Magickman's room tuning thread

[Magickman]

Hello forum members!

I've spent the last year doing an acoustic makeover of an existing studio. I'm quickly approaching the extent of my abilities and knowledge, and I'm hoping that some of the more knowledgeable folks here (e.g. - Andre, Sac, Lupo, Jens Eklund, G.E., et al) will assist me in moving further.

Some of you may have seen my designs in progress throughout the last 18 months or so as I occasionally pop up with a related question, or join in on other discussions.

My design criteria was unusual in that I really wanted to build modular treatments which could be moved to a new location at some later date. The reason for this approach is that I have no idea whether our lease will be renewed or not when it ends 3 years from now. In hind site, this was probably not the best approach, as it has exponentially increased both the amount of labor involved, and the overall cost of treating the room. However, I have enjoyed the design process immensely, and I do so love building things with wood!

Going hand-in-hand with the concept of these modular treatments is the requirement for them to be re-tunable for future applications. It was quite a design challenge to create something that is effective, looks professional, and that could actually be built by someone with limited tools and woodworking skills (i.e. -ME!). I have started a thread dedicated to the construction process here:

http://www.gearslutz.com/board/studi...ml#post6098843

I am a professional audio engineer, and have been working in the industry for over twenty years. I have had a strong interest in acoustics and studio construction for most of that time, and have even done some acoustics consulting.

Anyway, I'll start by showing a rendering of the room as it stands now (more or less), and then post some more details about the design concept behind the traps I've built.

[Magickman]

So, the concept behind these traps is that they are modular, multi-chamber devices. Each chamber can be used as either a purely porous device, a perforated panel-type device, or as a Helmoltz absorber device.

As currently configured, there are sealed chambers top and bottom, which are intended to be tuned, HH absorbers. Everything else is just open framing, intended to be very deep, porous absorbers, which are then faced with perforated panels in order to enhance low frequency absorption and reduce high frequency absorption. The idea is to keep as much energy in the room as possible.

[Magickman]

As you can see, the individual modules are relatively small, and stack together to form an array. Part of the design criteria was that they should be portable as I wanted them to easily be moved and stored if/when required.

[Magickman]

Here's some close-ups of the individual modules.

[Magickman]

Here's a couple more renderings of just the traps for clarity's sake.

[Magickman]

Oh, here's how the clouds look in real life. I'm not going to go into construction details in this thread. If you're interested, please see my construction thread linked above. It may be a while before I get to the trap/clouds in that thread, but I will cover their construction in detail in that thread.

You can see the DIY QRDs I built about 7 years ago in the background of one the shots. These pictures are in my old house where I had my studio for 7 years.

[Magickman]

Okay, now that you see what these things are, let's look at their affect on the room. I'll be back later with some room measurements.

Stay tuned!!!

-MM

[gullfo]

very nice workmanship!

couple of ideas to add: you angle the front clouds down a bit to increase the air space there. on the modular resonator panel units - put the panels with the smaller holes on the top and bottom and the larger holes in the middle.

[Magickman]

Originally Posted by gullfo

very nice workmanship!

couple of ideas to add: you angle the front clouds down a bit to increase the air space there. on the modular resonator panel units - put the panels with the smaller holes on the top and bottom and the larger holes in the middle.

Thanks Glenn. The front clouds have already been angled down. That photo was take months ago.

About the resonators: If you look at the images in my first post you'll see how these traps are currently configured. Keep in mind I'm playing catch up here as these measurements were done quite some time ago. I am showing the progression of our efforts here.

Stay tuned!!!

[Magickman]

Okay, I'm back!

The first thing we did was try to determine optimal listening position and speaker placement.

Unfortunately, I never took any measurements of the room prior to hanging the clouds, so all the measurements presented are with the clouds in place. It's really too bad, because the clouds had such a huge impact on the room. They each contain 6 inches of 2.5 lbs per CF rockwool and bridge the crown of the 9 foot vaulted ceiling creating about 12" air space between the clouds and the ceiling(!).

But, suffice to say, that as bad as the room looks in the first measurements (with only the clouds in place), it was far, far worse prior to hanging the clouds. Really, just simply amazing the affect they had on the room.

