Acoustic treatment and REW measurements for music studio.

[circleHstudios]

Hello everyone, brand new to the group, thanks for having me and thank you for your patience and any advice that i may receive. This will be a 2 part question.

1. I have a 20'L X 24'W X 8'H music studio/living room/kitchen (mostly studio!) with sheetrock and hardwood floors that I am trying to acoustically treat. I am getting decent recordings from my vocals and instrument takes but when I try to record drums they sound TERRIBLE (could be my mics, they are a Shure PGA 7 piece set, offset with some SM57's and an NT1A, but kind of irrelevant for this post) I have (12) twelve sheets of 4'x8'x2" Corning 703 fiberglass insulation that I'm making my acoustic panels from. I currently have (7) seven 2' x 4' x 6" already made and I am planning on using (2) two of these panels directly behind the drum set, stacked on top of one another (8' tall) for a bass trap as well as the adjacent corner for (4) four total panels. I am also planning on using (2) two 2' x 4' x 6" panels as clouds above the drum kit hung approximately 4" from the ceiling and (1) one 2' x 4' x 6" panel above my listening position towards the back center of the room hung 4" from ceiling. With the remaining insulation i would like to make (12) twelve 2' x 4' x 4" panels to hang on the walls and maybe ceiling around the room. Having said this, I have a multitude of questions! My first question is this:
   1. Where should my remaining (12) twelve 4' x 2' x 4" panels go. I feel like I am lacking clouds. Is there a ratio for cloud to wall coverings? I have heard of 30% coverage of area will do, but what does that mean, 30% of all surface area in the room (20' x 24' x 8' = 500 sq ft/ 355 sq ft without floor). I will do first reflection points for my drum kit but have no clue what the best plan of attack is for the rest of the room. And do clouds need to be 6"? (Buying more acoustic material is out of the question right now)
2. Second question has to do with REW. I set my RTA mic up in front of my drum kit and put my studio monitor on my drum throne and ran through the REW test to get results. Can someone PLEASE help me figure out what I'm looking at and if I need to do something different in the testing.
   1. REW FILE BELOW
   2. https://drive.google.com/file/d/10TSE_kORv_VsDU__ufaBGC27DT9EnAKp/view?usp=drive_link

I posted too before i meant to and its too late! I'll update tomorrow.
Thank You.

Nate,
Circle H Music Studios



Sample: https://on.soundcloud.com/xK0zkSfhhc2mFekxPu
(Please see PDF for room layout)

 

[Keith_W]

First question: what mic did you use to take the measurement? Can you confirm that your RTA mic is an omnidirectional condenser microphone? I can see from your MDAT that no calibration file was loaded.

Your ETC looks absolutely horrendous, early reflections all over the place. The RT60 is similarly pretty bad.

However, there are a lot of confounding factors that makes me wonder if we are looking at a whole bunch of artefacts.

1. Unknown microphone used for measurement, with no calibration file.
2. Unknown test signal.
3. Presence of drum kit in the studio when taking measurements. These things have their own resonant frequency and may contribute to ringing at certain frequencies. The measurement does not know any better, a resonance is a resonance, and you don't know if it's a room resonance or your musical instrument.
4. Placement of microphone at the drum stool in the middle of all those drums will worsen early reflections and make resonances louder.

If you want to gain a proper understanding of your room's acoustics, do this:

1. Omnidirectional condenser microphone, pointing UPWARDS (NOT towards the floor as you have in your photo), with the appropriate 90deg calibration file loaded. And definitely DO NOT PLACE your microphone in the middle of a forest of equipment and musical instruments.
2. State what test signal you are using. Either pink noise or a sine wave sweep.
3. Remove the drum kit from the studio when taking measurements.
4. Take the measurement at multiple points in your studio. I would suggest you create a 1m x 1m grid and measure at every point in the grid.

