Monitor placement against the wall -> flatter low freq response?
yetanotherjosh
Thread Starter
#1
8th May 2013
Old 8th May 2013
  #1
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Monitor placement against the wall -> flatter low freq response?

This is a bit of newbie acoustics question but it's been puzzling me as I try to setup my studio in a new room. I have a lot of panels and traps in the room in all the standard positions (corners, reflection points, etc), but am still hearing an eerie, deep null in the bass centered around 50hz. I find that if I push the speakers all the way to the front wall, the null gets better, and if I move my listening position off the 38% spot and forward towards the speakers (flattening the equilateral triangle into a squashed isosceles), I can improve it even more. Doing REW software measurements confirms a generally better response across the whole low-mid spectrum in this configuration.

However, the general advice I've read is to keep speakers a well off the walls and the listening position at 38%. So I'm a bit confused.

Here is what I understand is happening (please correct me): the null is caused by reflections off the wall with inverse phase. By moving the speakers closer to the wall, I rotate the phase of the reflected subs so they cancel less. And as I move my listening position further forward, I can further reduce the phase alignment issue in particular at this one hugely problematic frequency.

My question is, why is is generally frowned upon to have speakers against the wall? Isn't closer to the wall going to reduce cancellation issues at lower frequencies, trading them for cancellation issues at more midrange frequencies? And aren't acoustic treatments easier to manage for mids - so you could more easily absorb the reflections and prevent the nulls withs mids as opposed to subs. So, to my newbie thinking, speakers against the wall with treatments to address the midrange reflections is a general win, and my room experiments seem to validate it.

Can someone explain the flaw in my thinking?
#2
8th May 2013
Old 8th May 2013
  #2
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 4,019

Quote:
Originally Posted by yetanotherjosh View Post
This is a bit of newbie acoustics question but it's been puzzling me as I try to setup my studio in a new room. I have a lot of panels and traps in the room in all the standard positions (corners, reflection points, etc), but am still hearing an eerie, deep null in the bass centered around 50hz. I find that if I push the speakers all the way to the front wall, the null gets better, and if I move my listening position off the 38% spot and forward towards the speakers (flattening the equilateral triangle into a squashed isosceles), I can improve it even more. Doing REW software measurements confirms a generally better response across the whole low-mid spectrum in this configuration.
Good going! This is exactly what you should be doing.

Quote:
Originally Posted by yetanotherjosh View Post
However, the general advice I've read is to keep speakers a well off the walls and the listening position at 38%. So I'm a bit confused.
That's the problem with internet forums; - you have to have enough knowledge to sort the gold from the bear crap. - And there are many shades of grey. Nothing is ever black & white.

Quote:
Originally Posted by yetanotherjosh View Post
Here is what I understand is happening (please correct me): the null is caused by reflections off the wall with inverse phase. By moving the speakers closer to the wall, I rotate the phase of the reflected subs so they cancel less. And as I move my listening position further forward, I can further reduce the phase alignment issue in particular at this one hugely problematic frequency.
You are positively correct here.. Good thinking and measuring! 3/8th of the room is simply an 'odd' multiple of the room length where you will most likely not have resonance anomalies. However, many processes are at play in an enclosure and listening position is only one of them. AND everything works together, ie; change ONE thing and everything else changes.

Quote:
Originally Posted by yetanotherjosh View Post
My question is, why is is generally frowned upon to have speakers against the wall? Isn't closer to the wall going to reduce cancellation issues at lower frequencies, trading them for cancellation issues at more midrange frequencies? And aren't acoustic treatments easier to manage for mids - so you could more easily absorb the reflections and prevent the nulls withs mids as opposed to subs. So, to my newbie thinking, speakers against the wall with treatments to address the midrange reflections is a general win, and my room experiments seem to validate it.
Speakers against the wall will increase the bump in the LF response of the speaker because most speakers (99.9%) are designed to be free-standing. This is easily corrected with EQ (subtract/roll-off).

Correct again about the midrange frequency trapping.

Quote:
Originally Posted by yetanotherjosh View Post
Can someone explain the flaw in my thinking?
No flaw.

Cheers,
John
yetanotherjosh
Thread Starter
#3
8th May 2013
Old 8th May 2013
  #3
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Hey John, thanks for replying and helping me sanity check.

Quote:
That's the problem with internet forums; - you have to have enough knowledge to sort the gold from the bear crap. - And there are many shades of grey. Nothing is ever black & white.
I should know better, as a software developer I hate it when coworkers grab solutions to problems off internet forums without really understanding them. However, I do know enough to know that acoustics is tricky business and I am trying to make the best of my amateur status in a living-room-turned-studio.