The first set of measurements were done using a measurement mic placed precisely at the 38% point of the room. The mic was at tweeter height and at 90° orientation. I also purchased a calibration file with the mic which has been applied to the measurements. The measurements were taken with FuzzMeasure using pretty long sweeps in order to keep the S to N as high as possible. We then took a series of measurements, moving the mic 3 inches closer to the front wall for each successive measurement. To keep things somewhat tidier, I'll only theft speaker.

[Magickman]

Next we moved the mic in the opposite direction, starting at the 38% mark and moving away from the front wall, 3 inches at a time.

The first graph below is at 38% + three inches.

[Magickman]

Based on this series of measurements, we determined that the optimal listening position was in fact 12 inches closer to the front wall.

Oops! I've got a rehearsal starting in a while, so I have to go; but I'll be back.

Think of it like a cliff hanger!

-MM

[SAC]

The frequency response and derived waterfall are useful for modal analysis from 0-~300 Hz.

Above that, properly made ETC responses (L& R - one per isolated source) provide information regarding the specular energy behavior with respect to time.

Looking at a frequency response above the modal region will tell you little about the specular behavior of the space - with the exception being that you have one (and I could have saved you the trouble of making the FR for that insight- consider it a freebie ;-) . And as multiple sources superpose to create comb filtering, the FR provides no insight into the various components that superpose (combine) to create said anomalies, let alone sufficient information to identify and re-mediate them.

Thus I might suggest deciding what issue you are going to analyze and address, and then using the proper perspective to gain insight into the nature of the problem and to verify steps taken to remediate the situation. Thee is little to be gained by using the inappropriate response to try to analyze and resolve the contributing causal factors. Thus an understanding of the basic nature of the issue to be examined as well as the proper tool capable of displaying the characteristics of the causal variables will go a long way toward helping you to more accurately and quickly identify and resolve issues.

[Magickman]

Originally Posted by SAC

The frequency response and derived waterfall are useful for modal analysis from 0-~300 Hz.

Above that, properly made ETC responses (L& R - one per isolated source) provide information regarding the specular energy behavior with respect to time.

Looking at a frequency response above the modal region will tell you little about the specular behavior of the space - with the exception being that you have one (and I could have saved you the trouble of making the FR for that insight- consider it a freebie ;-) . And as multiple sources superpose to create comb filtering, the FR provides no insight into the various components that superpose (combine) to create said anomalies, let alone sufficient information to identify and re-mediate them.

Thus I might suggest deciding what issue you are going to analyze and address, and then using the proper perspective to gain insight into the nature of the problem and to verify steps taken to remediate the situation. Thee is little to be gained by using the inappropriate response to try to analyze and resolve the contributing causal factors. Thus an understanding of the basic nature of the issue to be examined as well as the proper tool capable of displaying the characteristics of the causal variables will go a long way toward helping you to more accurately and quickly identify and resolve issues.

Thanks Sac. I actually meant to attach the ETCs and Waterfall plots, but I was trying to beat my rehearsal downbeat and simply spaced it out until after I had already posted the FR graphs. Once I get back home, I'll repost everything properly.

Keep in mind, at this point I'm only trying to show our progression. These graphs (and the accompanying ETCs and waterfall plots) are in an essentially untreated room. I will then show the affect of each treatment as it's added.

Thanks for joining in to my discussion! Just give me a little while to get you all up to date on where the room is at now.

Stay tuned!!!

-MM

[Magickman]

Okay, these plots show the comparison of FR from 50Hz-300Hz of the various mic positions. Hopefully this will make it a little easier to see how we arrived at the decision that 12 inches closer to the front wall was the best listening position. I do believe the black trace shows the "smoothest" FR (and I do use the term "smoothest" loosely!). Remember, this is essentially an untreated room at the time of these measurements (only the clouds are installed). In the interest of not creating too much clutter or confusion, I've opted to not show the ETCs or Waterfalls at this time. Our first step was simply to find the position offering the smoothest frequency response.

Blue Trace=38% Position
Green Trace=3 inches closer to front wall
Red Trace=6 inches closer to front wall
Yellow Trace=9 inches closer to front wall
Black Trace=12 inches closer to front wall

[Magickman]

And here are the measurements moving away from the front wall.

Left and right speaker measurements.

Blue Trace=38% Position
Green Trace=3 inches closer to rear wall
Red Trace=6 inches closer to rear wall
Yellow Trace=9 inches closer to rear wall
Black Trace=12 inches closer to rear wall

[Magickman]

At this point, having moved the mix position 12 inches closer to the front wall, we moved the speakers 12 inches closer together in order to maintain our equilateral triangle. This actually had a very positive impact on the FR. As you can see, the huge null at 100Hz is considerably better. This was my first clue as to what was causing the null. Hint: this room has vaulted ceilings (more on this later when we get to the subwoofers).