 

[alex-z]

OC 703 is relatively dense for bass trap usage. I would switch to something with a flow resistivity of 10000 Pa*s/m² or less, that will give more absorption in the mid-bass region. Also consider doubling the thickness for better low frequency management.

Looking at your Filtered IR data there is a strong reflection 4ms after the impulse peak. That corresponds to 137cm distance, whatever surface is that distance from your drum seat needs absorption. You want the panels spaced out so that your decay rate is gradual with no sharp peaks.

6" ceiling cloud is good but 4" is still decent. Definitely increase your cloud coverage.

Keith mentioned this already, remove the drums from the room for more accurate measurements.

I will add that standard speakers are not omni-directional at mid and high frequencies. Rotate your speaker and repeat the measurements, that will help differentiate speaker radiation pattern vs room acoustics. Special dodecahedron speakers do exist but their cost is "ask for a quote".

The technical design challenges of the omnidirectional sound source

The technical design challenges of the omnidirectional sound source

www.hbkworld.com

Recording technique is a huge part of drum quality. Don't just near-field everything, have an overhead/room mic so you can experiment with blending.

 

[circleHstudios]

First question: what mic did you use to take the measurement? Can you confirm that your RTA mic is an omnidirectional condenser microphone? I can see from your MDAT that no calibration file was loaded.

View attachment 514738

Your ETC looks absolutely horrendous, early reflections all over the place. The RT60 is similarly pretty bad.

However, there are a lot of confounding factors that makes me wonder if we are looking at a whole bunch of artefacts.

1. Unknown microphone used for measurement, with no calibration file.

• I used the RTA microphone that came with my DBX Driverack PA2 speaker management system. It did not come with a calibration file and the quick google search i did, did not come up with one.
  • Is there a generic calibration file that i could use for this microphone or any suggestions on that front?

2. Unknown test signal.

• I used my studio KRK Rokit 5 studio monitor placed on my drum throne (rather low, you know, seat level) with the RTA microphone pointed at the ceiling and located at the front of the drum kit (more on this in the reply to 3 & 4). The actual signal type I cant for the life of me find in REW, any tips on where to find this in REW? It sounds like the same signal my DBX Driverack puts out, the signal starts low HZ and kind of sweeps up to the high HZ. I'm thinking it is the sine wave, i did not mess with any settings in REW so it would be whatever is the default REW is set to. I do know it put out a signal in the 20-20,000 HZ range. Is this clear as mud?
• I have a couple additional questions on speaker and mic placement.
  • microphone height - I have 8' ceilings so should my microphone tip be at the halfway point between ceiling and floor (4') or what is the best position as far as height goes.
  • monitor placement - this i have no clue as to where to place the monitor/monitors. I have my listening position as noted as "desk" on my very rough sketch provided, speakers facing away from the live recording area. Can i just leave my monitors there and turn them around so they will be facing the RTA microphone so the sound source is a constant if I'm moving my microphone around in a 1m x 1m grid as you suggested. Doers it matter if i use a single speaker or both? Any suggestions here are greatly appreciated.

3. Presence of drum kit in the studio when taking measurements. These things have their own resonant frequency and may contribute to ringing at certain frequencies. The measurement does not know any better, a resonance is a resonance, and you don't know if it's a room resonance or your musical instrument.
4. Placement of microphone at the drum stool in the middle of all those drums will worsen early reflections and make resonances louder.

• Yes, both microphone and speaker were pretty much in the middle of my drum kit. This is kind of laughable and embarrassing. In hindsight I should have known this!

If you want to gain a proper understanding of your room's acoustics, do this:

1. Omnidirectional condenser microphone, pointing UPWARDS (NOT towards the floor as you have in your photo), with the appropriate 90deg calibration file loaded. And definitely DO NOT PLACE your microphone in the middle of a forest of equipment and musical instruments.
2. State what test signal you are using. Either pink noise or a sine wave sweep.
3. Remove the drum kit from the studio when taking measurements.
4. Take the measurement at multiple points in your studio. I would suggest you create a 1m x 1m grid and measure at every point in the grid.

