New room, new sweeps -- Why the difference in starting points and how to adjust with controls?

[hilde45]

Hi all,
I've searched a bit on threads and don't see the answer to this, though it's probably out there. I just built a room and am doing sweeps for 2 channel.

TASK: Taking measurements to optimize speaker placements of speakers and listening position.

ISSUE: The starting points of my SPL graphs (at 20 hz.) show great variance — e.g. with a test level of 80 db. for sweeps, my SPL graphs range from 60 db - 69 db. I want to compare what seem like disparate readings. I want to rule out user error and also adjust things with any tools, if that's what I need to do.

FYI:
My mains have a bottom range of 35 hz. They are stand mounts.
When I measure, I always "Check Levels" to ensure the same db reading. (80db to start.)


QUESTIONS:

1. Are disparate db starting points normal? In other words, is this not ME but rather my room, just reacting differently because of changes in placement?

2. If disparate starting points are not normal, what should I do to align them with either my initial process or my calibration?

You can see I only want to find the best position for speakers and LP using the graphs generated. I'm just trying to make sure I'm doing an apples-to-apples comparison rather than trying to interpret mistakes in my procedure.

Thank you.

 

[John Mulcahy]

1. Are disparate db starting points normal? In other words, is this not ME but rather my room, just reacting differently because of changes in placement?

Yes, completely normal, and the variation with position can be much larger than you have observed.

 

[hilde45]

Yes, completely normal, and the variation with position can be much larger than you have observed.

Thank you, John.

Then I need to do nothing to calibrate those graphs to align them -- just read them as is? Is that what I'm to infer from your remark? Thank you.

 

[John Mulcahy]

Hard to say without seeing anything. REW can align graphs over a chosen SPL range after they have been measured if desired.

 

[hilde45]

Hard to say without seeing anything. REW can align graphs over a chosen SPL range after they have been measured if desired.

Understood. I don't really know when I'd need to align them -- that's the fundamental issue. I'm probably too novice to be using a tool this sophisticated!

 

[John Mulcahy]

To optimise positions look at the shape of the responses rather than their level, fewer/smaller dips and peaks is better. Room responses show a lot of variation with frequency and position, don't expect them to look flat.

 

[hilde45]

To optimise positions look at the shape of the responses rather than their level, fewer/smaller dips and peaks is better. Room responses show a lot of variation with frequency and position, don't expect them to look flat.

Thank you, John. I've done this for about a year at this point, and don't expect flatness.

Your advice about gauging the relative shapes — the drama of troughs and peaks — is well taken.

The reason I'm asking is that as I move things in the room to gauge better and worse, it's sometimes hard to compare to curves in order to know what is happening in the room. That's where the urge to align them comes from.