Black Trace=Original speaker position
Red Trace=New speaker position

[Magickman]

At this point, we started adding treatments to the room.

The purpose of these treatments is mainly to treat modal issues in the room, but we are also simultaneously creating a RFZ. The room model we are attempting is a modified LEDE/RFZ type room.

If you examine the drawings in my second post, you'll see that these traps are filled with fiberglass (actually, the smallest triangular traps are filled with OC703; the larger triangular traps are filled 2.5 lbs/CF rockwool, and the huge square corner traps are filled with regular old fluffy glass), and then faced with a perforated panel. For this next set of measurements, the perforated facing panels are left off and we are basically just adding heaps of wool to the room (with the exception of the back wall, which is getting treated with Helmholtz devices and RPG QRDs in order to get a "return" of the back wall). I'll talk more about the perforated panel facings later on.

In my next post I'll strip everything away accept the clouds and then add treatments one by one and show the corresponding waterfall so you can see the cumulative effect of the treatments on the room modes. But for the sake of clarity, I'm first going to attach a few drawings of all the traps in place so you can see where we'll end up.

[Magickman]

Okay, here we go.

The first drawing and measurement is with only the clouds in place. All subsequent measurements are with clouds in place even though they aren't depicted in the drawings. So I'll show the drawing depicting what we've added, and then immediately follow it with the pertinent waterfall plot.

BTW, there is one treatment I neglected to mention which is nothing more than a 3 foot high stack of OC703 in the front wall/floor corner (right in front of the console). This will most likely be replaced with a tuned absorber of some kind, and we'll probably build something similar for the same position on the back wall (to treat length mode of ~32Hz).

[Magickman]

Okay, we're getting close to being all caught up to where the room currently is, and then we'll be "real time" baby!

The next battery of tests I did were to try and optimize the two subs. Everything you've seen so far is without the subs incorporated, and I found that incorporating the subs really helped to smooth things out while simultaneously causing new issues.

We'll start looking at ETCs after that.

Stay tuned!!!

[DanDan]

Well done Magickman. A lot of effort on your behalf but a balanced and systematic approach. I am particularly pleased to see the improvements from changing speaker position alone, often the only variable available in rented rooms. Also interesting is the sort of hybrid generic traps. The holes look big so it appears to be leaning towards broadband. It is good to see such an inbetween design, taking advantage of a bit of this and a bit of that.

DD

[Magickman]

Originally Posted by DanDan

Well done Magickman. A lot of effort on your behalf but a balanced and systematic approach. I am particularly pleased to see the improvements from changing speaker position alone, often the only variable available in rented rooms. Also interesting is the sort of hybrid generic traps. The holes look big so it appears to be leaning towards broadband. It is good to see such an inbetween design, taking advantage of a bit of this and a bit of that.

DD

Thanks DanDan. Yes, a lot of work! Just organizing all the data was a huge task. Keep in mind, you're only seeing a fraction of the measurements taken. Unfortunately (for me), I'm kinda crazy that way. I feel compelled to explore every possibility no matter what I'm doing.

I'm working right now, but I'll be back home later to continue where I've left off.

-MM

[Magickman]

Here's a photo of the rear wall treatments. This is not the studio, simply a pic I snapped at my house. I was trying to get an idea of how it was going to look.thumbsup

BTW, I didn't build the treatments in this pic. The triangular traps were the inspiration for my own design. As stated above, they are a type of perforated/HH device designed by noted studio designer Chris Pelonis. Mine look essentially the same.

[Magickman]

Originally Posted by johndykstra

Magic,

were the faces of the front corners added on the last graph, or second to last?

is the last graph only the addition of the front middle stack?


Beautiful by the way!

As stated above, there are no faces in place for any of these measurements. Please be sure to read the entire posts as it will get confusing otherwise.

Originally Posted by Magickman

For this next set of measurements, the perforated facing panels are left off and we are basically just adding heaps of wool to the room (with the exception of the back wall, which is getting treated with Helmholtz devices and RPG QRDs in order to get a "return" of the back wall). I'll talk more about the perforated panel facings later.