• Just for a little more information, I have the RTA microphone into input 1 of my Focusrite 18i20 first gen. with the timing reference output in output 3 and loopback input into input 3. I read somewhere that the Focusrite 1st gen does not support loopback but I can't imagine that to be the case in this scenario.
  • do you have any knowledge on this?
  • also in REW it has some spots for calibration files for my focusrite unit. Do i need to have those?
• 

Keith, thank you VERY much for your reply and hopefully I answered any questions you had. Looking forward to your response.

On a side note, i know this will be a little vague, I found, from my memory, a youtube video ( i could be wrong on this) a while back that I was watching that was over an hour long with an older gentleman talking on the ends and outs of REW that i found very helpful, i thought i bookmarked the video but apparently did not, and can not find the video again with my searches. Would you have any good resources for REW usage and comprehension.

 

[circleHstudios]

OC 703 is relatively dense for bass trap usage. I would switch to something with a flow resistivity of 10000 Pa*s/m² or less, that will give more absorption in the mid-bass region. Also consider doubling the thickness for better low frequency management.

Looking at your Filtered IR data there is a strong reflection 4ms after the impulse peak. That corresponds to 137cm distance, whatever surface is that distance from your drum seat needs absorption. You want the panels spaced out so that your decay rate is gradual with no sharp peaks. You

6" ceiling cloud is good but 4" is still decent. Definitely increase your cloud coverage.

Keith mentioned this already, remove the drums from the room for more accurate measurements.

I will add that standard speakers are not omni-directional at mid and high frequencies. Rotate your speaker and repeat the measurements, that will help differentiate speaker radiation pattern vs room acoustics. Special dodecahedron speakers do exist but their cost is "ask for a quote".

The technical design challenges of the omnidirectional sound source

The technical design challenges of the omnidirectional sound source

www.hbkworld.com

Recording technique is a huge part of drum quality. Don't just near-field everything, have an overhead/room mic so you can experiment with blending.

Alex, thank you very much for the reply. Unfortunately at the moment, what i already have is what i will be working with for the time being. I think the drum kit was pretty instrumental in the "horrendous" reflections. In hindsight it is pretty laughable and am a bit embarrassed about that oversight. I will take the kit out of the room and re measure and post once Keith gets back to me with some remaining questions I had. Thank you.

 

[Keith_W]

I used the RTA microphone that came with my DBX Driverack PA2 speaker management system. It did not come with a calibration file and the quick google search i did, did not come up with one.
Is there a generic calibration file that i could use for this microphone or any suggestions on that front?

A bit of googling says that your microphone is the RTA-M microphone:

This is an omnidirectional condenser microphone. It looks like a Behringer ECM8000 with a different coat of paint

Yes, this microphone is suitable for the measurements you want to take. And no, you can not use a generic calibration file. The reason cal files are needed is because mics are like all manufactured products - there is variability. For sensitive test equipment like a microphone, ideally every mic that comes out of the factory should be tested and individually calibrated. But for cheap microphones, they don't bother. Some of them are batch calibrated, for every 50-100 mics that roll off the line, 1 or 2 are tested and the calibration is available for that entire batch. You can test this by entering two different serial numbers, if the calibration file is the same, then your mic is not individually calibrated. In fact strongly suspect that if you spend anything less than $300 it's likely not individually calibrated.

Actually I should have been more careful with my previous reply. Calibration is needed if you want to look at the frequency response. The errors are greatest at both extremes of the audible range. But for what you want to do - i.e. look at decay measurements, calibration is probably not necessary.

I'll say it again so that it's clear: calibration is a requirement for some types of measurements. For example, if you are a speaker designer, or for DSP, where it is vital to take accurate measurements of the frequency response. But for what you want to do, it is not required.

