Are there any step-by-step guides\white papers on methods which allow for most efficient way to find the spot where lowe-end is most linear? Any methods which allow to narrow down the choice and move in the right direction after any consecutive step made?
I am talking only modal issues 20-200 Hz (however REAL modal behaviour since the mode tables\calculators are for ideal rooms).
Because by just moving a speaker to many spots in a room and taking a measurement one still can miss the best spot since there are millions of possible measurement spots available in even a smal room and without knowing what to look for and when to move to judging by the measurement one still might be searching for a black cat in a dark room.

Every room is different, but a good starting point is to face the shot wall and start by sitting 38% of the room length. There is no "rule" so you will find moving from there may or may not improve things.

There must be some secrets still.
When pro guys tune the room and find the sweet spots for speakers and listening, this must be a series of conscious moves and measurements with every step efficient in finding the optimal position.

You would think so, but they get paid by the hour so they never bothered learning any of that.

Sent from my PC36100 using Gearslutz.com

By looking at the readings you can figure out where the null or peak is coming from (or what is causing it) and move from there. I most admit, I have seen Bryan (our lead room designer) calculate a room and get pretty close, but there are a lot of factors that play into this.

There is a procedural way to do it. An RTA helps during this process too though. A program like Room EQ Wizard helps tremendously as it allows you to see frequency response, decay times, waterfalls, impulse response, etc. while also proving the RTA function. Can't beat it for free...

Push speakers into the corners. This maximally excites the room modes and gain.

Slide the mic back and forth til you find the spot with the smoothest response, fewest nulls, pushes problems as high in frequency as possible, etc.

Once you have that spot, mark it and put the mic there. Then pull the speakers out to a starting point of an equilateral triangle and measure again as a baseline.

From there, move the speakers forward/backward OR farther apart, closer together. This let's you asses the impact of boundary interactions (SBIR). Keep moving until you find the best spot and leaves items that are known quantities and are treatable.

Notice I said OR. It is important to always change only one thing at a time or you will not know which change made what differences.

Another option is a program called CARA. It requires that you model your room and all of it's surfaces. You then define the areas that are acceptable for you to place your seating and speakers and let it rip. It will chug for a while and come up with the best place for everything. This can take quite a bit of time to compute depending on the complexity of the room, how much information you want, the speed of your computer, etc.

I've just purchased this program and as time permits am starting to do my own known room as a baseline.

Bryan

Bryan,
Thank you very much!!! I will try that and post the results back.

Bryan,

This is the first post I've come across that's laid out where and how to start with regards to listening and speaker position. Thanks so much for sharing your expert knowledge!

Going to give this a go this weekend.

Bill

Another less accurate way would be to take an omni mic and map your room into "cubes" or a 3-d grid, say, every foot or so.
You take the mic, your assistant plays a sinewave. You start at one corner, on the floor. Play the sine, start at (just suggesting...) 40, then 50, then .. and so on - it only makes sense up to lower mids) your assistant writes down the level your mic shows. If your grid is one foot wide, then your next point to measure is one foot beside it, next point, another foot further... if you got the lowest "layer" of your room mapped, start again, one foot higher, same location. It helps if you start with mapping the room with painters tape, some pieces of string and whatnot. Yes, it's inaccurate. Yes it's unscientific. Yes it takes some time. But you get a rough overview where there are trouble spots. You need no software and no calibrated mic. It's not about absolutes, it's about relative levels.

You could also grab a floor tom, walk around the room, hit it and listen to where it sounds fuller and where it sounds thinner. This method is borrowed from the book "mixing with your mind".

The best method would be to have a professional help you.. just costs some money. Is totally worth it.

I've gotten into a big project rearranging my garage and doing this stuff. I should post more about the experience when I'm done. I've been using a program called RPlusD (formerly called ETF) and a Behringer ECM8000 onmi mic. RPlusD is cheap. It plays a sweeping sine wave and white noise and then shows you the frequency response. Playing with this over the past few days has given me feel for how the room "behaves." It's fascinating stuff and it was all new to me when I started.

I used this in combination with CARA (to suggest some good starting places.) I ran CARA several times with different settings (man, it's complicated) and got widely varying results each time. Only one of those suggested positions turned out pretty good, the others not. In retrospect, I would have skipped CARA and just laid out the floor on a grid with taped marks, since this is what I wound up doing anyway. CARA was not a silver bullet.

