Measurement when speakers are muted.

[Pros1]

Is this normal when taking measurements of a room without speakers?
What could cause that if it's unusual?

 

[Wayne A. Pflughaupt]

It’s the ubiquitous 60-cycle noise from a refrigerator, air contidioner, etc. Never seen it at 96 dB though…

Regards,
Wayne

 

[Pros1]

Thanks Wayne,
I'm still new to REW but i think something is wrong with the dB reading, i have an SPL meter on my monitor controller and my room noise is at around 31-38 dB.
I should get a UMIK-1 next week so i will redo the level calibration in REW.
I also thought about a refrigerator or air conditioner so i measured the room when the AC is off and i unplugged the refrigerator.. i guess it's the AC from other apartments.
When i play a 60Hz sine wave and walking in my room, in some areas it doesn't sound like a solid wave but more like LFO/Tremolo effect.
I took a 2nd measurement and this time the peak is at 120Hz.. about 15-20dB louder than the rest of the spectrum, i guess the only way to get rid of it is to move to a house.

 

[Matthew J Poes]

Thanks Wayne,
I'm still new to REW but i think something is wrong with the dB reading, i have an SPL meter on my monitor controller and my room noise is at around 31-38 dB.
I should get a UMIK-1 next week so i will redo the level calibration in REW.
I also thought about a refrigerator or air conditioner so i measured the room when the AC is off and i unplugged the refrigerator.. i guess it's the AC from other apartments.
When i play a 60Hz sine wave and walking in my room, in some areas it doesn't sound like a solid wave but more like LFO/Tremolo effect.
I took a 2nd measurement and this time the peak is at 120Hz.. about 15-20dB louder than the rest of the spectrum, i guess the only way to get rid of it is to move to a house.

You could still see it in a house too. I assume this measurement is not SPL calibrated as it seems off. You would hear that very loudly if it was really doing that.

Since 60hz is the carrier frequency of AC electricity you will find all line voltage devices may make this hum. Stereos may even pick it up and reproduce it.

Are there no speakers in the room or are they simply muted? If still connected to an amplifier that is on, but output is muted, you can still have a system hum.

In my soundproof theater room built into a basement and with dedicated purpose run AC lines for the stereo, I can still measure 60hz and it’s double (120hz) from various sources. The HVAC and whole house humidifier, very slight amplifier hum in the system, etc. one thing i figured out was that my receiver has a kind of noise gate to make it seem noiseless when no sound is being produced. When a slight signal is sent that turns off the gate, the noise floor rises dramatically and the 120hz component is readily seen.

 

[Pros1]

Thanks Matthew, this is very interesting, but i think your 60Hz issue comes from the equipment / electronics, and mine is more related to the outside noise, and i think this new measurement will show it.

The previous measurement wasn't SPL calibrated, i just took a measurement right after i installed REW, but now it is.

There is no hum in my system, this is a home studio, so no receivers or amplifiers.. it's just powered monitors to a monitor controller (both are high end) and all cables are balanced.
I never heard a hum thru the monitors nor recorded something that picked up a hum noise.

I'm adding two screenshots, one is me holding the microphone (showing two different measurements) and a second screenshot (again showing two different measurements) with the microphone on a stand.

As you can see, the measurement with the mic on a stand picks up more noise and more of the 60/120Hz issue.

There is some noise that is present in the entire building when air conditioner is running in other apartments.. temperatures can reach 105° during this month so i hear it much more lately.

 

[Matthew J Poes]

This is what I was talking about. I was just pointing out that there are many sources of noise.

What kind of mic are you using? And would you mind posting the meat files from these.

I’m a little surprised you don’t have more signal noise from the mic. You may find that the umik is much noisier.

 

[Pros1]

Thanks for pointing out other options for this issue.
The only thing i can hear thru my speaker when i put my ear close to it is a tiny hiss from the tweeter only, i can't hear anything from the mid driver or the woofer.

