ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1] 2  All   Go Down

Author Topic: How can I get good coherence in FFT measurements  (Read 1741 times)

JamesJudson

  • Newbie
  • *
  • Offline Offline
  • Posts: 23
How can I get good coherence in FFT measurements
« on: May 03, 2024, 08:47:13 AM »

hey prosoundweb forums, I am trying to tune my church PA using OpenSoundMeter. I set up all my routing and gain staging but when I was trying to take a measurement the phase and magnitude lines are just not there (because bad coherence). I am like driving my PA at very high volume and still the coherence is bad. I can't record a measurement like this so I reduced coherence threshold from 0.7 to 0.2. After this I can get a readable data but I am worried that maybe I was measuring noise instead of measurement signal. What should about this?
Logged

Tim McCulloch

  • SR Forums
  • Hero Member
  • *
  • Offline Offline
  • Posts: 23931
  • Wichita, Kansas USA
Re: How can I get good coherence in FFT measurements
« Reply #1 on: May 03, 2024, 09:05:23 AM »

The coherence trace is telling you the mic is in spot with lots of reflections contributing to the sound field in that spot.
Logged
"If you're passing on your way, from Palm Springs to L.A., Give a wave to good ol' Dave, Say hello to progress and goodbye to the Moonlight Motor Inn." - Steve Spurgin, Moonlight Motor Inn

Riley Casey

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2125
  • Wash DC
Re: How can I get good coherence in FFT measurements
« Reply #2 on: May 03, 2024, 09:23:34 AM »

Things you can do to improve coherence -

1 - reduce reflections by putting the mic in the middle of a seating area with soft seats or moving blankets
2- reduce reflections by making a pressure zone measurement - put the mic on a hard floor or a large board with the capsule within 5 mm of the surface
3 - reduce reflections by putting the mic as far as possible from any reflective surface ( 15ft tall mic stand trick), measure the distance from the nearest surface and ignore measurement data below that wavelength. Not really applicable unless you have 30+ ft high ceilings
4 -make the measurement closer to the speaker ( but too close means you're measuring the speaker rather than the speaker response in the acoustic space )
5- turn up the level of the system output to move the 'near field' further from the speaker & toward the mic ( sounds like you've already tried this tho)

John Schalk

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 639
Re: How can I get good coherence in FFT measurements
« Reply #3 on: May 03, 2024, 09:50:04 AM »

....I reduced coherence threshold from 0.7 to 0.2. After this I can get a readable data but I am worried that maybe I was measuring noise instead of measurement signal. What should about this?
Riley's list is really good, but based on my "still learning" experience with measuring sound systems for the past year with Smaart, a coherence blanking threshold of 0.20 is normal in the real world.  Even with that, my coherence trace is usually a pretty jagged red line.  Part of the learning process for using these analyzers is being able to recognize what the data on the screen means.

Do you need to see live data?  If not, try using Room Eq Wizard (REW) to capture your system's response.  It uses a tone sweep that can be time-windowed.
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2543
  • Edmonton, AB
Re: How can I get good coherence in FFT measurements
« Reply #4 on: May 03, 2024, 12:51:24 PM »

{...} 5- turn up the level of the system output to move the 'near field' further from the speaker & toward the mic ( sounds like you've already tried this tho)

I'm not sure I understand this one; other than helping to out-compete any ambient noise, how does using a louder test signal improve coherence?

-Russ
Logged

Riley Casey

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2125
  • Wash DC
Re: How can I get good coherence in FFT measurements
« Reply #5 on: May 03, 2024, 01:09:38 PM »

If you've ever tried to tune a system without quiet time built into the production schedule you'll know what I mean. That plus air handling noise and such its just a more energy arriving at the mic means better signal to noise thing.


I'm not sure I understand this one; other than helping to out-compete any ambient noise, how does using a louder test signal improve coherence?

-Russ

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2543
  • Edmonton, AB
Re: How can I get good coherence in FFT measurements
« Reply #6 on: May 03, 2024, 09:43:39 PM »

If you've ever tried to tune a system without quiet time built into the production schedule you'll know what I mean. That plus air handling noise and such its just a more energy arriving at the mic means better signal to noise thing.

Ah okay, so it is just about out-competing ambient noise; makes sense!

(I tend to conceptualize the "near field" of a speaker as being more about direct-to-reverberant ratio instead of signal-to-noise ratio, and was therefore confused by the idea of "moving the near-field by adding level" since the reverberant characteristics of a room are level-agnostic.)

-Russ
Logged

Tim McCulloch

  • SR Forums
  • Hero Member
  • *
  • Offline Offline
  • Posts: 23931
  • Wichita, Kansas USA
Re: How can I get good coherence in FFT measurements
« Reply #7 on: May 04, 2024, 10:23:21 AM »

Ah okay, so it is just about out-competing ambient noise; makes sense!

(I tend to conceptualize the "near field" of a speaker as being more about direct-to-reverberant ratio instead of signal-to-noise ratio, and was therefore confused by the idea of "moving the near-field by adding level" since the reverberant characteristics of a room are level-agnostic.)

-Russ

What you said.  I've found that 6dB above ambient is more than sufficient to provide coherency; even mixing in a little pink noise with the pre-show music will increase coherency.

Increasing the pink noise level will do nothing for reflection-induced incoherence.
Logged
"If you're passing on your way, from Palm Springs to L.A., Give a wave to good ol' Dave, Say hello to progress and goodbye to the Moonlight Motor Inn." - Steve Spurgin, Moonlight Motor Inn

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2543
  • Edmonton, AB
Re: How can I get good coherence in FFT measurements
« Reply #8 on: May 04, 2024, 05:05:44 PM »

{...} even mixing in a little pink noise with the pre-show music will increase coherency. {...}

You just need to find music that has a more broadband spectral make-up; thrash metal, maybe?  ;D

-Russ
Logged

Riley Casey

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2125
  • Wash DC
Re: How can I get good coherence in FFT measurements
« Reply #9 on: May 04, 2024, 06:25:44 PM »

IOW music that's so close to pink noise the difference is negligible to software ...

You just need to find music that has a more broadband spectral make-up; thrash metal, maybe?  ;D

-Russ

ProSoundWeb Community

Re: How can I get good coherence in FFT measurements
« Reply #9 on: May 04, 2024, 06:25:44 PM »


Pages: [1] 2  All   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.04 seconds with 24 queries.