I used my studio KRK Rokit 5 studio monitor placed on my drum throne (rather low, you know, seat level) with the RTA microphone pointed at the ceiling and located at the front of the drum kit (more on this in the reply to 3 & 4). The actual signal type I cant for the life of me find in REW, any tips on where to find this in REW? It sounds like the same signal my DBX Driverack puts out, the signal starts low HZ and kind of sweeps up to the high HZ. I'm thinking it is the sine wave, i did not mess with any settings in REW so it would be whatever is the default REW is set to. I do know it put out a signal in the 20-20,000 HZ range. Is this clear as mud?

You are describing a sine wave sweep. This is a suitable test signal. With any measurement, you want a good signal to noise ratio (SNR). I would choose a longer sweep, 512k or 1M. And given you are concerned about your room's acoustics, you should take a measurement of your room's noise floor. See the eBook in my link for instructions on how to do that. If you want to measure your room's noise floor, you will need an SPL meter.

I have a couple additional questions on speaker and mic placement.

• microphone height - I have 8' ceilings so should my microphone tip be at the halfway point between ceiling and floor (4') or what is the best position as far as height goes.
• monitor placement - this i have no clue as to where to place the monitor/monitors. I have my listening position as noted as "desk" on my very rough sketch provided, speakers facing away from the live recording area. Can i just leave my monitors there and turn them around so they will be facing the RTA microphone so the sound source is a constant if I'm moving my microphone around in a 1m x 1m grid as you suggested. Doers it matter if i use a single speaker or both? Any suggestions here are greatly appreciated.

Think a bit about what your goal is. You want to record a drum kit, which is placed somewhere in your room, using a microphone, placed somewhere else in your room. Your drum kit is an omnidirectional sound source, and your speakers aren't. So your speakers need to be placed where your drum kit would go. To get your speakers to approximate an omnidirectional source, you can turn your speakers away so that they are firing into the wall.

Single speaker vs. both - I would probably use both and place them back to back and facing away from each other, just to get more volume and to approximate an omni source.

The relevant measurement is RT60, or "reverberation time for sound to decay by 60dB". Firstly, 60dB decay is not achievable by most of us because our speakers don't go that loud. So the RT60 is actually a 20dB or 30dB decay (T20, T30 respectively) and extrapolated to 60dB by simple multiplication. The other problematic requirement is "reverberation". A reverberant field is defined as thousands of room modes overlapping each other, and it should be the same no matter where it is measured in the room. The problem is - small rooms don't do that. Long wavelengths form room modes, and short wavelengths form specular reflections. This is why we measure from multiple points in the room - if you overlay all these measurements, you will see where the curves diverge, and that will tell you if you are looking at a specular reflection / room mode, or a reverberant field.

Just for a little more information, I have the RTA microphone into input 1 of my Focusrite 18i20 first gen. with the timing reference output in output 3 and loopback input into input 3. I read somewhere that the Focusrite 1st gen does not support loopback but I can't imagine that to be the case in this scenario. do you have any knowledge on this?

You do not need loopback if you are measuring your room's acoustic properties. Loopback is for timing measurements, e.g. if you are trying to time multiple speakers and subwoofers. And of course your Focusrite supports loopback, all you need to do is plug a DAC output into the mic input using a suitable cable. Whether it supports software loopback without a cable is another story, you will need to consult your manual for that one.

• also in REW it has some spots for calibration files for my focusrite unit. Do i need to have those?

Ideally yes, but in practice no. It is pretty safe to assume that your Focusrite has a flat frequency response.

On a side note, i know this will be a little vague, I found, from my memory, a youtube video ( i could be wrong on this) a while back that I was watching that was over an hour long with an older gentleman talking on the ends and outs of REW that i found very helpful, i thought i bookmarked the video but apparently did not, and can not find the video again with my searches. Would you have any good resources for REW usage and comprehension.

Yes, see the eBook in my signature. It's free.

