New REW SPL curves that don't align with previous ones

[hilde45]

This is a question about why new SPL measurements don't line up (at 20db) with previous ones and what to do about it.

Here's the situation: I'm using REW to integrate a second sub. The subs are Rythmik F12SE and have a number of controls that I'm adjusting.
I'm varying "Delay/Phase" "Crossover" and/or "Volume."
Before each new measurement, I use "check levels" and ensure pink noise is hitting at 80 db.
On the subs, for each new measurement, I only change one variable at a time.
Puzzling outcome: sometimes, when I vary "Delay/phase" or "Crossover" the new reading's SPL line won't line up at the starting point (20 db) with previous measurements.

A couple questions:

1. Why does this happen? While I understand why changing the "Volume" setting might alter the starting point of the SPL curve (because I'm adding volume!), why do changes in these non-volume settings do it? My assumption is that the room is reacting in ways that are amplifying the response with this new setting. Because if it's not my initial settings, it must be the room's response to the new setting on the subs.

2. Regardless of the explanation to (1) above, what should I do about it? Should I adjust the new curve to fit the older ones for comparison's sake? What meaning should I take away from a new curve that doesn't start where the others do, db-wise at 20db? Should I use the "measurement actions" control to raise or lower the curve to match the others? And do I do that at the initial part of the curve (i.e., at 20db)?

Thanks so much. It's really hard to know what to make of a new curve when it seems as if it needs alignment with previous measures.

 

[John Mulcahy]

The overall SPL from more than one sub depends on how they sum, which will depend on the relative amplitude and phase of the signal form each sub at the mic position. Changing the delay varies the phase so the summation will be different. The main interest isn't so much the level of the curve as its shape, more even and fewer dips is better.

 

[hilde45]

The overall SPL from more than one sub depends on how they sum, which will depend on the relative amplitude and phase of the signal form each sub at the mic position. Changing the delay varies the phase so the summation will be different. The main interest isn't so much the level of the curve as its shape, more even and fewer dips is better.

Thank you, John. Does your answer imply that I do *not* need to adjust the measurement in order to compare it effectively with previous curves? Or should I line them up?

 

[John Mulcahy]

No, you do not need to adjust them.

 

[hilde45]

No, you do not need to adjust them.

Great. So -- this will be my last question, and thank you for your forbearance -- so when I'm looking at two curves, and comparing them, and one starts lower or higher than the other, you're saying that what I need to do is assess them based on their shape (relative to flatness)?

I only want final clarity on this question because it will save me a lot of future trouble in how I handle new measurements.

Continued gratitude for the most powerful audio tool I have ever used.

 

[John Mulcahy]

Yes, that's right. The particular level a graph happens to start at isn't usually of much consequence.