It just strikes me as odd that speakers against the wall isn't a more recommended practice, since it seems to have some significant advantages in terms of cancellations, and I agree w/ you that a general LF boost is not the end of the world, especially if the alternative is a serious black-hole null or two in the LF response. I guess that's why in the pro studios they flush mount the speakers.

Quote:
everything works together, ie; change ONE thing and everything else changes.
Sad but true. I found as I pushed the speakers and listening position towards the wall, an excitation got worse at a higher frequency around 135hz. However, what I noticed was this was a broader spectrum defect, as opposed to a narrow spike or dip. And I think I can account for broader spectral imbalances easier (with EQ and ear training) than steep, deep holes in the response curve. So I think I picked a lesser poison, and might try to put more trapping on the front wall to see if that can be mitigated.

Thanks again
#4
8th May 2013
Old 8th May 2013
  #4
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 4,019

Josh,

You're on the right track EXCEPT for the EQ... Only ever EQ a steady-state problem - like flush mounting. You must use acoustic treatment to control reflections, comb-filtering, and for bass trapping so that your sound will be balanced.

If you use EQ for a 'room reflection' or 'modal' issue', you'll have to clamp your head in a vice AT the measurement position for the EQ to actually 'work'. Move a fraction of an inch and you're screwed again.

What you want in a composing / mixing space is balance, comfort, and the largest possible 'sweet spot'. This is only ever attained with proper modal spread (shell dimensions) and acoustic treatment.
- "Ye canno' change th' laws 'o physics!" - Scotty from StarTrek

Cheers,
John
#5
9th May 2013
Old 9th May 2013
  #5
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

Internuts

Glad you found that convenient truth josh.
Ethan and I call almost touching the Front Wall, 'Pseudo Soffit'.
Do check the Waterfalls, you might get better FR but worse modal decays. We have to balance these issues.
Sometimes it is worth inserting a 100mm Trap between speaker and Wall.
SBIR
Width and height matter also, and don't be afraid to go all the way to the Front Wall, or to invert the speakers.
It is possible to test whether your new anomaly is susceptible to Eq or not. Look up the section following Minimum Phase in the REW manual.
Even if that particular issue is not, others will be, particularly boosts and modal boom at VLF. Sorry to diverge from my friend John's opinion, but I have to say I strongly recommend demoing Eq such as Audiolense, Dirac Live, Coneq and so on. I have found that the common wisdom on this matter didn't last long once I started listening and testing.

DD
yetanotherjosh
Thread Starter
#6
9th May 2013
Old 9th May 2013
  #6
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Quote:
If you use EQ for a 'room reflection' or 'modal' issue', you'll have to clamp your head in a vice AT the measurement position for the EQ to actually 'work'. Move a fraction of an inch and you're screwed again.
I've historically shied away from correctional EQ for room modes and reflections for this reason, but I've kind of reached my limit on time and money for the absorption treatments, and I can't do much with the room shape. Would you still say I should avoid it even if I'm addressing only low frequency issues, say below 200hz? Won't I get some leeway with the listening position there - maybe 1 foot? I'm mainly focused on getting as flat a low end as I can, since I mostly focus on club music. And I've just never been able get results out of headphones that didn't sound awful once I switched to the monitors.

Quote:
Look up the section following Minimum Phase in the REW manual
Thanks DD for pointing that out. That was pushing my limits of theory but I think I understand it, and it does look like this one woofy bump around 135hz is a minimum phase region. I did check the waterfalls, and what I found was that some decay times were made worse but it was really only after they'd already fallen off about 25db, which I found odd, but I figure it's the right tradeoff for me.

Thanks again guys
Audio X
#7
9th May 2013
Old 9th May 2013
  #7
Audio X
Guest
 
Posts: n/a

Quote:
Originally Posted by yetanotherjosh View Post
I did check the waterfalls, and what I found was that some decay times were made worse but it was really only after they'd already fallen off about 25db, which I found odd, but I figure it's the right tradeoff for me.
Out of curiosity, what kind of speakers are you using and what are the dimensions of your room? Also, what kind of treatment do you have behind the speakers?
#8
9th May 2013
Old 9th May 2013
  #8
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

Bump

Well bumps are much more amenable to Eq than Nulls.
When using these Eq systems, there is typically a choice of listening area.
I sample an area about 1 foot deep by 3 wide, nine samples.
This seems to work well. The imaging is sharpened, in particular the position and solidity of the central image. This of course makes it easy to remain in a fairly limited area without having to get visual bearings.
Most of the room seems to benefit in this mid clarity and the bottom end is more confident. The only downside is extra bass at the couch at the back, which is unusable in any case.

