John Sayers' Recording Studio Design Forum Archive

Hi friends,

I've built a home studio and bought a Radioshack Analog Sound Level Meter to see the frequency response on it.

Here's my studio:


The room was made with the following measures:
Lenght = 3,07 meters
Widht = 2,52 mt
Height = 2,2 mt

Here's a graphic of the expected modes of my room:


The roof used to be 2,80mt, then I've made it be 2,20 using this product (http://www.isover.com.br/isover/por_lin ... enu=boreal (this is rockwoll in one side and pvc in the other side), i've put foam in the reflexion points and DIY basstraps on the corners (1.20mt) + foam over them (1m). On the ground there is no treatment, its Slate over a solid concrete ground. I live in a house, not apartment.

Here was my procedure to make the test:
0 - I've put the SPL Meter in my MIC stand in the recommended 1m of the nearfield monitors (KRK Rokit Passive on a Alesis RA-100), the place where my ears would be.
1 - Opened Protools and with the tone generator i've selected Noise, when the noise was exactly 90db I was sure that I could change to the sine to proceed with the test.
2 - In the sine generator I began to analyse through every 1/3 octave, but then, since there were a lot of valleys in the analysys, I began to go further in the HZ.
3 - Input it in Excel and made a graphic of db x freq.



From 25hz to 63hz it was having an acceptable curve, then, suddenly in 80hz there was a -10db valley and from there a lot of flutiation over the different hz, not a progressive curve.
It also presented bad results under 25hz.

In the next month, i'll buy a friend's Behringer B2030A + Subwoofer B2092A, because the KRK is good, but the sub part is essential in my productions, although I don't have a quite big room to hear it (I produce Rap and Electronics).

I would like to ask for some suggestions to have a more acceptable curve on the 20hz-500hz range, I'm worried about it.

What I've thought myself that would be a great idea:
a) Build a Helmholtz DYI (But don't really know which frequency to tune it)
b) Build more 2 basstraps like these: http://www.teresaudio.com/haven/traps/traps.html
b) Eq the monitors with Behringer Feedback Destroyer Pro DSP 1124P

Please, give me some advices and I'll be ever thankful.

If you need additional details, please let me know.

And I'm really sorry for my bad english.

Hi "deck-o", and welcome!

The roof used to be 2,80mt, then I've made it be 2,20

Why did you do that? You did not need to. If you would have kept the original ceiling, your ratio would have been 1 : 1.11 : 1.21 (swapping height for width), which is close to Sepmeyer's first ratio. As it is now, you have exactly Sepmeyer's first ratio, but you sacrificed one cubic meter of room volume to get there, and reduced your ceiling height, making it very low. Also, the product you used is a micro-perforated panel, so it isn't even actually doing what you think it is doing: You have just created a partially reflective, partially tuned surface one meter above your head.

DIY basstraps on the corners (1.20mt) + foam over them

How did you build those "bass traps", and what type of "foam" did you use?

I've put the SPL Meter in my MIC stand in the recommended 1m of the nearfield monitors (KRK Rokit Passive on a Alesis RA-100), the place where my ears would be.

Your speakers are not set up correctly: they are both facing directly down the room, parallel to the walls, and they are much too close together. They are also on a shelf right now. What I would suggest that you do to fix that is to remove the shelf, and place the speakers on very heavy speaker stands, in the correct position and aimed correctly. You can make the speaker stands very cheaply, by stacking up some concrete blocks, or by using a hollow metal stand filled with sand. The speakers must be set up so that the acoustic axis is 1.2m above the floor, and aimed directly at your ears: So the speakers will be set much further apart than you have them now, and angled inwards so that their acoustic axes intersect a few cm behind your head, at an angle of at least 60° (and maybe as much as 90°). The distance between the speakers should be about the same as the distance from each speaker to your head. Page 7 of the manual for your speakers shows you how to do that.

and with the tone generator i've selected Noise,

White noise, or pink noise? You MUST use pink noise for the test, never white noise.

when the noise was exactly 90db I was sure that I could change to the sine to proceed with the test.

The level should be 85 dB, but 90 is OK too. No problem there. But why did you switch to sine wave testing? That isn't realistic. Pink noise is the correct way of testing: Since sweeps only tell you about individual frequencies, one at a time, but they don't tell you about the interaction of the room as a whole with all frequencies at once.

