Low SPL level measurement

[hbap349]

hi all,

I always have low SPL reading in my REW measurement to speakers, only getting range of 35 to 45dB while i see other are making beyond that, anybody can highlight to me what could be the cause of that. I am using Umik-1 with calibration file input into REW.

 

[John Mulcahy]

That's closer to a background noise level than a measurement level. Assuming you can hear the sweep there may be a problem with your measurement setup. Easiest is probably to post an mdat file from a typical measurement.

 

[DanDan]

There might be a level control in your OS, not at nominal. On Mac take at look into AudioMidi Setup.

 

[hbap349]

That's closer to a background noise level than a measurement level. Assuming you can hear the sweep there may be a problem with your measurement setup. Easiest is probably to post an mdat file from a typical measurement.

Hi John,

Here's the mdat file. I run Pink PN for those measurement.

 

[John Mulcahy]

You can use the RTA Appearance setting "Adjust RTA levels" to have REW shift the plot so it is more similar to the results that would be obtained with sweep measurements. When you use noise the energy is spread across the bandwidth of the signal, so though the overall SPL of the signal may be at 75 dB (say) the energy at any particular frequency is much lower. A sweep, in contrast, has the specified level at every frequency as it progresses through the measurement range, and gives the results you more typically see.

 

[jschwender]

What is the explanation for these stair steps?

 

[John Mulcahy]

It is an RTA measurement with the "Use bars on RTA" option selected.

 

[hbap349]

You can use the RTA Appearance setting "Adjust RTA levels" to have REW shift the plot so it is more similar to the results that would be obtained with sweep measurements. When you use noise the energy is spread across the bandwidth of the signal, so though the overall SPL of the signal may be at 75 dB (say) the energy at any particular frequency is much lower. A sweep, in contrast, has the specified level at every frequency as it progresses through the measurement range, and gives the results you more typically see.

So there isn't anything wrong on my spl level output being low after measure?

 

[John Mulcahy]

No, those are correct figures for the measurement method you used.

 

[hbap349]

No, those are correct figures for the measurement method you used.

Alright, thanks and appreciated for your feedback.

 

[jtalden]

The posted file measurements were created used 'spectrum' mode. The 'RTA' mode should be used with the PinkPN signal.

 

[hbap349]

The posted file measurements were created used 'spectrum' mode. The 'RTA' mode should be used with the PinkPN signal.

jtalden,

I used PinkPN signal for the measurement, is there anywhere in REW I should correct my setting?

 

[John Mulcahy]

In the RTA graph controls change the mode from Spectrum to RTA 1/48.

 

[hbap349]

..........When you use noise the energy is spread across the bandwidth of the signal, so though the overall SPL of the signal may be at 75 dB (say) the energy at any particular frequency is much lower.

John,

I have a question, when we run the measurement for each individual channel (speaker), do we level the volume up or down to SPL signal to 75dB or just a fix volume level to measure each speaker individually?

 

[John Mulcahy]

A fixed volume is usually fine unless there are very large differences between speakers, as long as the input doesn't clip the results should be OK.

 

[hbap349]

A fixed volume is usually fine unless there are very large differences between speakers, as long as the input doesn't clip the results should be OK.

Thanks John, understood from now.