Adding the faces is where things start getting interesting. I'll post more when I have time to devote to it, probably tonight.

In answer to your last question: Yes, last measurement is with the stack in front of the console added.

Thanks for joining in. Things will get interesting once I get you all caught up with where we're at.

Cheers!

-MM

[nixoblivion]

beautifull design! Never seen something quite like it

Btw. about your listening position. Im not quite sure why you chose 12" infront. As far as i see it, the 3rd measurement in the first post seems to be the best to me (especially the 100hz region!!) and that would only be 9" right?

[Magickman]

Originally Posted by nixoblivion

beautifull design! Never seen something quite like it

Btw. about your listening position. Im not quite sure why you chose 12" infront. As far as i see it, the 3rd measurement in the first post seems to be the best to me (especially the 100hz region!!) and that would only be 9" right?

Holy crap. Good eye! That was the wrong graph, which I have now corrected. Thanks so much for paying attention!!!

And actually, the third graph would be six inches closer, not nine, as the first graph is at the 38% point, with each successive measurement being three inches closer to the front wall.

The graph that you chose was six inches closer with the speakers moved 12 inches closer together, and is indeed much better (which I show in post #17). Thanks for spotting that error. As I said, I have corrected that post. I really do apologize for the confusion.

If you look at the superimposed graphs in posts #15 and #16 you will see that 12 inches closer is indeed better.

Here are the comparisons again, but with only two traces per graph for easier reading. In all of the following graphs, the black trace is at 12 inches closer, and in all cases, the black trace looks better to me.

Thanks again for being such an astute observer! You get a gold star!

Blue Trace=38% Position
Green Trace=3 inches closer to front wall
Red Trace=6 inches closer to front wall
Yellow Trace=9 inches closer to front wall
Black Trace=12 inches closer to front wall

[Magickman]

Having now determined the optimal listening and speaker positions, we began experimenting with the subs. Before I get to that though, let me give you a little more info on our monitoring system.

The speaker system we are using is a 2.2 setup comprising Focal Twin6 Be mains and two Focal Sub6 subwoofers. Both the Twin6 Be and the Sub6 offer a number of controls for tuning and integrating. One such control on the Twin6 offers the ability to flip the position of the mid and bass drivers. Early on, I had done some subjective listening tests and determined the position that I felt gave the best imaging; in this case that was with those drivers in the "reversed" position. All subsequent tests to this point were done with the speakers in that position.

I decided to run some tests with the drivers in the "normal" position just to confirm my subjective experience. As it turns out, this "normal" position measured much better.

Here's the graphs for the left speaker. The right speaker shows a very similar result, so I won't bother posting that comparison. You will note that these graphs look a bit different than the previous graphs; this is due to various small changes in control room furnishings, speaker placement, etc. But these measurements do represent the current state of affairs. Once again, these measurements are without the subs, and without the perforated panel covers. Additionally, the scale has changed a bit from the previous graphs.

Blue trace is the "normal" position.
Purple trace in the "reversed" position.

[Magickman]

I did do a number of measurements in different listening positions with the speakers in both the "normal" position, and the "reversed" position, and in all cases the response was better in the "normal" position.

Here's a comparison of the 5 listening positions from the 38% point, moving 3 inches closer to the front wall with each subsequent measurement. I know it seems redundant, but this is basically just to confirm my previous findings.

Once again:

Blue Trace=38% Position
Green Trace=3 inches closer to front wall
Red Trace=6 inches closer to front wall
Yellow Trace=9 inches closer to front wall
Black Trace=12 inches closer to front wall

[Magickman]

Okay, at this point I want to introduce my findings incorporating the subwoofers, and then we'll look at ETCs. Things start to get interesting now.

Here's the left channel before and after adding the sub. LP filter at 100Hz, and the main's HP filter set at 100Hz.

[Magickman]

As you can see, the huge null at 100Hz has been completely erased. Of course, now there's a huge bump at 100Hz. My conclusion is that the 100Hz null was being caused by SBIR from the ceiling! This conclusion is also reinforced by the fact that (as seen in my previous graphs) moving the speakers closer together improved this null as the ceilings are vaulted.

That's one hell of a strong endorsement for running a 2.1/2.2 system!

So, now to deal with that big bump at 100Hz I began lowering the LP filter on the sub. I also reduced the volume on the sub slightly to find the best compromise for the small dips at 55Hz and 75Hz.

Just like magick!