DD
yetanotherjosh
Thread Starter
#9
9th May 2013
Old 9th May 2013
  #9
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Quote:
Originally Posted by Audio X View Post
Out of curiosity, what kind of speakers are you using and what are the dimensions of your room? Also, what kind of treatment do you have behind the speakers?
I'm using Dynaudio BM15As in a longer room that's 11'4" x 27'. (It's worth noting the rear wall is not closed, about 1/3 of it is open into the kitchen.)

I've got the speakers now about 1.5' from the wall, and behind each one is a pair of stacked GIK Acoustics "Monster" bass traps, which are at about a 30 degree angle to the front wall. (The ceiling is just under 8' so I can't vertically stack 2 4' traps in portrait mode with 45 degrees to the corner walls, so unfortunately I have to lay them on their side in "landscape" and at a slighter angle to the front wall, putting a bigger corner air zone behind them and forcing the speakers forward a little more, and leaving the ceiling corners untreated).

Quote:
Originally Posted by DanDan
When using these Eq systems, there is typically a choice of listening area.
Oh, interesting. Trying one of these systems out is going on my list of weekend projects to do soon. Trying it out will be informative in the least. Thanks.
Audio X
#10
9th May 2013
Old 9th May 2013
  #10
Audio X
Guest
 
Posts: n/a

Quote:
Originally Posted by yetanotherjosh View Post
I'm using Dynaudio BM15As in a longer room that's 11'4" x 27'. (It's worth noting the rear wall is not closed, about 1/3 of it is open into the kitchen.)
Ok. Thanks. I've owned the Dynaudio Air15's and Air20's and used them extensively in the past with and without the Dyn base 24 sub.
If the BM's are anything like the Airs they probably have a healthy dose of low mids inherent in the speakers like most Dyn's studio speakers that I've heard.

You've already received some good advice. Getting the right speaker placement and listening position in any given room can be a daunting task to get right because there are a lot of variables involved. It's also one of the most important things to have tuned in, ..in any studio

I wouldn't worry to much about adhering to too many rules circulating on the internet although left and right symmetry is one of the considerations to follow and having the high mids and tweeters around ear height is also a good one.

I'd try it for a while how you like it now... close to the wall and see how those mixes translate out side of your room. You have a lot of front to back space to work with, so farther into the room could also be a consideration when testing if against the wall isn't giving you what you want. It can be a game of inches .. or feet.

Getting it to sound good and trusting what you hear inside your room and having that translate seamlessly outside your room with consistency is the goal .. Translation is one way that will let you know right away if you got it right.

I think there are two types of thinking when going about finding the best location. One is more on the scientific side and is to test through measurement and follow what the numbers say, the other is to trust your intuition and instincts. Combining both might give you the best results although experience with critical listening comes into play.

I did just run across this quote from Harvey Gerst who has designed some very fine speakers over the years and this might be helpful to your situ.. gl

"As you get close to any large flat surface, you get into boundary issues: 1/2 space, 1/4 space. etc.

While I can work in very close to the HG3's, I like to be about 4' to 5' out from them for best bass response. Any closer to the speakers gives me a sense of some bass loss, although the imaging becomes more amazing to me as I get closer.

How small a room, and how close to the wall, is a trial and error process for me. In a very small room, try about 2 feet out from the wall, and 3 to 5 feet apart. Moving the HG3's closer to the wall might boost the bass a bit, but will increase the room's standing waves."
#11
10th May 2013
Old 10th May 2013
  #11
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

Near

Josh have you tried your speakers really close to the front wall. I mean almost touching, or touching the Monsters? Your logic goes all the way.
The closer to the wall the more soffit like the behaviour. Which includes the modal down side of course. But as you are trying Eq, you can control the LF boost as you chose, and software does shorten low modes.
DD
yetanotherjosh
Thread Starter
#12
10th May 2013
Old 10th May 2013
  #12
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Quote:
Originally Posted by DanDan
Josh have you tried your speakers really close to the front wall. I mean almost touching, or touching the Monsters?
Encouraged by this thread I did indeed boldly push them all the way to the wall, both directly in front of the monsters and also right flush with the wall. My hunch is that if I could have done that *and* maintained the heavy trapping I otherwise had in the front corners when the speakers are off the wall a bit, it would have been better. But I had to remove some panels in order to get them closer, so I think it was a net loss (the response curve got smoother in some spots but took on more sharp peaks and dips elsewhere.)

What I haven't tried yet is squishing them closer together laterally, because well, it's hard to test changes in two dimensions at the same time! (Not to mention listening position as well, it's just so time consuming to try the variations!)