I'm using Apogee Mic96k USB microphone at the moment and the preamp gain is at max.
When you measure your room (just the room noise, after turning off the speaker / receiver / amp) do you see any bumps in the graph at around 60Hz?
I didn't save the other measurement but i made a new one, it should be similar, when the mic is on a stand, i see a peak of 60dB at 60Hz while the general average is maybe 20-30dB? i'm not sure but you can tell me what you think.

 

[Matthew J Poes]

Thanks for pointing out other options for this issue.
The only thing i can hear thru my speaker when i put my ear close to it is a tiny hiss from the tweeter only, i can't hear anything from the mid driver or the woofer.

I'm using Apogee Mic96k USB microphone at the moment and the preamp gain is at max.
When you measure your room (just the room noise, after turning off the speaker / receiver / amp) do you see any bumps in the graph at around 60Hz?
I didn't save the other measurement but i made a new one, it should be similar, when the mic is on a stand, i see a peak of 60dB at 60Hz while the general average is maybe 20-30dB? i'm not sure but you can tell me what you think.

If I measure my room's noise floor using a USB measurement mic like the UMIK or Dayton Audio, I just get mic noise, very little room noise.


I had to use a special measurement device and mic capsule to get at the room noise floor and even that device was questionable. I'm still working to capture the room accurately. the NTi device I used came with the noisiest of their microphone options, so I really need something else to be sure what I got with that setup is accurate. It showed a rise in low- and high-frequency noise and seemed to be artificially picking up noise, such as its own self-noise and the laptop noise.

Goal is to buy a quieter affordable lab grade mic, power supply, and preamplifier in the near future. I need it for my acoustics work anyway, as I need to be able to rate a rooms noise level and show that my specified design met the criteria, and right now my limit is like NC25, which is well above the level required for many studios and theaters.

In any case, you can see a little rise at 120hz in that room measurement. I'm fairly sure most of that is mic noise, but the spike at 120hz I think is line noise. The suspected culprit being a sub that uses the Dayton plate amp, which is known to make a little noise. There is no 60hz noise in this measurement but I had picked up some when I was using the NTi rig. I don't have ready access to the report from the NTi.

I'll try to post a measurement from my phone. I have a homemade low noise microphone with the audiotools software that I can use for room measurement ratings. That still isn't "right" but its better than the USB mic.

 

[Matthew J Poes]

Thanks for pointing out other options for this issue.
The only thing i can hear thru my speaker when i put my ear close to it is a tiny hiss from the tweeter only, i can't hear anything from the mid driver or the woofer.

I'm using Apogee Mic96k USB microphone at the moment and the preamp gain is at max.
When you measure your room (just the room noise, after turning off the speaker / receiver / amp) do you see any bumps in the graph at around 60Hz?
I didn't save the other measurement but i made a new one, it should be similar, when the mic is on a stand, i see a peak of 60dB at 60Hz while the general average is maybe 20-30dB? i'm not sure but you can tell me what you think.

Ok I looked at your measurements. Clearly, the noise floor on your mic is lower than the measurement USB mics. Something to keep in mind! It appears that its lower at all frequencies but has the same rise in low frequencies that mine does, it just doesn't rise as much.

I also see what you mean about holding with your hand vs a mic stand. With that in mind, it seems you are picking up transformer or motor noise. I know you said you believe that it is AC from your neighbors. That is totally possible, however, I don't think you are picking it up through the air. It appears to be noise coming through your floor.

The solution to this is to decouple your room from the rest of the rooms. You said this is a studio? If so there are a variety of viable options, depending on how its used. The most expensive is to dig up your current floor and replace it with a decoupled cement floor. Assuming that isn't an option, the next step is to add a decoupled floor on top of the existing floor. You can achieve this a variety of ways:

1. Add a decoupling mat between the current floor and newly installed floor. Most studios have hardwood or otherwise hard floors, so I would install this floor as a floating floor above the decoupling mat. I would be sure to leave a gap around the outside to avoid touching the walls. If this idea is viable for you I can let you know some companies that make good products for this.
2. Add a decoupling layer on top of the existing floor in the form of a special rubber mat. Not all products are created equal here and unfortunately, this approach only decouples well at certain frequencies. The frequencies it works best on depends on the thickness of the mat and the nature of the material.
3. Decouple just the devices that matter. In this scenario, you use microphone isolators and decoupling pads for those things that pick up the noise that get into your recording. It appears this would work fine for you since none of this noise seems to be detectable in the background noise of your handheld measurement.

 

[Pros1]

Thanks for sharing your measurement.

For some reason i thought that a measurement microphone will have much lower noise floor than a regular condenser mic.. i just now after reading your post did my research about noise level and indeed the UMIK-1 is not the best when it comes to noise floor.. i'm starting to think about Earthwork M23 which is probably an overkill for what i need but i see online that it's a great mic in general and has a 20dB SPL noise floor (about a half of the UMIK-1)


The main purpose of this whole thing for me (REW, measurement microphone, room noise etc.) is to get a nice mixing room with frequency response as flat as possible, but i'm willing to buy the better mic (M23) if the advantage is noticeable in a real life situation like mine.
I know this thread is about room noise when there is no speaker involve but this is due to the unusual situation in my apartment, i wanted to get an opinion and to confirm what i thought from someone more experienced.. i won't stay in this apartment for more than a few months so it's not worth the hassle to try and make this room as good as possible. (floating floor, soundproofing insulation)

The peak in your graph is more at around 145Hz so i don't know if it's a random peak or a line noise, but either way it's much smaller than the 30dB peak that i have at 60Hz so i wouldn't worry about this peak.

I wanted to ask you about the values in your graph, i see THD: 0.779 for example and for some reason i can't see any value in REW when i look at my measurement but i remember that this value was much higher in the measurement file that i posted here (maybe 20-40%, i can't remember) What does it mean?

One more thing that i would like to understand is about a graph of a UMIK-1 in a lab room that i found online, how is it that some of the graph is showing below 0 value?

 

[Matthew J Poes]

Thanks for sharing your measurement.

For some reason i thought that a measurement microphone will have much lower noise floor than a regular condenser mic.. i just now after reading your post did my research about noise level and indeed the UMIK-1 is not the best when it comes to noise floor.. i'm starting to think about Earthwork M23 which is probably an overkill for what i need but i see online that it's a great mic in general and has a 20dB SPL noise floor (about a half of the UMIK-1)


The main purpose of this whole thing for me (REW, measurement microphone, room noise etc.) is to get a nice mixing room with frequency response as flat as possible, but i'm willing to buy the better mic (M23) if the advantage is noticeable in a real life situation like mine.
I know this thread is about room noise when there is no speaker involve but this is due to the unusual situation in my apartment, i wanted to get an opinion and to confirm what i thought from someone more experienced.. i won't stay in this apartment for more than a few months so it's not worth the hassle to try and make this room as good as possible. (floating floor, soundproofing insulation)

The peak in your graph is more at around 145Hz so i don't know if it's a random peak or a line noise, but either way it's much smaller than the 30dB peak that i have at 60Hz so i wouldn't worry about this peak.

I wanted to ask you about the values in your graph, i see THD: 0.779 for example and for some reason i can't see any value in REW when i look at my measurement but i remember that this value was much higher in the measurement file that i posted here (maybe 20-40%, i can't remember) What does it mean?

One more thing that i would like to understand is about a graph of a UMIK-1 in a lab room that i found online, how is it that some of the graph is showing below 0 value?

A few reasons explain it. Ok so first, I learned this recently from John after taking a measurement in my theater using a lab grade mic and a spectrograph mode on the RTA. I too had -20 or so dB levels and was shocked. "My Gosh, I have an anechoic Chamber, its the quietest room on earth, how could this be?" John brought me back to earth by reminding me that the FFT mode is dividing the sound spectrum up into buckets, those buckets are very narrow. Take the total energy in the sound, think about the energy in high frequencies, now divide that up into tiny little buckets a fraction of a fraction of an octave wide. How much energy is left? Not very much, and thus, the SPL level is not very high.

This isn't a useful way of looking at sound, as I figured out, so room noise floor measurements are not done this way. In fact you are supposed to do it in 1/3 octave or really, 1 octave spacing. The official room ratings are based on 1 octave spacing (like NC or NCB levels, for example). That packs more energy into each bucket and the SPL levels are thus higher.

As for how something can be below 0, that is a different issue. 0dB is not equal to zero sound. It is equal to the minimum level of sound a human ear can detect. Better thought of as the threshold bottom for human hearing. Things can be much quieter than we can hear, hence how we can have sound below 0.

I know I'm addressing your comments in random order. Ok so as for distortion, 20-40% is not good. Let me recheck, I missed that.

As for solutions, if this is an apartment and you won't be here long, I would just use a mic isolator and maybe some equipment isolators for now. Buying some puzzle flooring, the mats that have little teeth to lock together, would be good. You can take it with you. I would avoid the light play-foam kids type and look for something a little more rubbery. A sheet of silicone that is a 1/4" thick could work too.

 

[Matthew J Poes]

Oh wait, Duh! You need to take a measurement of your system. Your distortion measurement is meaningless. When you use the newest Beta of REW, it takes a room noise measurement before a sweep. That is what I shared with you. That came from an actual sweep however. Distortion measurements require a fundamental. A room noise measurement has no fundamental. A quiet room has no distortion, its just random noise.

Sorry about that, should have caught that right away.

 

[Pros1]

Oh wait, Duh! You need to take a measurement of your system. Your distortion measurement is meaningless. When you use the newest Beta of REW, it takes a room noise measurement before a sweep. That is what I shared with you. That came from an actual sweep however. Distortion measurements require a fundamental. A room noise measurement has no fundamental. A quiet room has no distortion, its just random noise.

Sorry about that, should have caught that right away.

Oh i get it now, thanks for the explanation, so basically my previous mdat file is invalid.

Today i got the UMIK-1 in the mail, two days earlier than expected so i could take some measurements and i also tried to record with it just to get an idea of its tone.. pretty nice mic, neutral tone, but i still would be happy to hear your opinion about a more expensive mic like Earthworks M23 and if it would make a noticeable difference for basic measurements.
I really like the integration with REW, i just wish it had a lower noise floor (i don't even know how important it is for what i need but the UMIK-1 has much higher noise floor than my other mic.)

I see two things that are different when measuring with UMIK-1 compare to Apogee Mic96k.
1) The values in the distortion tab are much closer to the values in your measurement (although i think there is some graphic bug in REW, or it's something with my GPU, i just can't see the values most of the time)
2) The Spectrogram graph looks different with the UMIK-1, much less Red in the 2nd half of the graph.. any idea about it or what does it mean?

 

[Matthew J Poes]

Oh i get it now, thanks for the explanation, so basically my previous mdat file is invalid.

Today i got the UMIK-1 in the mail, two days earlier than expected so i could take some measurements and i also tried to record with it just to get an idea of its tone.. pretty nice mic, neutral tone, but i still would be happy to hear your opinion about a more expensive mic like Earthworks M23 and if it would make a noticeable difference for basic measurements.
I really like the integration with REW, i just wish it had a lower noise floor (i don't even know how important it is for what i need but the UMIK-1 has much higher noise floor than my other mic.)

I see two things that are different when measuring with UMIK-1 compare to Apogee Mic96k.
1) The values in the distortion tab are much closer to the values in your measurement (although i think there is some graphic bug in REW, or it's something with my GPU, i just can't see the values most of the time)
2) The Spectrogram graph looks different with the UMIK-1, much less Red in the 2nd half of the graph.. any idea about it or what does it mean?

umik-1 and similar USB measurement microphones all have a high noise floor. So high that they are not really appropriate for use in recording. However, for frequency response measurements and most acoustic measurement needs, they are fine. Their main advantage is that they have a nice wide bandwidth response and high SPL capability. They can easily measure up to and beyond 135dB's without any real distortion issues.

For critical acoustic measurement work, these mics are not appropriate. The high noise floor is problematic for measuring room noise floor, transmission loss, etc. When I first starting doing acoustic work professionally, I hadn't realized how limiting this was. I've borrowed, rented, and previously owned numerous high-end measurement devices and can say that for 90% of what you do, the umik-1 will be fine. For that other 10% you can always consider borrowing or renting a better mic, or purchasing something better.

The M23 is a great mic, my good friend James Larson from Audioholics uses this mic, so I have some experience with it. I also previously owned an M30 (Sold it for money when I thought I wouldn't be doing this anymore). The mic is really well made, has the flattest response in the industry (Lab grade flat), and it does, in fact, have a lower noise floor than the umik. However, what I've learned as of late is that it still isn't good enough to use for room noise floor assessment or transmission loss measurements. It's very good, you may find it adequate, but what is needed is a mic that has an equivalent noise floor of around 15dB or less, ideally under 10dB. 1/4" capsules are not quiet enough for this, so you will find a need for higher end mic's with 1/2" or 1" capsules from B&K, ACO Pacific, DPA, etc. The cheapest way to get into this is around $1500 or so. I don't recommend buying this unless you plan to do this professionally or need/want a professional grade measurement microphone for use in recording. These mic's are a bit of a white elephant. Very expensive, delicate, require special care and handling, of very limited use/value compared to cheaper options.

As for your distortion send the mdat and I'll happily take a look when I get a chance. I think you aren't looking at the distortion correctly.

 

[Matthew J Poes]

Oh i get it now, thanks for the explanation, so basically my previous mdat file is invalid.

Today i got the UMIK-1 in the mail, two days earlier than expected so i could take some measurements and i also tried to record with it just to get an idea of its tone.. pretty nice mic, neutral tone, but i still would be happy to hear your opinion about a more expensive mic like Earthworks M23 and if it would make a noticeable difference for basic measurements.
I really like the integration with REW, i just wish it had a lower noise floor (i don't even know how important it is for what i need but the UMIK-1 has much higher noise floor than my other mic.)

I see two things that are different when measuring with UMIK-1 compare to Apogee Mic96k.
1) The values in the distortion tab are much closer to the values in your measurement (although i think there is some graphic bug in REW, or it's something with my GPU, i just can't see the values most of the time)
2) The Spectrogram graph looks different with the UMIK-1, much less Red in the 2nd half of the graph.. any idea about it or what does it mean?

I forgot to respond to this, sorry about that. I have no idea why they are different. I suggest setting it to wavelet mode instead of fft spectrogram. It's more useful. It's just a graphic representation of what is shown in a waterfall plot or even the impulse itself. It's frequency data over time. If they look different, it's because the amplitude response is different. That makes sense, the measurement mic should have a different response than a non-measurement mic. I would suggest overlaying steady-state measurements for each. That should show you, if one has more red, there is probably higher amplitude.

 

[Pros1]

Thanks for the info about the UMIK Vs Earthworks, i think i will just keep the UMIK-1 for now.
I'm more into electronic music so i'm mostly recording synths and don't need a high end microphone for recording, i wanted to know the benefits of a better measurement microphone just for the purpose of measuring my room.

The screenshot of the distortion graph is just to show that i can't see the values next to the checkboxes (Fundamental, 2nd harmonic etc) i don't know if it's a bug or something with my GPU.. this is not the only graphic related issue i'm having with REW.

I'm adding the mdat file of the UMIK-1 measurement, the main thing that doesn't look right is the Spectrogram, in wavelet mode it looks even worse.. if you see something that doesn't look right or i should improve, please let me know.. i'm trying to improve my room sound, i have a new subwoofer on the way and i'm thinking of adding more room acoustics

 

[John Mulcahy]

The screenshot of the distortion graph is just to show that i can't see the values next to the checkboxes (Fundamental, 2nd harmonic etc) i don't know if it's a bug or something with my GPU.. this is not the only graphic related issue i'm having with REW.

There are no values in the legend because your cursor is at 3.48 Hz, below the range of the distortion data.

 

[Pros1]

There are no values in the legend because your cursor is at 3.48 Hz, below the range of the distortion data.

Thank you, i didn't understand what's going on and why sometimes i can see the values and sometimes i can't.