• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Help interpreting phase/SPL correlation

karrebarre

New Member
Joined
Nov 21, 2022
Messages
2
Likes
1
Hey,

I'm very new to all this and so I apologize if none of this makes sense :)

Right now I'm in the middle of construction of my new music studio and even though there's still ways to go until it's finished, I wanted to try and measure the room to help with decisions down the line.

The measurements were of two 2-way 6 inch studio monitors and an accompanying 8 inch sub. Because the measurements were made a bit om a whim all I had at hand was my flattest condenser mic, a Rode NT1 w/ calibration (yes, perhaps not absolutely optimal...). Anyway, looking at the SPL/Phase graph there seems to be some weird correlation between some aggressive phase shifts and a dip in frequency response (esp. evident around 4k, 12k and 20k).

REW_PHASE.jpg


My question to y'all more experienced than me is whether this is all down to the use of a LDC to do the measurements or if something else might be at play here?

Best,
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
2,383
Likes
1,731
Location
Canada
My question to y'all more experienced than me is whether this is all down to the use of a LDC to do the measurements or if something else might be at play here?

I haven't tried taking measurements using a large diaphragm condenser mic. However, you mentioned two speakers... Measure left and right individually. Vector average or measure both L+R when the correct time and phase alignment is confirmed. It could all be down to a microphone that's not perfectly equidistant to both speakers.
 
OP
K

karrebarre

New Member
Joined
Nov 21, 2022
Messages
2
Likes
1
I haven't tried taking measurements using a large diaphragm condenser mic. However, you mentioned two speakers... Measure left and right individually. Vector average or measure both L+R when the correct time and phase alignment is confirmed. It could all be down to a microphone that's not perfectly equidistant to both speakers.
Yes, this makes a lot of sense and I'm positive that'll get rid of the phasing. When you say "confirming time and phase alignment", would that simply be by measuring out the distances between the sources and and microphone and making sure they're the same, or how would you go about that in the best way?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
2,383
Likes
1,731
Location
Canada
Yes, this makes a lot of sense and I'm positive that'll get rid of the phasing. When you say "confirming time and phase alignment", would that simply be by measuring out the distances between the sources and and microphone and making sure they're the same, or how would you go about that in the best way?

Not sure about the "best way" as there are many ways to go about it.

A convenient way is to apply automatic time offsets to both individual left and right measurements prior summing/vector averaging.

In the All SPL tab view:

1669113280866.png


In most cases, this will align all selected measurements very close to their approximated start time, thereafter you can vector average the measurements without a problem -- I mean, assuming the phase profiles match, there should be no anomalous cancellations visible in the vector average or aligned summed result (the latter if using the 'Alignment tool').

If you want to get rid some of the room-caused phase wrap "noise", apply windowing to discard later arriving energy:

1669114183370.png



This will give you an "early" frequency response curve -- fewer cycles means more of the late arriving energy is discarded. Windowing is useful for a variety of reasons, say, when creating EQ filters. You would want to remove FDW before vector averaging, though.


----


All that said, I often just simply manually/physically center the microphone while looking at the magnitude and phase traces (or alternatively, the IR peaks) -- thereafter, perform a single L+R sweep.

1669115096910.png


Alignment Tool view (FDW 15 cycles):
1669116834497.png



An acoustic reference helps USB mics avoid or reduce clock drift errors like what I use (UMIK-1):

1669116121646.png

*Analysis tab in Options Window


Mics that use the same I-O audio interface (presumably yours) likely do not require clock re-adjustment, but a fixed reference output can still be useful so that you can determine the approximate delays/distance offset needed relative to a fixed reference.

1669116935038.png
 
Last edited:

Radixons

Member
Joined
Jul 19, 2020
Messages
37
Likes
11
Hey,

I'm very new to all this and so I apologize if none of this makes sense :)

Right now I'm in the middle of construction of my new music studio and even though there's still ways to go until it's finished, I wanted to try and measure the room to help with decisions down the line.

The measurements were of two 2-way 6 inch studio monitors and an accompanying 8 inch sub. Because the measurements were made a bit om a whim all I had at hand was my flattest condenser mic, a Rode NT1 w/ calibration (yes, perhaps not absolutely optimal...). Anyway, looking at the SPL/Phase graph there seems to be some weird correlation between some aggressive phase shifts and a dip in frequency response (esp. evident around 4k, 12k and 20k).

View attachment 245155

My question to y'all more experienced than me is whether this is all down to the use of a LDC to do the measurements or if something else might be at play here?

Best,
This is interesting. Will you keep us posted on the outcome of your efforts in terms of successful strategy and equipment used? I am also interested in conducting response measurements and would hope to participate in your realisations.
 
Top Bottom