In the sine generator I began to analyse through every 1/3 octave, but then, since there were a lot of valleys in the analysys, I began to go further in the HZ.

1/3 octave is no use. For the low end of the spectrum, you need to analyze with precision of 1 Hz, and you can allow that to get broader as you get to higher frequencies (maybe 30 Hz at 300 Hz.).

Here was my procedure to make the test:

It would be much better, much more accurate, and much more complete to just download the REW program from Home Theatre Shack ( http://www.hometheatershack.com ). It is free, easy to use, very accurate, and does many more types of analyses, such as waterfall plots, which are very useful for diagnosing modal problems.

nearfield monitors (KRK Rokit Passive

You mean the R6? I guess that must be it, since it is the only passive Rokit. If so, the frequency response is only down to 50 Hz, so you can ignore everything below that on your graph (the entire left half of your graph is meaningless).

From 25hz to 63hz it was having an acceptable curve,

Actually, that isn't acceptable at all! IT means nothing, since the response of the KRK Rokit R6 is not linear below 50 Hz (and probably more like 60 Hz), plus your sound level meter is not very linear that low either, and besides it is extremely hard to get accurate measurements of low frequency sound, especially in such a tiny room. You can ignore everything below about 60 Hz on there, and even then that isn't accurate.

suddenly in 80hz there was a -10db valley

That might be your 1,0,0 axial (first vertical mode), but more likely to just be a reflection from the stuff you put on your ceiling.

and from there a lot of flutiation over the different hz, not a progressive curve.

If that curve is accurate, then it isn't too bad at all! You have +/- 10 dB from 60 Hz up to 1 kHz, which is reasonable for a partially untreated small room like that.

It also presented bad results under 25hz.

That is totally meaningless. The readings down that low mean zero. Zilch, Nada. Nothing at all. Ignore everything below 60 Hz for those speakers and that method of measuring response.

In the next month, i'll buy a friend's Behringer B2030A + Subwoofer B2092A

Better. That goes down to around 35 Hz. But! Your room does not even support any modes below 56 Hz! So there's no need to even bother testing much below that.

but the sub part is essential in my productions, although I don't have a quite big room to hear it

You will still hear it, if your speakers can produce it! As long as your room is treated suitably, you will be able to hear all of the frequencies that the speaker can produce. Why do you think you would not be able to hear low frequencies?

What I've thought myself that would be a great idea:

Before you do any of that, you need to set up your speakers correctly, and use REW to do the testing. We need to see the waterfall plots to figure out what is going on. So set up your speakers correctly, get a good measurement mic (or any omni mic with flat response), and test again. When you do, make a note of the exact position of the measurement mic, so that you can put it in the exact same place again for future readings, after you treat the room. The mic should go in the same position as your head will be, pointing upwards at an angle of roughly 30°, but measure EXACTLY where it is, accurate to 5mm. You MUST put it back in the same position each time you measure, or the readings will not be valid. Also, make separate measurements for the left and right speakers, one at a time. Do not try to measure both at once.

Build a Helmholtz DYI

Probably not necessary. Hard to do right, and hard to position correctly. That would be the absolute last resort, after doing all the other treatment first.

Build more 2 basstraps

probably, but not tube traps. More like superchunks or maybe hangers. But first we need to know how you built your existing traps, and what that foam is. Also, what other treatment do you have in the room? For example, how have you treated the back wall?

Eq the monitors with Behringer Feedback Destroyer Pro DSP 1124P

No! Never! That will not work. It CANNOT work, since modal issues are problems in the time domain, not just the frequency domain. EQ cannot do ANYTHING to the signal to fix modal issues after the tone has stopped coning out of the speaker but while it is still bouncing around in the room! That's the problem. A mode carries on "ringing" for many milliseconds after the speaker has stopped, and during that time there is NOTHING that EQ can do to fix it. Modal issues can ONLY be fixed with suitable acoustic treatment, never with EQ. Sure, with EQ you can cut some problem frequencies a bit, but what do you do with the nulls? A null is always a null, and no amount of power increase will fill it. EQ will also only work for ONE position in the room, not for any other, and distorts the response anyway: it gives you a false representation of your music, which is totally wrong. That's why you built a studio in the first place; to get an ACCURATE representation of your music! EQ messes that up. Bad idea.

