Can REW measure background room noise?

[PeteD]

Can you just run a sweep with the output disconnected?

I am interested in seeing what frequencies and levels of background noise I have in my listening room. I think I am seeing evidence of background noise on my waterfall plots and want to check levels without a sweep.

Thanks,
Pete

[Otto]

Sure. Just disconnect the line-out signal and run the sweep.

[brucek]

I am interested in seeing what frequencies and levels of background noise I have in my listening room. I think I am seeing evidence of background noise on my waterfall plots and want to check levels without a sweep.

Otto's quite correct, but you do have to be aware of setup levels so as not to over-estimate backround noise impact. In the waterfall thread when someone told me that waterfalls were of limited use and that a lot of the information was noise, I did exactly what Otto suggests to see if it was true.

I setup REW just as I would normally do for a subwoofer measurement, except I removed the output signal after setting up the input/output levels, and then did a measure. Well, the noise in the room was quite low, and of course REW gives a warning that is easily bypassed with a click.

But there was some low level noise (that wouldn't affect a measurement), but I thought it would be interesting to setup the levels to a proper reading and examine the noise more carefully by adding a known noise source in the room to act as a reference.

So, with the output cable from REW disconnected and the mic at my listening position, I setup the levels to allow about a 6dB of headroom in the measure and take a reading with no extra noise in the room and then (without changing level settings) take measures with my fridge on, and my furnace on, and lastly and most revealing, I brought my office subwoofer close to the mic and took a reading. That sub has a power supply filter fault and has a nasty 60Hz hum while simply plugged in.

It's best to use long sweeps with multiple passes. I used 512 size sweeps with 4 passes. This reduces random noise quite a bit and give a better reading of persistant noise sources in the room.

Below is a response and a subsequent waterfall overlay of the base noise in the room and then the same measure with my noisy faulty sub. Note the 60Hz hum with multiples of 120Hz, 180Hz.

But remember, I had to turn up my microphone level (or with a Radio Shack meter or Galaxy meter use the lowest scale). I also setup my Check Levels routine as if I actually had an REW output instead of the noise level I was picking up. At normal setup levels, not much of this noise is readable, which debunks of course the notion that waterfalls are a lot of noise...(in my house anyway)

Anyway, kinda interesting. My furnace sure outputs some low frequency noise. I do indeed shut it off when I'm serious about measuring.

This little experiment shows you can examine background noise over the frequency bandwidth of interest if you think there's a problem. First do your sweep after a normal setup of levels and then remove the output cable before measuring. This gives an idea if the noise has an effect at normal levels. Then you can setup the levels (leaving the output cable removed) to examine down a bit deeper.

Green = background noise. Purple = background noise plus humming sub.
Note that absolute SPL levels will not be accurate because of level adjust. This is to examine noise against reference floor noise.




brucek

[JohnM]

Of course what you really need is this:

[brucek]

You're killin' me John...............

[jerome]

Is this a new feature in the next version? I can't get it in my 4.00 build 673

[JohnM]

Yes, it is a new feature. Beta testing will start soon.