Measuring in room freq response questions

0 Members and 1 Guest are viewing this topic. Read 432 times.

SoCalWJS

Measuring in room freq response questions
« on: 7 Jun 2018, 03:53 pm »
Once again I found the need to play around with the settings on the Servo amp running on the Super V's. Wife is away for several hours, so I can get away with Test Tones. Pulled out my old Radio Shack digital SPL meter and the Alan Parsons & Stephen Court MFSL Sound Check 2 CD along with my printed copies of the settings sheet for the Servo Amp, ready to go to town. I've noticed it before, but decided to see what specific techniques people use to get their best ("most accurate") readings. I run only one channel at a time and am measuring from my main listening position with the meter held at the approximate location of my head with the microphone pointing mostly straight up, but tilted slightly forward towards the speakers (straight forward between the speakers, not directly at the one being tested), in the "C Weighting" position. I hope this is all correct.

I notice that at the lowest frequencies of the CD (20 &25 Hz), the readings jump all over the place, well in excess of 10 db.
- is there a way to minimize this a bit?
- should I be doing something different?
- is slow or fast response more accurate at these low frequencies?

I know the RS meter is way off at those frequencies (I usually read 4-5 db at 25 Hz and 6-7 db at 20 Hz), but I'm still confused as to what I should use as my actual reading when it's swinging back and forth so much.

Any tips?

If I should post this in the Acoustics Forum, please fell free to move it. Thanks.

HAL

  • Industry Contributor
  • Posts: 4026
Re: Measuring in room freq response questions
« Reply #1 on: 7 Jun 2018, 04:02 pm »
Are you using the C weighting and slow averaging on the RS SPL meter to make the LF measurements?   

You can correct the measurements with one of the Excel spreadsheets as that will help know the LF response from the RM meter.

Another option is Room EQ Wizard and use the RS meter as input and PC output to drive the system.   REW has worked well here. You can get the calibration data from the spreadsheet to use in REW.




SoCalWJS

Re: Measuring in room freq response questions
« Reply #2 on: 7 Jun 2018, 04:14 pm »
Are you using the C weighting and slow averaging on the RS SPL meter to make the LF measurements?   

You can correct the measurements with one of the Excel spreadsheets as that will help know the LF response from the RM meter.

Another option is Room EQ Wizard and use the RS meter as input and PC output to drive the system.   REW has worked well here. You can get the calibration data from the spreadsheet to use in REW.
I've never been able to get the hang of REW. Bought one of the external contraptions (can't think of the exact name now - it's been years. Sound card sort of thing) and ran through the directions and wasn't able to get it to work. Ended up buying the OmniMic system. I use it as well, but have the same issue - the readings jump all over the place.

I like checking in-room response with the RS meter and Sound check CD so that I can check specific frequencies rather than a sweep.

I'll try switching back and forth on fast vs slow response times to see if that helps. Yes, using "C" weighting.

****EDIT*****

Very similar results in "Slow" response mode, maybe a bit better, but still 10 db swings

HAL

  • Industry Contributor
  • Posts: 4026
Re: Measuring in room freq response questions
« Reply #3 on: 7 Jun 2018, 09:08 pm »
If OmniMic software has either an MLS style or Log Sweep style measurement capability like REW, you should be able to select 1/3 Octave smoothing for the data to help with the LF variations.   




SoCalWJS

Re: Measuring in room freq response questions
« Reply #4 on: 7 Jun 2018, 10:15 pm »
If OmniMic software has either an MLS style or Log Sweep style measurement capability like REW, you should be able to select 1/3 Octave smoothing for the data to help with the LF variations.
Yeah, OmniMic has that capability. Not sure if it helps with variations or if it hides details  :green:

I know the last time I ran a check with OmniMic and ran it at 1/3 Octave, my in room response was practically flat at main listening position, so I just kept reducing so I could get better detail and usually end up at 1/48th or 1/96 smoothing. I know that 1/3 is the spec that many Manufacturers use to cite specifications, but think that hides what is really happening with respect to room nodes and that is what I am trying to minimize via placement and treatments.

(still need to get better at OmniMic and figure out how to run a Waterfall. The documentation is not very clear to me. I know I have to run multiple tests and then somehow combine them, but the methodology is not explained very well)

mojave

  • Full Member
  • Posts: 326
Re: Measuring in room freq response questions
« Reply #5 on: 8 Jun 2018, 04:19 pm »
(still need to get better at OmniMic and figure out how to run a Waterfall. The documentation is not very clear to me. I know I have to run multiple tests and then somehow combine them, but the methodology is not explained very well)
I think you are thinking of something other than waterfall charts. You don't run multiple tests or combine them for a waterfall. All you do is press the waterfall button as it says in the manual:

The Waterfall feature becomes available when you click the Waterfall button (above the Frequency Response graph, next to the Smoothing control).

Danny Richie

  • Facilitator
  • Posts: 12034
    • http://www.gr-research.com
Re: Measuring in room freq response questions
« Reply #6 on: 8 Jun 2018, 04:31 pm »
To run a waterfall plot you need to do so on the speaker in an anechoic chamber or using a gated time window to remove any and all room reflections. You can't get any real data running one on a room response.

SoCalWJS

Re: Measuring in room freq response questions
« Reply #7 on: 8 Jun 2018, 07:58 pm »
I think you are thinking of something other than waterfall charts. You don't run multiple tests or combine them for a waterfall. All you do is press the waterfall button as it says in the manual:

The Waterfall feature becomes available when you click the Waterfall button (above the Frequency Response graph, next to the Smoothing control).
I've clicked that button many times. It comes back with "For Polar display, must have 3 or more added curves with different angles assigned!"

.......whatever that means.  :scratch:

******EDIT********

CR@P!!!!!! I've been clicking on the wrong icon all this time. (Polar plot vs Waterfall)

MUST LEARN TO GO SLOWER ON THESE THINGS!

Next time I'll know better.  :duh: :duh: :duh: :duh: :duh:

SoCalWJS

Re: Measuring in room freq response questions
« Reply #8 on: 8 Jun 2018, 08:02 pm »
To run a waterfall plot you need to do so on the speaker in an anechoic chamber or using a gated time window to remove any and all room reflections. You can't get any real data running one on a room response.
Maybe I don't correctly understand what it does then - I thought it gave you decay times of the selected frequency sweep. Gives you some idea of whether or not you have too much/too little absorption at various frequencies.

"Gated time window". I'll have to look that one up.  :oops:

Danny Richie

  • Facilitator
  • Posts: 12034
    • http://www.gr-research.com
Re: Measuring in room freq response questions
« Reply #9 on: 8 Jun 2018, 08:23 pm »
Maybe I don't correctly understand what it does then - I thought it gave you decay times of the selected frequency sweep. Gives you some idea of whether or not you have too much/too little absorption at various frequencies.

"Gated time window". I'll have to look that one up.  :oops:

No, it's a speaker measurement and not a room related measurement.

mojave

  • Full Member
  • Posts: 326
Re: Measuring in room freq response questions
« Reply #10 on: 13 Jun 2018, 07:35 pm »
Maybe I don't correctly understand what it does then - I thought it gave you decay times of the selected frequency sweep. Gives you some idea of whether or not you have too much/too little absorption at various frequencies.
Yes, that is exactly what it does. It is a different method of showing the impulse response in the time domain. Room EQ Wizard provides a good explanation of how it works. Also see
REW: Understanding decay and waterfall plots.

The waterfall graph aids in seeing the decay time and room resonances. This article by GIK Acoustics demonstrates how:  Understanding Decay Times and Waterfall Graphs. RealTraps also recommends using the waterfall for viewing decay time.