 

[circleHstudios]

A bit of googling says that your microphone is the RTA-M microphone:

View attachment 514896

This is an omnidirectional condenser microphone. It looks like a Behringer ECM8000 with a different coat of paint

Yes, this microphone is suitable for the measurements you want to take. And no, you can not use a generic calibration file. The reason cal files are needed is because mics are like all manufactured products - there is variability. For sensitive test equipment like a microphone, ideally every mic that comes out of the factory should be tested and individually calibrated. But for cheap microphones, they don't bother. Some of them are batch calibrated, for every 50-100 mics that roll off the line, 1 or 2 are tested and the calibration is available for that entire batch. You can test this by entering two different serial numbers, if the calibration file is the same, then your mic is not individually calibrated. In fact strongly suspect that if you spend anything less than $300 it's likely not individually calibrated.

Actually I should have been more careful with my previous reply. Calibration is needed if you want to look at the frequency response. The errors are greatest at both extremes of the audible range. But for what you want to do - i.e. look at decay measurements, calibration is probably not necessary.

I'll say it again so that it's clear: calibration is a requirement for some types of measurements. For example, if you are a speaker designer, or for DSP, where it is vital to take accurate measurements of the frequency response. But for what you want to do, it is not required.



You are describing a sine wave sweep. This is a suitable test signal. With any measurement, you want a good signal to noise ratio (SNR). I would choose a longer sweep, 512k or 1M. And given you are concerned about your room's acoustics, you should take a measurement of your room's noise floor. See the eBook in my link for instructions on how to do that. If you want to measure your room's noise floor, you will need an SPL meter.



Think a bit about what your goal is. You want to record a drum kit, which is placed somewhere in your room, using a microphone, placed somewhere else in your room. Your drum kit is an omnidirectional sound source, and your speakers aren't. So your speakers need to be placed where your drum kit would go. To get your speakers to approximate an omnidirectional source, you can turn your speakers away so that they are firing into the wall.

Single speaker vs. both - I would probably use both and place them back to back and facing away from each other, just to get more volume and to approximate an omni source.

The relevant measurement is RT60, or "reverberation time for sound to decay by 60dB". Firstly, 60dB decay is not achievable by most of us because our speakers don't go that loud. So the RT60 is actually a 20dB or 30dB decay (T20, T30 respectively) and extrapolated to 60dB by simple multiplication. The other problematic requirement is "reverberation". A reverberant field is defined as thousands of room modes overlapping each other, and it should be the same no matter where it is measured in the room. The problem is - small rooms don't do that. Long wavelengths form room modes, and short wavelengths form specular reflections. This is why we measure from multiple points in the room - if you overlay all these measurements, you will see where the curves diverge, and that will tell you if you are looking at a specular reflection / room mode, or a reverberant field.



You do not need loopback if you are measuring your room's acoustic properties. Loopback is for timing measurements, e.g. if you are trying to time multiple speakers and subwoofers. And of course your Focusrite supports loopback, all you need to do is plug a DAC output into the mic input using a suitable cable. Whether it supports software loopback without a cable is another story, you will need to consult your manual for that one.




Ideally yes, but in practice no. It is pretty safe to assume that your Focusrite has a flat frequency response.



Yes, see the eBook in my signature. It's free.

You sir are amazing. Thank you very much.

 

[circleHstudios]

A bit of googling says that your microphone is the RTA-M microphone:

View attachment 514896

This is an omnidirectional condenser microphone. It looks like a Behringer ECM8000 with a different coat of paint

Yes, this microphone is suitable for the measurements you want to take. And no, you can not use a generic calibration file. The reason cal files are needed is because mics are like all manufactured products - there is variability. For sensitive test equipment like a microphone, ideally every mic that comes out of the factory should be tested and individually calibrated. But for cheap microphones, they don't bother. Some of them are batch calibrated, for every 50-100 mics that roll off the line, 1 or 2 are tested and the calibration is available for that entire batch. You can test this by entering two different serial numbers, if the calibration file is the same, then your mic is not individually calibrated. In fact strongly suspect that if you spend anything less than $300 it's likely not individually calibrated.