One thing is for sure, wow it really makes a difference where your speakers are and where you are sitting. There is a lot of discussion about what the best monitors are, etc etc, but... imagine slapping a 32 band equalizer on your monitors with 10-15dB peaks and nulls! That's what's going on unless you find the best position and treat your room. Seeing it graphed out on RPlusD is very revealing. It's also amazing how small differences, like moving a speaker 1 ft, can have big effects sometimes.

Yes, young cats just discuss opinions on speakers listened to briefly in different room not understanding that monitors are monitors-all of them workable and relatively flat- and its room\placement that needs to be sorted out first in the 20-300 Hz region.

IMHO CARA is not a good program to do this. And for mid and high frequency you have better but of course more expensive software to model the rooms (I work with CATT mainly)*

For low frequency behavior you can use RPG Room Optimizer, it works of for square or rectangular rooms. If you get into strange rooms then you need more complicated software to work with like ABEC.

Anyway since the room already exists the best way is really to measure it. Place yourself at 38 % and start from there. Place the speakers near the walls and use your ears along with the measurements. Start by moving your position and the speaker's position 2 to 4 inches and see if they reflect any changes.

My experience is if the room is really bad, like a square and not treated the results are not that different from each position. I will send a measurement if I find one

So I pushed the monitors into the corners, taped a line down the center of the room, and measured from 50Hz to 315Hz in 3 inch increments. The mic pointed down that line and was 38" from the floor which was the measured height to the midway point between the woofer and tweeter.

My best result funnily enough was exactly at 38%. I averaged 86dB (+2/-3dB). The speakers roll off fast below 50Hz. At this location they're at 72dB at 40Hz and the most I got out of them at that frequency was 79dB at the very back of the room.

Today I decided to try and figure out the speaker positions by laying out a grid on the desk. In the best locations I've got a null at 160Hz and a peak at 250Hz. Those positions are also nowhere near an equilateral triangle. It's 40" from mic (listening position) to tweeter and 63" from tweeter to tweeter.

Is broadband absorption going to be able to iron this out significantly? Will I be able to pull down remaining peaks with PEQ?

My room is small: 121" wide, 140" long with 107" high ceilings.

38% is a starting point but generally speaking it is a lot of times close to that point (GENERALLY heh)
Broadband is going to help but going to help a lot more with the decay times over frequency response unless REALLY THICK or something tuned. For frequency response you will want the bass trap (tuned or broadband) on the wall where the problem is coming from.
With all of that said decay time is really important to cure as that is what will give you the tightness/clarity in the room. I would recommend starting with bass traps in corners for that. The following is our test/lab room where I tested it empty then with 8 monster traps straddling corners. If you notice there is improvement all the way down to 40hz or so on the decay times and the great deal of improvement at that 70hz area where there was a ton of ringing. Note that I would not call the room done but just showing how much you can improve things without a great deal of hassle. From here I could treat with more broad band in problem areas (frequency response/decay times) or something tuned to not absorb anymore of the upper end.

I'm thinking of building an acoustic soffit 26"H x 16"D x 120"W, using OC703 on the two "faces" and back filling with the pink fluffy stuff placed along the front wall.

Then I'd build 17x17x24 Superchunks in the front corners that run from the floor to the soffit.

Lastly would be to put 24x24x4 panels on the front wall behind the speakers.

Am I on the right track?

I played around with speaker position this weekend and the best I measured was 30 deg spread but I'm getting 7dB nulls @ 100Hz followed by 10dB peaks at 125Hz.

Bill

I was recently reading a NARAS paper on suggestions for surround sound production and one of the things they mentioned was finding the best location for the sub. One method they suggested was to move the sub around the space while taking measurements at the listening location. The alternative method was to set the sub at the listening location and move around the space measuring for smoothest response (which would be the best location for the sub). The alternative method seems like it would be much less labor intensive, and probably save your back some aches.

So after a bit of time away doing other things I finally had some time to sit down and try and measure my room/speaker relationship.

After several failed attempts at trying to get my iMac, Tascam 144mkII, and REW to play together nicely I threw in the towel and got out the Dell laptop and went through the setup and calibration steps again with success!

