Plotting Pink Noise Measurements

[Alchematic]

I would like to use the AD2 to measure speakers and rooms in professional recording studios.

I use Bruel & Kjaer 1405 noise source and a Bruel & Kjaer 2235 SPL meter with line output into the AD2

Putting the output of the noise source into the AD2 plots a flat frequency response with White noise and a -3 db per octave plot with Pink noise.

It is traditional to make acoustic measurements with Pink noise.

It the past the plotting systems had a choice of plotting flat with either White or Pink Noise.

I would like to see if Waves can be calibrated or selected to plot Flat with Pink noise.

Otherwise my solution is to insert a +3db/ octave analog filter before the AD2. This is not optimum as there is an insertion loss.

Or use another system like Room EQ Wizzard

Can Waveforms plot with Pink Noise displaying as flat?

Thanks

Mark

[JColvin]

Hi @Alchematic,

I am not certain of the answer, but the developer of the WaveForms software should be able to give you a more definitive answer when they come back to the office next week.

Thanks,
JColvin

[Alchematic]

Thanks Mr Colvin

I look forward to talking to the software developer

Would you share his name so I know who may be contacting me?

I also have some discrepancies with measurements made by Waveforms and by by Tektronix test equipment

Thanks

Mark

[attila]

Hi @Alchematic

This is what you need ?

var mag = Spectrum.Trace1.magnitude;
var c = mag.length;
for(var i = 1; i < c; i++){
    mag[i] = mag[i]-10.0*log10(i/c)-30;
}
Spectrum.Trace2.magnitude = mag;

 

Edited December 5, 2022 by attila

[Alchematic]

Hi Attila,

I am assuming the blue trace is as normal for pink noise and the orange trace compensated for pink noise to be flat.

Then yes that is what I need.

Do I create a script as you have sent me?

I have not done that before.

I have copied the script and placed it in the spectrum window as shown but it does not seem to do anything. 

It seems to run in the SAMPLE mode

Can it run in EXPONENTIAL RMS AVERAGE mode?

Thanks

Mark

 

 

Edited December 5, 2022 by Alchematic

[Alchematic]

[attila]

Hi @Alchematic

With averaging use Execute: Manual and use Save to execute the script.

Setting the magnitude generates an internal signal similar to adjusting an option and the averages reset, so setting magnitude automatically for each acquisition will prevent averaging to work. I've changed this behavior for the next sw version.

[Alchematic]

Hi

 

I tried the script in manual

 

Attached are 3 plots. All have Blue C2 as a normal plot of pink noise and Yellow as 1) no script. 2) with script and 3 with the analog filter which has a 25 db insertion loss.

I wish I could get a display like the analog filter with out the level loss

Mark

[attila]

Hi @Alchematic

For this the sign needs to be changed to: mag[i] = mag[i]+10.0*log10(i/c)+30; 

Edited December 6, 2022 by attila

[Alchematic]

I did this but it still is not right.

First image with script sign change

Second C1 running C2 HOLD mode then SAVE in script pressed. Closer but still a huge spike at levels below 100HZ

Mark

 

[attila]

Hi @Alchematic

You have a new line in the +30; so you are only adding +3 and not +30

[engi]

Interesting subject. I am going to use AD2 for audio too, whereby I wonder about the quality of input noise. Generally I would go with a good sound card or a professional device. Also for the microhones i suggest a noise free pre amp with a good and noise free 48V phantom power.

Me I am going with RME' ADI Pro and fireface. For Audio analysis I use ARTA, and selfbuilt gear.

Regarding AD2, I wonder about the linearity of the frequency curve. (?)

[attila]

Hi @engi

The WaveForms application has a demo like sound card support.