So right now with them just under 1.5' from the wall with them pretty well flush against some (angled) monster traps. It's now a manageably good response with the exception of an ugly dip at 65hz. Next step is to try lateral adjustment as well.

Quote:
Originally Posted by Audio X
I think there are two types of thinking when going about getting the best location. One is more on the scientific side and is to test the crap out of everything and follow what the numbers say, the other is to trust your intuition/instincts. Combining both might give you the best results. Experience comes into play.
Experience is something I'm only just getting started on . And it's definitely fun, albeit slow, to be methodical with the test-driven approach to positioning.

Thanks to everyone on the thread for encouraging me to break the standard advice and go for broke towards what works best.

Quote:
Originally Posted by Audio X
I've owned the Dynaudio Air15's and Air20's and used them extensively in the past with and without the Dyn base 24 sub.
If the BM's are anything like the Airs they probably have a healthy dose of low mids inherent in the speakers like most Dyn's studio speakers that I've heard.
FWIW I bought the BM15A's after hearing a friend's surround studio setup with Air 15's and felt I was pretty much getting the same sound. Someday I'll get myself a second pair of speakers to give me a sanity check against the Dynaudio hegemony...
#13
10th May 2013
Old 10th May 2013
  #13
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

All

One of my clients did it all, different widths and heights at each different distance from the front wall. And the corresponding listener positions.
A large amount of data but very informative. The optimum spot was about one foot from the front wall if I remember correctly.
I find the Spectrum Analyser in REW really good for getting a quick but sure impression. You can look at the screen and move the mic about. Ditto a speaker. The no fly zones become quickly obvious.
Note also Jeff's trick of placing a speaker at the listener position and moving the mic in the potential real speaker spots. its interchangeable.

DD
#14
10th May 2013
Old 10th May 2013
  #14
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 4,019

This can easily get crazy very fast. So....

Dan has some great points... But this is easier: (I think)

Place only ONE of your speakers on the floor in one of the front corners. FACE the corner with the speaker and tilt it into the tri-corner, ie; floor and two walls.

This will excite all room modes.

Position the microphone beginning at .375 X the room length from the front wall where you will have your mix position. Set the microphone at ear height. point the mic to the ceiling. (omni testing mic)

Take a 20Hz to 20kHz measure. - I prefer a full measure for data.. you can chop it up later.

Move the mic forward 5 cm at a time and take measures. do this about 6 times.

MARK EACH MEASUREMENT SPOT ON THE FLOOR AND LABEL IT WITH TAPE.

Put the mic back at the starting position and move it BACKWARDS 5 cm at a time and take measures.

Organize the mdat files with distance measurements as file names.

You can now look at the waterfall and FR to see where things settle as you move forward and back. This can give you a 'picture' of your room. With this data you can position yourself in the 'best' place for YOUR room.

Once you find the 'best' position for the mic, mark it on the floor. That's going to be your listening spot. It might not be great, but it's the best you can do right now. Don't go splitting hairs.

Next set up your speakers in an equilateral triangle pointing at a spot 15-18 cm or 6-7" BEHIND the 'best' position spot that you marked earlier. (It is not critical but this will point the speakers at your EARS and not your nose).

Run tests again with the speakers right up next to the wall in front and move them out (toward the mic) about 5cm at a time. Maintain the equilateral triangle. Do as many measures as you want/need. Mark and label.

Look at the data to find the 'best' position for the speakers.

I find that this is the optimum procedure. First you set the Listening position based on the room, then you set the speakers based on the room.

You must have treatment, even if it is unopened bags of fiberglass stacked around. Get whatever you can to help you trap the room down to a manageable decay rate. - You'll see that on your RT-60 graph. -- See where it just 'takes off' on low frequencies? Treatment will mostly affect damping and early reflections that cause the comb filtering that you may see on the graphs.

NOTE: Use a heavy desk or table placed between the speakers and microphone for this measurement. POINT the desk AT the speaker cones by putting bricks or boards under the back legs. This will eliminate both desk reflections and floor reflections. The floor reflections can cause destructive LF comb-filtering.

Cheers,
John
yetanotherjosh
Thread Starter
#15
10th May 2013
Old 10th May 2013
  #15
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
I like this a lot, thanks John! It's the first description of a reproducible process I've read for how to start from scratch in a room to optimize speaker placement and listening position methodically.

