweep convolution for the tweeter measurement then cutnwindow to get the pulse: