Why did my room frequency change in 2 weeks?

[Sean123]

Hello to all, I'm Sean. I noticed on April 10th earlier this month that when I measured the frequencies in my room I received a measurement of -8.7 as a final result. My room measurement came back flat. However yesterday when measuring again I received a measurement of 31.2. My room is treated with acoustic foam in an isolated corner. Another thing that stopped working was my Java drivers within the preferences. I now have to use Asio drivers. I had a preamp hooked to my audio interface but removed it, and still got the same results. Could this be an issue with any of my equipment like the audio interface or my microphone? I even updated my software to the latest version. I’ll try to attach a picture of the calibration process for extra information about my problem. Thanks

 

[Matthew J Poes]

Hi Sean. It doesn’t look like you have been measuring your rooms response. My guess is that you have some connection problems or issues with how REW is setup.

The dB values you are giving for spl are not accurate and suggest you are just getting some noise.

If you could share some information about your measurement setup we can try to troubleshoot and get things working properly. John may also recognize the problem from the graph and be of more help.

 

[Sean123]

Hi Sean. It doesn’t look like you have been measuring your rooms response. My guess is that you have some connection problems or issues with how REW is setup.

The dB values you are giving for spl are not accurate and suggest you are just getting some noise.

If you could share some information about your measurement setup we can try to troubleshoot and get things working properly. John may also recognize the problem from the graph and be of more help.

Okay, my setup is a Roland octacapture audio interface, an Avalon preamp, and a Nuemann u87 microphone. Here is my results from a few weeks ago measured by REW.

 

[Sean123]

Hi Sean. It doesn’t look like you have been measuring your rooms response. My guess is that you have some connection problems or issues with how REW is setup.

The dB values you are giving for spl are not accurate and suggest you are just getting some noise.

If you could share some information about your measurement setup we can try to troubleshoot and get things working properly. John may also recognize the problem from the graph and be of more help.

Since my frequencies have changed I've unplugged my preamp, and plugged my microphone straight to my Roland Octacapture audio interface. Here are my new results!

 

[Matthew J Poes]

Since my frequencies have changed I've unplugged my preamp, and plugged my microphone straight to my Roland Octacapture audio interface. Here are my new results!



View attachment 7797

Those look more like frequency response results.

For what it’s worth the Neuman mic is not a great option for room response measurements, even if it is an otherwise great microphone. You really want to use a Omni mic with a flat response. The Neuman is a cardioid pattern and a non flat response that is known to roll off the lows and highs (by design of course). Without the right mic any corrections you make to your setup could simply be compensating for the mic.

Otherwise things look good.

 

[Sean123]

Those look more like frequency response results.

For what it’s worth the Neuman mic is not a great option for room response measurements, even if it is an otherwise great microphone. You really want to use a Omni mic with a flat response. The Neuman is a cardioid pattern and a non flat response that is known to roll off the lows and highs (by design of course). Without the right mic any corrections you make to your setup could simply be compensating for the mic.

Otherwise things look good.

Okay, there is an Omni setting on my microphone. Would it help if I changed the setting to Omni then do another test?

 

[Matthew J Poes]

Oh I just noticed the levels. So this is relatively minor but you should probably spend some time optimizing your gain structure and setting the reference SPL level.

If you don’t have an SPL meter but have an iPhone, there is an app from NIOSH that is accurate enough for setting the reference level. You did this in the spl meter module of REW. You basically play pink noise. Measure on your phone app. Take a reading. Measure on REW. Adjust the setting in the spl level setting until it matches your phone. Then your levels will read correctly.

As for gain, you want to get them as high as they can be without causing any clipping. If possible the computer levels should be as high as possible (without clipping) but best to adjust the levels in the analogue domain to avoid the noise floor as much as possible.

 

[Matthew J Poes]

Okay, there is an Omni setting on my microphone. Would it help if I changed the setting to Omni then do another test?

Sort of. These mics aren’t as “omni” as measurement mics. Basically it’s probably fine between 50hz and 6khz. Below 50hz it appears to roll off until it is -10db at 20hz. At 6khz it appears to peak about 2-3dbs and then roll off. At around 15khz it seems to roll off a lot. I would personally use it just for room assessment of decay and low frequencies. I would worry about it’s accuracy above 6khz.

 

[Sean123]

Oh I just noticed the levels. So this is relatively minor but you should probably spend some time optimizing your gain structure and setting the reference SPL level.

If you don’t have an SPL meter but have an iPhone, there is an app from NIOSH that is accurate enough for setting the reference level. You did this in the spl meter module of REW. You basically play pink noise. Measure on your phone app. Take a reading. Measure on REW. Adjust the setting in the spl level setting until it matches your phone. Then your levels will read correctly.

As for gain, you want to get them as high as they can be without causing any clipping. If possible the computer levels should be as high as possible (without clipping) but best to adjust the levels in the analogue domain to avoid the noise floor as much as possible.

I don't own an Iphone, but I do have an Android. I downloaded an app called Niosh and it says National Institute Occupational Safety and Health. There is an area that says Program Courses, and says Monitoring of Noise Exposure, however it doesn't let me click on this an option to measure noise. Maybe this is the wrong app. There is another app called Sound Meter that says it has a calibration feature. Should I try this instead? Thanks

 

[Matthew J Poes]

I don't own an Iphone, but I do have an Android. I downloaded an app called Niosh and it says National Institute Occupational Safety and Health. There is an area that says Program Courses, and says Monitoring of Noise Exposure, however it doesn't let me click on this an option to measure noise. Maybe this is the wrong app. There is another app called Sound Meter that says it has a calibration feature. Should I try this instead? Thanks

Yeah the NIOSH apps spl function only works on iPhones. It’s because all iPhones are the same for a given model and the mics are consistently high quality. Androids are not, lots of variation and only a small number are of good quality. NIOSH did a study and ultimately decided not to support android. I would give the other app a try, you really don’t need to be that precise. Don’t worry about its calibration feature. The Calibration feature is probably to calibrate the app against a reference calibrator.

 

[Sean123]

So the lowest db I recorded was 24db. However, my average db was 28db. So do you want me to put this info into REW? Do you want me to put this info into the calibration feature in REW? Or is this my official sound frequency?

 

[Matthew J Poes]

So the lowest db I recorded was 24db. However, my average db was 28db. So do you want me to put this info into REW? Do you want me to put this info into the calibration feature in REW? Or is this my official sound frequency?

Hi Sean,

I think we still have a little confusion. It’s ok, this is how we learn.

DB is a measure of sound pressure. A single value without reference to frequency doesn’t tell us much. With an SPL meter reading we are actually getting a kind of average over a particular frequency range. Often weighted in a psychoacoustic scale. It just tells you how loud something is.

Now 28db is very quiet. That is the sound level of a very quiet room with no other noises at all in it. In fact it is quieter than most rooms. My guess is that you didn’t have a test tone playing at the time.

What you want to do is follow these instructions.
https://www.roomeqwizard.com/help/help_en-GB/html/inputcal.html

You will play a test tone using the calibration feature that is in those instructions. Take a reading in both the REW spl meter and your phone meter. Then adjust the REW meter until it matches. That will calibrate the levels to be close to accurate.

 

[Sean123]

Okay I

Hi Sean,

I think we still have a little confusion. It’s ok, this is how we learn.

DB is a measure of sound pressure. A single value without reference to frequency doesn’t tell us much. With an SPL meter reading we are actually getting a kind of average over a particular frequency range. Often weighted in a psychoacoustic scale. It just tells you how loud something is.

Now 28db is very quiet. That is the sound level of a very quiet room with no other noises at all in it. In fact it is quieter than most rooms. My guess is that you didn’t have a test tone playing at the time.

What you want to do is follow these instructions.
https://www.roomeqwizard.com/help/help_en-GB/html/inputcal.html

You will play a test tone using the calibration feature that is in those instructions. Take a reading in both the REW spl meter and your phone meter. Then adjust the REW meter until it matches. That will calibrate the levels to be close to accurate.

Okay I'll follow the instructions. Thanks Mr. Poes for all your help. I've readjusted the gain on my preamp like you advised me to and I think I fixed the problem. But I'll run the test regardless. Here is my new results!

 

[Sean123]

Hi Sean,

I think we still have a little confusion. It’s ok, this is how we learn.

DB is a measure of sound pressure. A single value without reference to frequency doesn’t tell us much. With an SPL meter reading we are actually getting a kind of average over a particular frequency range. Often weighted in a psychoacoustic scale. It just tells you how loud something is.

Now 28db is very quiet. That is the sound level of a very quiet room with no other noises at all in it. In fact it is quieter than most rooms. My guess is that you didn’t have a test tone playing at the time.

What you want to do is follow these instructions.
https://www.roomeqwizard.com/help/help_en-GB/html/inputcal.html

You will play a test tone using the calibration feature that is in those instructions. Take a reading in both the REW spl meter and your phone meter. Then adjust the REW meter until it matches. That will calibrate the levels to be close to accurate.

So I read the instructions that you sent to me Mr. Poes and unfortunately I cant follow them because the app that I downloaded called Sound Meter has a db measuring feature. It also has a calibration feature, however you told me to disregard this. I don't know what to do. Perhaps I should seek another app with SPL features. Thanks

 

[Matthew J Poes]

So I read the instructions that you sent to me Mr. Poes and unfortunately I cant follow them because the app that I downloaded called Sound Meter has a db measuring feature. It also has a calibration feature, however you told me to disregard this. I don't know what to do. Perhaps I should seek another app with SPL features. Thanks

DB and SPL are referencing the same thing in this case. You should be ok.

The measurement you posted looks wrong. Not sure what happened but it looks too perfect. The best speaker in an anechoic chamber wouldn’t look that good so I’m guessing something is wrong.

Maybe adjusting the gains wasn’t a good idea. Not sure what has changed.

 

[Sean123]

DB and SPL are referencing the same thing in this case. You should be ok.

The measurement you posted looks wrong. Not sure what happened but it looks too perfect. The best speaker in an anechoic chamber wouldn’t look that good so I’m guessing something is wrong.

Maybe adjusting the gains wasn’t a good idea. Not sure what has changed.

So Mr. Poes, I played some music and tested the SBL on my microphone inside of my vocal booth and the measurement was 33db. When I put my phone inside of the vocal booth with the music on the measurement average was 66db. I then followed the instructions and did a calibration. I put the data from my phone into the REW calibration at 66db. Then I turned the music off, and did a measurement test. Here's a picture of my result. Thanks

 

[Matthew J Poes]

that doesn't look right either. The SPL level that is shown is 121 db's, which is off the charts, but even so, I can see enough in that pic to tell that you seem to not be capturing sound from your speakers. Something has gone wrong again. I would maybe play with the settings again and try unplugging and re-plugging everything. You were getting measurements that seemed plausible earlier.

 

[Sean123]

that doesn't look right either. The SPL level that is shown is 121 db's, which is off the charts, but even so, I can see enough in that pic to tell that you seem to not be capturing sound from your speakers. Something has gone wrong again. I would maybe play with the settings again and try unplugging and re-plugging everything. You were getting measurements that seemed plausible earlier.

I do want to mention that when I did my calibration I put "Use an external source." Is this the correct thing to do when testing the sound on from microphone?

 

[Matthew J Poes]

You can do that. Given that you used music, its ok. It is better to use a calibration signal because its more steady in nature.

Can you redo it using the speaker calibration signal just to be sure that REW is sending sound to your speakers correctly? That seems to be part of the issue. I'm not clear how you are getting measurements without errors that look like this.

 

[Matthew J Poes]

Sean what's the chance you are located somewhere in the Chicago area? I'd offer to help you out if you were close. It's hard for me to troubleshoot like this.

 

[Sean123]

that doesn't look right either. The SPL level that is shown is 121 db's, which is off the charts, but even so, I can see enough in that pic to tell that you seem to not be capturing sound from your speakers. Something has gone wrong again. I would maybe play with the settings again and try unplugging and re-plugging everything. You were getting measurements that seemed plausible earlier.

So Mr. Poes I did a calibration with music playing using the " Use REW speaker cal signal option". The signal came back at 70. I then ran a measurement test and I got the same results. Only this time it went from ending in a 9 to an 8. Here's my results. I do have my preferences input set to my audio interface and my microphone. Was this the correct thing to do or should I have changed any settings within the preferences menu option? Thanks

 

[Sean123]

Sean what's the chance you are located somewhere in the Chicago area? I'd offer to help you out if you were close. It's hard for me to troubleshoot like this.

So I just saw this message about my location Mr. Poes. Unfortunately, I'm not located anywhere near Chicago;I'm in Atlanta, GA. Thanks

 

[Matthew J Poes]

So Mr. Poes I did a calibration with music playing using the " Use REW speaker cal signal option". The signal came back at 70. I then ran a measurement test and I got the same results. Only this time it went from ending in a 9 to an 8. Here's my results. I do have my preferences input set to my audio interface and my microphone. Was this the correct thing to do or should I have changed any settings within the preferences menu option? Thanks


View attachment 7800

Yeah like I mentioned before, something is off here. This is not a measurement of your speakers. It is just too flat. Not sure what is going on. When you run a sweep are you hearing something come from the speakers?