Couple questions:

Quote:
Originally Posted by jhbrandt View Post
Next set up your speakers in an equilateral triangle pointing at a spot 15-18 cm or 6-7" BEHIND the 'best' position spot that you marked earlier. (It is not critical but this will point the speakers at your EARS and not your nose).
Because my room is long (27' x 11'4"), if I were put my LP at 0.375 the length from the front wall (just over 10'), the largest equilateral triangle I can make with the speakers will put them flush with the side walls, not the front wall. In my own tests so far I've found things sounding better when they are closer to the front wall, which either forces me to break the triangle or to move my LP forward. Also, FWIW I'd rather optimize for frequency response over stereo image if I had to choose. Given those, what do you think about the importance of the triangle?

Quote:
Originally Posted by jhbrandt View Post
Use a heavy desk or table placed between the speakers and microphone for this measurement. POINT the desk AT the speaker cones by putting bricks or boards under the back legs.
I did notice this was a problem for me earlier, as I was initially taking measurements without the desk and they'd change when I put it in of course. When you say angle the desk, do you mean having it sloping so that a pen will roll towards you or towards the speakers? In either case this may be a challenge for me since I'm using a standard plank desk and need to put a monitor, keyboard, mouse, etc on it. Instead I could see getting some plywood or other hard flat pieces and laying them on the sides of the desk with wedges underneath, so the center of the desk is flat but the sides are flayed. The overall desk would be parallel with the floor in that case but the reflection points to the speakers would be angled. Will that make much of a difference?

Quote:
Originally Posted by DanDan
I find the Spectrum Analyser in REW really good for getting a quick but sure impression.
Oh sweet, I didn't notice that feature (the "RTA Window" it's called, now I see). Thanks Dan!
#16
10th May 2013
Old 10th May 2013
  #16
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

FR

I would prioritise even bass response over stereo imaging.
I favour a minimal desk, just big enough to hold keyboard and mouse. An angle can be helpful for HF reasons. Desktop Reflection Revealed
If I had to use a large desk for some reason, then I would try to make it double duty blocking floor reflections.

I am all for making a map of the modes in the room to inform the choices of speaker and listener positions. However I believe the complex interactions between modes, BIR etc. mean that each speaker position has a corresponding optimum listening position. There have been a few attempts at describing the optimising process in posts. boggy has a nice one somewhere here on GS. IMO though, it almost doesn't matter where you start as it is an iterative process. Everything influences everything else. One hand washes the other.....

DD
#17
10th May 2013
Old 10th May 2013
  #17
Gear addict
 
Joined: Feb 2011
Location: Norwich, UK
Posts: 439

Quote:
Originally Posted by jhbrandt View Post
This can easily get crazy very fast. So....

Dan has some great points... But this is easier: (I think)

Place only ONE of your speakers on the floor in one of the front corners. FACE the corner with the speaker and tilt it into the tri-corner, ie; floor and two walls.

This will excite all room modes.

Position the microphone beginning at .375 X the room length from the front wall where you will have your mix position. Set the microphone at ear height. point the mic to the ceiling. (omni testing mic)

Take a 20Hz to 20kHz measure. - I prefer a full measure for data.. you can chop it up later.

Move the mic forward 5 cm at a time and take measures. do this about 6 times.

MARK EACH MEASUREMENT SPOT ON THE FLOOR AND LABEL IT WITH TAPE.

Put the mic back at the starting position and move it BACKWARDS 5 cm at a time and take measures.

Organize the mdat files with distance measurements as file names.

You can now look at the waterfall and FR to see where things settle as you move forward and back. This can give you a 'picture' of your room. With this data you can position yourself in the 'best' place for YOUR room.

Once you find the 'best' position for the mic, mark it on the floor. That's going to be your listening spot. It might not be great, but it's the best you can do right now. Don't go splitting hairs.

Next set up your speakers in an equilateral triangle pointing at a spot 15-18 cm or 6-7" BEHIND the 'best' position spot that you marked earlier. (It is not critical but this will point the speakers at your EARS and not your nose).

Run tests again with the speakers right up next to the wall in front and move them out (toward the mic) about 5cm at a time. Maintain the equilateral triangle. Do as many measures as you want/need. Mark and label.

Look at the data to find the 'best' position for the speakers.

I find that this is the optimum procedure. First you set the Listening position based on the room, then you set the speakers based on the room.

You must have treatment, even if it is unopened bags of fiberglass stacked around. Get whatever you can to help you trap the room down to a manageable decay rate. - You'll see that on your RT-60 graph. -- See where it just 'takes off' on low frequencies? Treatment will mostly affect damping and early reflections that cause the comb filtering that you may see on the graphs.

NOTE: Use a heavy desk or table placed between the speakers and microphone for this measurement. POINT the desk AT the speaker cones by putting bricks or boards under the back legs. This will eliminate both desk reflections and floor reflections. The floor reflections can cause destructive LF comb-filtering.

