Attn. Acousticians: Help with Ideas and Methods for Monitor Positioning

[DarkEcho]

I had some ideas and questions regarding monitor placement and how one can acheive the best positioning.

I understand that the ONLY way to find the "best" position is through measurement, and not by what sounds good. Often, what sounds *good* is bad for mixing. So we aim for the flattest frequency response we can achieve. This is something hard for us to hear, so we rely on measurement tools to give us a visual aid in the form of a graph.

Ethan Winer recommends ETF for Windows and FuzzMeasure for Mac. I have used FuzzMeasure and find it to be a great tool with a fantastic interface.

Now, the only way to test properly is to move the speakers around while maintaining the equilateral triangle with the points of the triangle emanating from the tweeter cone to just behind your head (so that the lines connecting the three points pass right by your ears). This proves to be a somewhat difficult thing to accomplish, especially if you are doing it yourself. It can be a very time consuming process full of dimensions, scratch paper, and potentially disorganized Graphs of various positions.

This threads purpose is to propose some ideas to the Acousticians here, as to whether certain methods might help reduce the confusion and exhaustion from testing various locations.

With that said- Here are some ideas I had: Ethan, Glenn, James, (and anyone else!) please let me know if any of these ideas will or won't work!

1.) Is it possible to just set up one speaker facing the appropriate direction and disconnect the other (so we dont have to move twice as many speakers and write down twice as many dimensions). Will the results of the graph be roughly the same when a second speaker is added to the mix? This way, we can move one speaker around and find the best location, then we can just reciprocate the position to the other speaker. I understand that this method may be shot to hell if the reflections from the second speaker are necessary in that they produce the potential for cancellation and nulls that wouldn't otherwise be present with just one speaker.

The rest of these can be applied to one or two speaker testing, but I will assume the answer to 1.) Is No, you must have both speakers going and positioned for accurate measurements.

2.) Does the angle that the speakers face affect the bass response? I assume yes, but just checking. If you have both speakers aimed parallel to each other towards the back wall, would it give a different response than if they were toed in to properly create the equilateral triangle?

3.) Is the 38% rule an absolute? if so, this would remove one important variable in the formula for best speaker placement: listening position. if not, we can look forward to twice as many measurements!

4.) Instead of "sliding" the speakers along their equilateral triangle axis, could we just move the speakers horizontally (left and right), while measuring for the best "x" location, then vertically (further or closer from the listening position), while measuring for the best "y" position, and then just plop the speakers in on that [x,y] coordinate?

Now that we have that out of the way. If the answer to all of these potential "short cuts" are No No NO NO NONO!!! Then perhaps there is a way to make sliding the speakers around easier.

At first I thought of an elaborate track system but this would be too difficult to recreate for most of us. Then I thought of a simple plywood board and some of those furniture coasters to aid in sliding things around. then if we draw or tape out a straight line that represents the equilateral axis, we can lay the board on the floor and slide the monitors along the taped lines. The only downside I can see to this is the reflections caused by the plywood might skew our graph. Perhaps if the wood/surface is thin enough the more important bass frequencies will be left alone and we can still capture an accurate representation of the response?

Thanks for your time!

[bpape]

The equilateral triange and 38% are starting points. Usually between 33 and 38% works out best. You can certainly start equilateral and then just slide the speakers out 2" each and remeasure - don't kill yourself with toe-in just yet - won't change bass response that much. You can do the same thing with distance to the wall behind the monitors.

If you can get Room EQ Wizard to run on the mac (it's supposed to work fine), there is an RTA function on it that allows you to measure in real time as you move things to make life a LOT easier.

Bryan

[Ethan Winer]

Originally Posted by DarkEcho

It can be a very time consuming process full of dimensions, scratch paper, and potentially disorganized Graphs of various positions.

Easy solution - buy a six-pack and invite two friends over to move your speakers while you measure repeatedly. Even if you buy Heinekin or Bass Ale the cost will be more than reasonable.

(BTW, this is an absolutely serious answer.)