On the ground there is no treatment, its Slate over a solid concrete ground.

Good! Do not change that! It is a really good acoustic surface..

So I'd suggest that first you need to fix your speaker mounting problem (never on shelves, always on stands or in soffits, acoustic axis at 1.2m above the floor), then fix your speaker geometry problem (further apart but NOT in the corners, angled so that they are aimed directly at your ears, with the acoustic axes intersecting at 60° to 90° a few cm behind your head), the re-measure with REW and a good mic. Post those results here so that we can see where your REAL problems are.

And also post the details of how you built your bass traps, and what foam you have used there.


- Stuart -

Soundman2020 wrote:Hi "deck-o", and welcome!

The roof used to be 2,80mt, then I've made it be 2,20

Why did you do that? You did not need to. If you would have kept the original ceiling, your ratio would have been 1 : 1.11 : 1.21 (swapping height for width), which is close to Sepmeyer's first ratio. As it is now, you have exactly Sepmeyer's first ratio, but you sacrificed one cubic meter of room volume to get there, and reduced your ceiling height, making it very low. Also, the product you used is a micro-perforated panel, so it isn't even actually doing what you think it is doing: You have just created a partially reflective, partially tuned surface one meter above your head.

I tried the best to get the lowest modes in the room, and the way I found was to lower the roof. It gave me less modes than it would if it was 2,80... i`ve got less reflexions due to the rockwoll, but its micro-perforated, so it's is working also as a tuned surface.

DIY basstraps on the corners (1.20mt) + foam over them

How did you build those "bass traps", and what type of "foam" did you use?

They were done with 50mm Rockwoll, as this pictures:







That is the way "manual" i've used to build them (the grey one is the wood and the yellow one is the rockwoool:


The foam over the bass traps are from a bed mattress that i'm not using anymore (I'm in such a low budget, that i'm even ashamed to use this as an alternative to another diy basstraps... they have a high density and are about 12,5cm of height.

I've put the SPL Meter in my MIC stand in the recommended 1m of the nearfield monitors (KRK Rokit Passive on a Alesis RA-100), the place where my ears would be.

Your speakers are not set up correctly: they are both facing directly down the room, parallel to the walls, and they are much too close together. They are also on a shelf right now. What I would suggest that you do to fix that is to remove the shelf, and place the speakers on very heavy speaker stands, in the correct position and aimed correctly. You can make the speaker stands very cheaply, by stacking up some concrete blocks, or by using a hollow metal stand filled with sand. The speakers must be set up so that the acoustic axis is 1.2m above the floor, and aimed directly at your ears: So the speakers will be set much further apart than you have them now, and angled inwards so that their acoustic axes intersect a few cm behind your head, at an angle of at least 60° (and maybe as much as 90°). The distance between the speakers should be about the same as the distance from each speaker to your head. Page 7 of the manual for your speakers shows you how to do that.

I'm planning to build this speakers stands: http://onefallinghope.com/2009/02/03/diy-speaker-stands

and with the tone generator i've selected Noise,

White noise, or pink noise? You MUST use pink noise for the test, never white noise.

White Noise... i'll not use it again... thanks

when the noise was exactly 90db I was sure that I could change to the sine to proceed with the test.

The level should be 85 dB, but 90 is OK too. No problem there. But why did you switch to sine wave testing? That isn't realistic. Pink noise is the correct way of testing: Since sweeps only tell you about individual frequencies, one at a time, but they don't tell you about the interaction of the room as a whole with all frequencies at once.

In the sine generator I began to analyse through every 1/3 octave, but then, since there were a lot of valleys in the analysys, I began to go further in the HZ.

1/3 octave is no use. For the low end of the spectrum, you need to analyze with precision of 1 Hz, and you can allow that to get broader as you get to higher frequencies (maybe 30 Hz at 300 Hz.).

Here was my procedure to make the test:

It would be much better, much more accurate, and much more complete to just download the REW program from Home Theatre Shack ( http://www.hometheatershack.com ). It is free, easy to use, very accurate, and does many more types of analyses, such as waterfall plots, which are very useful for diagnosing modal problems.

i'll do that, sure.

nearfield monitors (KRK Rokit Passive

You mean the R6? I guess that must be it, since it is the only passive Rokit. If so, the frequency response is only down to 50 Hz, so you can ignore everything below that on your graph (the entire left half of your graph is meaningless).