Cheers,
John
+1

This is a great post John! Listen up people.
#18
10th May 2013
Old 10th May 2013
  #18
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 4,019

Thank you.

When searching for the best mix position, begin with odd multiples of the room length, ie;
most commonly accepted: 3/8 = 0.375

Here I'm using prime numbers;
2/7, 2/11, 2/13 = 0.285, 0.1818, 0.153
3/7, 3/11, 3/13 = 0.428, 0.2727, 0.23
5/11, 5/13, 5/17, 5/19 = 0.4545, 0.384, 0.397, 0.294, 0.263
7/17, 7/19, 7/23 = 0.411, 0.368, 0.304

(AVOID 1/4, 1/2, 3/4 = 0.25, 0.5, 0.75 and similar even increments.)

In order they are:

0.153
0.1818
0.23
0.263
0.2727
0.285
0.294
0.304
0.368
0.384
0.397
0.411
0.428
0.4545

You can also try Pi / 10 = 0.314

and phi, the 'golden number' = 1.618 and divide it by a prime such as:

1.618/5 = 0.3236
1.618/7 = 0.2311
1.618/11 = 0.147

- I use number theory in all my room designs...

Desks; Like Dan said, a 'TV' tray would be ideal.
But the main purpose of the desk that I mentioned above is only for this 'testing'. You want to eliminate floor reflections from your data. They will not be present once you are setup anyway. The 'test' desk must be angled so that a pencil will roll toward you.. or the mic. Look at the edge of the desk and aim it like a shotgun at the acoustic center of the speaker. This will be an EXTREME tilt, ok? Probably NOT what you would want to end up working with. but this eliminates reflections off the desk into the mic and blocks floor reflections. I hope that this is not confusing. This is only for the 'before room setup' measures. - So that you can set your listening and speaker positions.

Cheers,
John
#19
10th May 2013
Old 10th May 2013
  #19
Gear addict
 
Joined: Nov 2007
Posts: 385

Quote:
Originally Posted by DanDan View Post
I would prioritise even bass response over stereo imaging.
I favour a minimal desk, just big enough to hold keyboard and mouse. An angle can be helpful for HF reasons. Desktop Reflection Revealed
If I had to use a large desk for some reason, then I would try to make it double duty blocking floor reflections.

I am all for making a map of the modes in the room to inform the choices of speaker and listener positions. However I believe the complex interactions between modes, BIR etc. mean that each speaker position has a corresponding optimum listening position. There have been a few attempts at describing the optimising process in posts. boggy has a nice one somewhere here on GS. IMO though, it almost doesn't matter where you start as it is an iterative process. Everything influences everything else. One hand washes the other.....

DD
I 've been going through a similar process of extensive measurement of every variation I can think.
For the record the speakers ended up fairly close to the wall now.
Still not perfect, but sounding better than ever.

If I may inject question here - what's a reasonable amount of spikey-ness that can be expected? Not talking about world class Dolby studios here -but what is considered "good" or at least "good enough" ?

With every measurement some things are better but other anomalies may pop up so its always a trade-off.
When can I stop?
#20
10th May 2013
Old 10th May 2013
  #20
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

Enough

When you run out of time/patience/money/interest?
Experimenting with acoustics is one thing, doing actual audio work is another.
e.g. I optimised my positions and treatment years ago, by ear.
I learned to work around the (later found to be blatant) anomalies.
Audio Engineers do that.
Some time passed and Fuzzmeasure appeared. I tested the response, increased the amount of treatment but settled on similar positions again by ear.
The devil you know. But more substantially one can't shift from a well known (if flawed) response to a new one without experiencing a whole new translation adjustment. Audio work becomes unreliable during the transition, so we tend to leave things alone.
More time passed until one day I decided I should follow the advice I regularly give to others......LOL
So I optimised positions and achieved a much better response. This time, I found it remarkably easy to read myself into this new one.

Rambling over.

I recommend applying priorities.
Globally nulls are more damaging than peaks.
The biggest proportion of energy and defining fundamental tone is at LF. But not all LF is equal.
20-40Hz, rumble, gratuitous bigness, hardly vital, nor audible on smaller speakers.
40-160Hz Musical fundamental notes, crucial.
Kick area 60-70 deserves special focus.
Acoustic music deserves special focus on the 85-105 region. Acoustic guitars resonate, boom, in this band.
160-320, warmth, peaks can be tamed, nulls would be worrying.
Say 4K up is easily dealt with by HF reflection killers and simply adjusting the level or Eq.

DD
#21
11th May 2013
Old 11th May 2013
  #21
Lives for gear
 
jhbrandt's Avatar
 
Joined: Feb 2009
Location: Jakarta, Indonesia
Posts: 4,019

Quote:
Originally Posted by jeff_free69 View Post
...With every measurement some things are better but other anomalies may pop up so its always a trade-off.
When can I stop?
Read my post above. Get enough measures to 'see' what your room has to offer. THEN DEAL WITH IT! Major, major, super-big, expensive, world-class studios DO NOT have 'perfect' acoustics! They are very good! - But everything in acoustics is always a compromise. If someone told you that they built their studio with zero compromise and it's perfect, either they really don't know - or they're lying to you.

Quote:
Originally Posted by DanDan View Post
....The devil you know. But more substantially one can't shift from a well known (if flawed) response to a new one without experiencing a whole new translation adjustment. Audio work becomes unreliable during the transition, so we tend to leave things alone.
More time passed until one day I decided I should follow the advice I regularly give to others......LOL
So I optimised positions and achieved a much better response. This time, I found it remarkably easy to read myself into this new one.

Rambling over.

I recommend applying priorities.
Globally nulls are more damaging than peaks.
The biggest proportion of energy and defining fundamental tone is at LF. But not all LF is equal.
20-40Hz, rumble, gratuitous bigness, hardly vital, nor audible on smaller speakers.
40-160Hz Musical fundamental notes, crucial.
Kick area 60-70 deserves special focus.
Acoustic music deserves special focus on the 85-105 region. Acoustic guitars resonate, boom, in this band.
160-320, warmth, peaks can be tamed, nulls would be worrying.
Say 4K up is easily dealt with by HF reflection killers and simply adjusting the level or Eq.

DD
+1 Dan,
Great run-down.

Cheers,
John
#22
11th May 2013
Old 11th May 2013
  #22
Gear addict
 
Joined: Nov 2007
Posts: 385

Quote:
Originally Posted by DanDan View Post
When you run out of time/patience/money/interest?
Experimenting with acoustics is one thing, doing actual audio work is another.
e.g. I optimised my positions and treatment years ago, by ear.
I learned to work around the (later found to be blatant) anomalies.
Audio Engineers do that.
Some time passed and Fuzzmeasure appeared. I tested the response, increased the amount of treatment but settled on similar positions again by ear.
The devil you know. But more substantially one can't shift from a well known (if flawed) response to a new one without experiencing a whole new translation adjustment. Audio work becomes unreliable during the transition, so we tend to leave things alone.
More time passed until one day I decided I should follow the advice I regularly give to others......LOL
So I optimised positions and achieved a much better response. This time, I found it remarkably easy to read myself into this new one.



DD
That's the path I'm on now. My listening position only moved 6 inches and my speakers about a foot closer to the front. So I wasn't too far off. That helped but the room still needed significant more treatment.

Rather than hijack this thread I'm spinning my own, thanks!
HobbitHoleStudio - how does it sound now?
yetanotherjosh
Thread Starter
#23
14th May 2013
Old 14th May 2013
  #23
Gear interested
 
Joined: Aug 2009
Posts: 17

Thread Starter
Quote:
Originally Posted by jhbrandt View Post
Place only ONE of your speakers on the floor in one of the front corners. FACE the corner with the speaker and tilt it into the tri-corner, ie; floor and two walls.

This will excite all room modes.

...
Well John I took your advice and followed this procedure yesterday. And I want to thank you for your guidance because I have a setup now that, while far from flat, is the best I've ever had, and significantly better than what I had before this thread started (I had a 15db dip around 65hz, which is now cut in half to 7db, and the waterfall looks significantly better too with reduced resonance/booming).

For what it's worth, here are some of my observations:

- Using the single speaker pointed into the corner, I was indeed able to find the most optimum listening position in the room. That worked really well. It turns it out it was not very far from where I'd located my LP earlier using more ad-hoc testing approach, so I take that as confirmation by triangulation that the technique works.

- Putting the desk in place definitely changed the graphs but I'm not sure if my results would have suffered had I skipped the desk step. It was unruly to get it propped up that way, and I found that just placing it in normal parallel-to-the-floor position in the physical location it would be given the particular LP I chose turned out to be good enough in the end.

- Once I had the LP setup and started placing speakers in the equilateral triangle position flush with the front wall and inching them forward, I found the optimum positions for them given the equilateral constraints. However, when I moved the speakers laterally wider from that optimal spot on the triangle, into "wide" positions, I was able to get better results. (They worsened initially as I diverged from the triangle, but out of curiosity I kept pushing them wider and eventually found a markedly better configuration). So I am pretty confident than in my room, a squashed triangle with the speakers extra wide is giving me flatter low end.