--Ethan

[DarkEcho]

But Ethan, I live out in the boonies so, even if I had friends (gave them up for mixing) they wouldnt want to come all the way out to help me!

What do you think about my ideas?

[jwl]

Hi Dark Echo,

In response to your questions:

1. I'd run both speakers at the same time. You won't ever listen to just one speaker (except maybe occasionally to check mono), so why test the room with just one speaker?

2. As bpape says, angling the speakers won't affect bass much. When bass leaves the speaker, it is essentially omnidirectional (think ripples in a pond from a stone).

3. The 38% is NOT an absolute, but rather a theoretical starting point. Start there, and experiment to see where it sounds good.

4. Why is it difficult to slide the speakers along the equilateral triangle? Listen to bpape.... put them in one place, and measure. Then move them 2-3" back, and repeat. Do this a few times until you've covered all the possibilities for speaker placement, and compare the results.

Also, once you've determined the flattest spot(s), spend some time listening at each position. If you have 2 spots that both measure similarly flat, but one of them just sounds better, well, there ya go.

To make it easier on yourself, mark the floor with masking tape, and move the speaker stands along there. Write on the tape itself a mark so you can move them back to the tested position once you have the data.

[DarkEcho]

Where should I start though?

If I had an absolute listening position and then just had to slide the monitors around along their axis, that would be one thing, but how do you test different listening positions? you'd have to do tests for each listening position.

In my room, if I moved the monitors 2 inches along their axis for each test, I would have to test about 10 times per listening position, and there would be about 10 listening positions 2" apart... so I'd have to do 100 measurements?!

Is there any way that I can confidentally rule out one of these variables via room dimensions (like the 38% rule which isnt really a rule now apparently ) or something?

Do I really need to compare 100 graphs?

is there a way that I can test for the best sized triangle and then apply that triangle to multiple listening positions?

also, should I remove any book cases or other desks in the room while testing?

EDIT: By the way, I did not realize that JWL = James! How's it going? This is Hawk- the guy who bugs you about this stuff all the time via Email. (I thought I'd spare you this question but thanks for the assistance!)

[magicdog]

For a decent stereo perspective you need to have the monitors equidistant from the sidewalls if possible - without this your stereo will be unbalanced.

So, work out the best distance from the back wall for bass response (too close = boomy, too far = bass light), then, assuming the info above is being followed, it should be a simple matter of deciding the listening position at around 38% from the back OR front wall and creating your equilateral triangle...

[jwl]

well said, magicdog.

DarkEcho, obviously you want to get this right. That's a good thing. But don't overthink it. Sit and listen. Find a good spot or two that you like, by ear. The recordings should be mixes that have been mastered and that you know very well, using 24 bit files if possible. Once you find a few spots that you like, test those spots. Just a few tests should be necessary to give you an idea of what's going on.

Once you have a favorite spot, it's good to know where the peaks and nulls are in that setup, so you can budget for them while mixing. For example, if you know you have a null from 75-95Hz, you won't add too much of that when EQing.

Once you find the right setup, acoustic treatment will have another huge effect on your sound.

[DarkEcho]

Yes, I apologize. I have a horrible tendency to over analyze and critically think about everything.

It all comes down to the fact that there IS a best listening position in my room, and because moving objects around is the freest method of achieving better sound, I get a little OCD on finding that spot..

So when measuring for the best bass response, I am looking for the least amount of variation in level between frequencies (shallower dips and peaks) and preferably an area where there is less variation between the level of a given frequency and the average level of the bass frequencies (IE, no rolling off or heavy sloped graphs).


Am I looking for the flattest, straightest, and most horizontal trend in my graphs?
And what is a good range of frequencies to measure? 0hz-300hz?

[avare]

Originally Posted by DarkEcho

And what is a good range of frequencies to measure? 0hz-300hz?

Your joking. Right? No room dimensions, not indication of what, if any acoustic treatment in the room. If there is, how low the materials are effective sound absorbers.

Andre

[DarkEcho]

did we lose some posts?