Noise floor / Dist. harmonics, with REW and UMIC-1

[bothu]

I have bought a UMK-1 and installed REW and started to measure on my homemade speakers.

My first question is about noise floor.
The RTA plot is taken in our summer house far from big roads and other houses.
There was no wind outside and no other noises. I switched off the incoming mains.
I put the mic. in the sleeping room and closed the door and had the computer outside the door.
The USB cable was about 3,6 meters ( 12 feet). So the condition for a god result was there.
But .... the result .... hf-peaks an still a high noise floor.
I had expected a much lower noise floor.
With a better mic. and no cable issues what figures can I expect ?

My second question is about distortion measurements in REW.
When I look at the dist. plots of my speaker I can see some plots under the noise floor.
How can that bee ? For me it seems impossible to measure something that is hidden in noise.
A related question is : In the control box there is an option for "Plot harmonics at the the harmonic frequency"
Witch setting is to prefer ?

My speaker:

Noise floor :

Dist plot: Blue is noise floor.

Best wishes from Bo Thunér in Sweden. ( Still learning and having fun ! )

 

[KMO]

But .... the result .... hf-peaks an still a high noise floor.
I had expected a much lower noise floor.

That noise floor looks reasonable to me. A normal "silent" room is still going to be above 20dB SPL. With REW you're mainly going to be processing properly audible 75dB+ test signals - I wouldn't particularly worry about stuff down below 40dB.

When I look at the dist. plots of my speaker I can see some plots under the noise floor.
How can that bee ? For me it seems impossible to measure something that is hidden in noise.

Indeed - any harmonics below the noise floor are not significant. REW is showing you the result of its mathematical processing, but it's just noise.

There is a "Mask harmonics below noise floor" button that would grey out those lines.

Although it's not quite as straightforward as "below the line" - a harmonic might be below the noise level for the fundamental frequency, but above the noise level for its own frequency. A straightforward "below the line" check only works with the next option on...

A related question is : In the control box there is an option for "Plot harmonics at the the harmonic frequency"

You normally want that off. We normally assume any non-fundamental frequencies (above the noise floor) are harmonic distortion of our sweep signal. Normal distortion tends to generates multiple harmonics simultaneously, and having that setting would show all those harmonics aligned at the fundamental frequency. Generally if there's anything significant in the distortion, you should see it showing up on multiple harmonics, aligned, with that option off. And THD adds them up to give the total. We see which input frequencies are problematic.

If you ever saw what looked like staggered distortion, with harmonics appearing sequentially, then it could be some other source of noise rather than distortion of the sweep input. Flipping that switch to "on" would show the shape of that noise, by viewing output frequencies.

(Eg a constant 15kHz CRT whine would show up on a normal graph as a 7.5kHz 2nd harmonic bump, a 5kHz third harmonic bump, a 3.75kHz 4th harmonic bump, a 3kHz 5th harmonic bump. It's not obviously explicable by the input frequency. Flipping the switch to "on" would show a single 15kHz spike, revealing the stray signal).

 

[bothu]

Thank You KMO !
That was a professional answer, I bow my bold head to You with a smile.
ASR is based on technical fundamentals, I am real happy to bee a member here.