Actually I should have been more careful with my previous reply. Calibration is needed if you want to look at the frequency response. The errors are greatest at both extremes of the audible range. But for what you want to do - i.e. look at decay measurements, calibration is probably not necessary.

I'll say it again so that it's clear: calibration is a requirement for some types of measurements. For example, if you are a speaker designer, or for DSP, where it is vital to take accurate measurements of the frequency response. But for what you want to do, it is not required.



You are describing a sine wave sweep. This is a suitable test signal. With any measurement, you want a good signal to noise ratio (SNR). I would choose a longer sweep, 512k or 1M. And given you are concerned about your room's acoustics, you should take a measurement of your room's noise floor. See the eBook in my link for instructions on how to do that. If you want to measure your room's noise floor, you will need an SPL meter.



Think a bit about what your goal is. You want to record a drum kit, which is placed somewhere in your room, using a microphone, placed somewhere else in your room. Your drum kit is an omnidirectional sound source, and your speakers aren't. So your speakers need to be placed where your drum kit would go. To get your speakers to approximate an omnidirectional source, you can turn your speakers away so that they are firing into the wall.

Single speaker vs. both - I would probably use both and place them back to back and facing away from each other, just to get more volume and to approximate an omni source.

The relevant measurement is RT60, or "reverberation time for sound to decay by 60dB". Firstly, 60dB decay is not achievable by most of us because our speakers don't go that loud. So the RT60 is actually a 20dB or 30dB decay (T20, T30 respectively) and extrapolated to 60dB by simple multiplication. The other problematic requirement is "reverberation". A reverberant field is defined as thousands of room modes overlapping each other, and it should be the same no matter where it is measured in the room. The problem is - small rooms don't do that. Long wavelengths form room modes, and short wavelengths form specular reflections. This is why we measure from multiple points in the room - if you overlay all these measurements, you will see where the curves diverge, and that will tell you if you are looking at a specular reflection / room mode, or a reverberant field.



You do not need loopback if you are measuring your room's acoustic properties. Loopback is for timing measurements, e.g. if you are trying to time multiple speakers and subwoofers. And of course your Focusrite supports loopback, all you need to do is plug a DAC output into the mic input using a suitable cable. Whether it supports software loopback without a cable is another story, you will need to consult your manual for that one.




Ideally yes, but in practice no. It is pretty safe to assume that your Focusrite has a flat frequency response.



Yes, see the eBook in my signature. It's free.

I have removed the drum kit and ran a couple more tests. Please let me know your thoughts.

https://drive.google.com/file/d/1T6N2K_S_664KiXLgwMcTCM0_pQYS6V6_/view?usp=drive_link

Thank You.

 

[Keith_W]

You have included 3 measurements, and you did not label what the third measurement was, or explain where the mic was positioned. All I know is that it's from a different position in the room since the pattern of reflections is different. I get really annoyed when I am forced to guess, it would have taken you two seconds to provide that information! Anyway let's take a look.

This is a comparison of "Pos 1" with and without panels. I am guessing it's where the speakers/mics are as set up in your photo. You can see that the pattern of reflections is the same, but the green measurement (with panels) shows that the reflections are significantly attenuated. In particular, that specular reflection at 4.8ms is almost gone.

This is the RT60 with the indicator bars calculated for your room volume (108m^3) and intended application (target RT60 for studios should be 250ms or less). Overall your RT60 is too high, it's about 550ms - 600ms.

You did not take enough measurements from different positions, so it is difficult to see where the RT60 measurements start to diverge. Did you read that eBook?

That large spike at 1.5kHz in the green measurement is external noise and should be ignored. This is the evidence:

You will need a lot of absorbers to bring your RT60 down to 250ms or less. I am not in the business of recommending acoustic treatment products, so you need to speak to someone who has expertise in this area.