- In all my tests, I moved the mic (and later the speakers) by 3" (~7.5cm) increments for each measurement. I found this to be fine-grained enough. And REW has a limit on the number of measurements you can capture in a single session, so doing a littler fewer helped there too. You suggested a smaller increment of 5cm, but as I laid tape on the floor and started marking out the positions, bumping it up a tad out of laziness turned out to be fine.

- Speaking of tape: I used painter's tape to lay a stripe down the center line of the room, and marked the increments on there. This turned out to be a good idea that made doing measurements fast and reliable. When I got into speaker placement, I used tape to lay out the lines of the equilateral triangle radiating at +/- 30 degrees from the LP. Without the floor tape I would have been fiddling with a tape measure or yardstick to keep things in the right position which would have been a major pain, so fwiw to others on the thread I'd suggest doing the same tape on the floor approach.

- I didn't end up using any kind of mathematical ratios or percentages of room length. What I did was mark (in the 3" increments) on the floor tape between the two furthest positions that would be even remotely reasonable as actual working LP's for the room, and just figured the 3" granularity would land me on the right spot between them. That worked fine and I didn't have to do any math based on the room, which in my case is probably inaccurate anyway given the nature of the room (it's not a closed room, it has a partly rear back wall, and a small half-height wall about halfway through, etc).

- By the time I'd mostly optimized the LP and speaker positions, I put the absorption panels back in and found their main benefit was on the waterfalls, and in slightly tightening up the graph. Originally I was of the mind that the panels were really the "solution" and placement was less important, but now I see it the other way around. Both are vital but placement is what you have to get right first!

Well thanks again and I hope this helps others to get good results as well. I think I've now arrived at Dan's said threshold of "time/patience/money/interest" at least on the physical setup side of things, though I think I will still try one of those software-based correction suites to see how far I can close the gap now that I've done what I can in the physical domain...

cheers
-JW
#24
14th May 2013
Old 14th May 2013
  #24
Gear Guru
 
DanDan's Avatar
 
Joined: Aug 2003
Location: Cork Ireland
Posts: 11,094

Movement

jeff, six inches and a foot seem like large moves to me. Remember we have found it necessary to test in very small increments particularly when zeroing in on the final decisions.

josh, unfortunately I think you may have taken a few wrong turns. I believe that acoustic treatment can and most likely will change the optimum (least bad) Speaker Positions and Listener Positions.

The overall problem here is that we have an impossibly complex geometric.
Three positions have to be optimised in 3 dimensions in 50mm increments.
Most importantly they all interact. Everything depends on everything else.

To turn this into a doable exercise, explainable in an internet posting, we have to cut corners. We tie down one or two of the dimensions. Different people chose to tie down different aspects in different orders. This is OK if we remember that they all interact, and do iterative testing all the time.
It can also be helpful to start with broad strokes. Six inches at a time, going down to 2 or less when we are near the final. A Real Time Spectrum zoomed to LF is a fast to instant way of confirming and rechecking. It can also identify whole areas of no fly zones very quickly.

SBIR can be much more significant than modes. http://recording.org/studio-construc...en-matter.html
Also I believe every different speaker position has a corresponding optimum LP.
This line of thought has persuaded me to adopt the following sequence of fixing/priority.

I would settle on a optimum Height first, subject to periodic verification that it is holding up, and certainly final checking. Remembering many speakers work fine upside down which places the woofer alternatively.
Next, I settle on a Width, this time verifying more often if this optimum holds up as lateral movement progresses.

I posted this below recently in reply to an OP.
In PM destana asked me if such a 'Fussy' process applies to everyone.
IMO, Yes.

Quote:
Glad to hear your project is coming along well. Optimum speaker and Listener Positions are interactive. I would start with the speakers pretty much touching the Front Wall. Let's assume Height is fixed, tweeter at ear level. Pick a likely width within the practical range of options. Play Pink Noise PN over both speakers using the REW Generator. Find the optimum LP by moving the mic about while watching the Spectrum analyser in REW, Zoomed to LF only.
Monitor placement against the wall -> flatter low freq response?-rtspectrum.jpg
Now try changing the width, 50mm at a time, taking sweep measurements.
Better or worse is usually pretty obvious. Settle on the optimum Width and check the LP again. Now take the 0 distance from Front Wall Sweep measurement. I think both L+R speakers running is good in this work.
As height and width are now pretty much set we can now take a series of measurements moving the speakers away from the Front Wall in 50mm increments. Remember to optimise the LP for each new speaker position before each sweep.
DD
Quote
2
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
 
Register FAQ Search Today's Posts Mark Forums Read

SEO by vBSEO ©2011, Crawlability, Inc.