They are the old school Rokit Passive.... really they are good from 69hz on... http://www.dancetech.com/item.cfm?threadid=1254&lang=0

From 25hz to 63hz it was having an acceptable curve,

Actually, that isn't acceptable at all! It means nothing, since the response of the KRK Rokit R6 is not linear below 50 Hz (and probably more like 60 Hz), plus your sound level meter is not very linear that low either, and besides it is extremely hard to get accurate measurements of low frequency sound, especially in such a tiny room. You can ignore everything below about 60 Hz on there, and even then that isn't accurate.

suddenly in 80hz there was a -10db valley

That might be your 1,0,0 axial (first vertical mode), but more likely to just be a reflection from the stuff you put on your ceiling.

and from there a lot of flutiation over the different hz, not a progressive curve.

If that curve is accurate, then it isn't too bad at all! You have +/- 10 dB from 60 Hz up to 1 kHz, which is reasonable for a partially untreated small room like that.

It also presented bad results under 25hz.

That is totally meaningless. The readings down that low mean zero. Zilch, Nada. Nothing at all. Ignore everything below 60 Hz for those speakers and that method of measuring response.

In the next month, i'll buy a friend's Behringer B2030A + Subwoofer B2092A

Better. That goes down to around 35 Hz. But! Your room does not even support any modes below 56 Hz! So there's no need to even bother testing much below that.

but the sub part is essential in my productions, although I don't have a quite big room to hear it

You will still hear it, if your speakers can produce it! As long as your room is treated suitably, you will be able to hear all of the frequencies that the speaker can produce. Why do you think you would not be able to hear low frequencies?

What I've thought myself that would be a great idea:

Before you do any of that, you need to set up your speakers correctly, and use REW to do the testing. We need to see the waterfall plots to figure out what is going on. So set up your speakers correctly, get a good measurement mic (or any omni mic with flat response), and test again. When you do, make a note of the exact position of the measurement mic, so that you can put it in the exact same place again for future readings, after you treat the room. The mic should go in the same position as your head will be, pointing upwards at an angle of roughly 30°, but measure EXACTLY where it is, accurate to 5mm. You MUST put it back in the same position each time you measure, or the readings will not be valid. Also, make separate measurements for the left and right speakers, one at a time. Do not try to measure both at once.

Please, can you give me a link where I can find more details about this test?

Build a Helmholtz DYI

Probably not necessary. Hard to do right, and hard to position correctly. That would be the absolute last resort, after doing all the other treatment first.

Build more 2 basstraps

probably, but not tube traps. More like superchunks or maybe hangers. But first we need to know how you built your existing traps, and what that foam is. Also, what other treatment do you have in the room? For example, how have you treated the back wall?

No back wall treatment

Eq the monitors with Behringer Feedback Destroyer Pro DSP 1124P

No! Never! That will not work. It CANNOT work, since modal issues are problems in the time domain, not just the frequency domain. EQ cannot do ANYTHING to the signal to fix modal issues after the tone has stopped coning out of the speaker but while it is still bouncing around in the room! That's the problem. A mode carries on "ringing" for many milliseconds after the speaker has stopped, and during that time there is NOTHING that EQ can do to fix it. Modal issues can ONLY be fixed with suitable acoustic treatment, never with EQ. Sure, with EQ you can cut some problem frequencies a bit, but what do you do with the nulls? A null is always a null, and no amount of power increase will fill it. EQ will also only work for ONE position in the room, not for any other, and distorts the response anyway: it gives you a false representation of your music, which is totally wrong. That's why you built a studio in the first place; to get an ACCURATE representation of your music! EQ messes that up. Bad idea.

On the ground there is no treatment, its Slate over a solid concrete ground.

Good! Do not change that! It is a really good acoustic surface..

So I'd suggest that first you need to fix your speaker mounting problem (never on shelves, always on stands or in soffits, acoustic axis at 1.2m above the floor), then fix your speaker geometry problem (further apart but NOT in the corners, angled so that they are aimed directly at your ears, with the acoustic axes intersecting at 60° to 90° a few cm behind your head), the re-measure with REW and a good mic. Post those results here so that we can see where your REAL problems are.

And also post the details of how you built your bass traps, and what foam you have used there.


- Stuart -

[/quote]