I was able to make some measurements and it looks like I did everything correctly:





If everything looks correctly done, can anyone help me interpret these? It seems I can use acoustic trapping to smooth out the peaks @ 265, 220, 120, 110, and 50 Hz with some parametric EQ to further pull them down if necessary.

I'm concerned about the nulls @ 180 and 155 Hz though. Will the trapping help that at all or do I need to experiment more with speaker/listener placement?

I've planned on making "superchunks" for the front two corners that would run floor to soffit, and soffit traps running around the entire room. The soffit on the front wall would be roughly 16" x 16" and the soffits on the other three walls would be 16"D x 9"H.

The room is roughly 11'6"L x 10'W x 9'H with a 40"W a 26"L "entryway" into it. It's a home office that I'm using for a computer-based 2-ch music only setup. I've come here to the pro-audio side of things because I like the approach you lot take better than the audiophile approach. Floors are hardwood with an area rug, walls and ceiling are drywall on standard new construction 2x4s. To give you some idea of the room here are a few old pictures:





The speakers have since been moved so that they're @ 60* from the listening position which itself is 38% from the front wall, nearly forming an equilateral triangle.

The speakers are Dynaudio BM5AmkII that are only rated down to @ 47Hz. I have no plans for adding a sub as that would create more issues in such a small room.

Any help would be greatly appreciated.

Bill

Bill, your measurement is probably useful for Modes, but I would view only to 300Hz or so. Other things suggest RTFM also. https://www.gearslutz.com/board/studi...er-v2-1-a.html

The HF rise looks dodgy. What mic is this?
Do you have Inverse C checked in the prefs?

DD

Dan,

The mic is a Behringer ECM8000. My primary concern is getting 50Hz to 400Hz as right as possible for a decent listening experience.

I don't see Inverse C anywhere in the Preferences menu.

Off to RTFM.

Bill

REW/Prefs/Mic Meter/C Weighted SPL Meter.
I did a whole load of Measurements before I realised this was active.
It adds top and bottom unnecessarily. Might explain your HF rise.
Three Measures. L, R, L+R

DD

HA! It was checked because in the directions it said to check it off if using an SPL meter set to C weighting when calibrating SPL! Nothing in there about turning it off to take measurements!



Bill

Hmmm. One way or another OFF would be nice.
The only SLM I would ever use for acoustic measurement is a B&K set flat, no weighting curve. So I am wondering how that box got checked in my REW.
Is it on by default. Measurements taken with it on at least do have an Inverse C notation. Although it's small enough not to notice for some time.....
DD

I think that is a great plan.

Thanks Glenn. What are your thoughts on the measurement results? I can't say I'm surprised at what's going on from 40Hz to 400Hz since my room is close to square 137"L x 121"W x 108"H.

I plan on measuring again after I get the treatments up to see if there are new "best" positions for sitting and for the speakers but I feel like I'm at least in the ballpark at this point.

My latest debate is what to do on the back wall where I have the CD rack and bookcase. Currently the listening position is @ 4' from that wall. I'm out of CD space since that picture was taken. Plan A would be to lose both and put some absorption. Plan B would be to put in one large combination unit that's @ 15" deep and would house the CDs and books and perhaps, in the most vague sense, provide some sort of diffusion?

Bill

You need to turn off the smoothing and view it from 20hz to 400hz. If you would like got to file, save all measurements, zip that file you save and post it here. I can show you how it should be viewed and show you how to view the waterfall plot, which is important to view. More on that from a article I did.

Testing of Acoustic Foam Bass Traps. Acoustic Panels and Bass Traps.

Here's 20Hz-400Hz with no smoothing, both channels driven and then the same with just L&R separate measurements with overlay:

I would move the speakers around a little bit to get them to eve out a little better. Looks like you need bass traps bass traps and more bass traps. Like I said you need to also view the waterfall graph. And you can post the raw file here which I can post the waterfall graph for you.

Yes, bass traps are in my future for sure! I keep getting an invalid file error when I try to upload the zip file of the waterfall.

What should I change in Limits and Controls in order to view it correctly?

Bill

Actually now that I look again you are viewing just fine. Carry on.

So I finally got around to adding my traps, soffit-style traps all around the room and SuperChunk style traps in the front corners and my results look worse to me:








What gives? I now have a huge suckout from @ 108Hz to 173Hz. Is it just a phase issue where I can experiment with listening/speaker positions?

Bill