use of a sub in a small room

[axemanchris]

Hi there; (follow-up from "very small room - mixes translate a loss in high frequencies" - http://www.johnlsayers.com/phpBB2/viewt ... =3&t=17090 )

I posted this as an update to another thread, but it was in the acoustics subforum, and I think this is probably a better place for it.

My room is 8' by 9'6" by 6'10". It is in a basement, so one wall is concrete block covered by rigid foam insulation, furring strips and drywall. Another wall is wooden slats that represent a partition wall between the studio and fruit cellar. The other two walls are drywall over 2x4 framing with Roxul insulation.

I have built and installed acoustic panels based on input from these very forums. Panels are on 1x3 furring strip frames. Semi-rigid fiberglass (similar to OC703) 2" thick in two layers for a total of 4" of fiberglass. Wrapped one layer with 6 mil vapour barrier (sides and front, but not underneath), and wrapped with a breathable cloth (cotton/poly blend). They are stood out from the wall using 3.5" lengths of ABS drain pipe so that the fiberglass is out 4" from the wall.

Three panels across the front wall. 2x2, 1x2, and 2x2.

A pair of panels on either side. 1x3

I built a 4x3 cloud by adding 4" of semi-rigid fiberglass onto a sheet of 1/2" plywood, covered in vapour barrier and then covered in cloth. The front is lower than the back, so it slopes upwards and away from the mix position, ending directly above the mix position, at an angle of 11 degrees.



Behind me are two panels (18" x 24"), one that is 12" x 12", and one that is 12" x 24".



Here are the frequency graphs of my room:

Frequency graph: blue = before adding cheap sub from a home stereo "kit" and before adding room treatment. red: with cheap sub and after room treatment.



I'm thinking about getting a good quality subwoofer - either a Yorkville YSS10 or one of the new Mackie MRmk3 10's.

Will this help with some of the lacking low end I'm experiencing, or will I just be opening the door to more trouble?

Any suggestions?

Thanks!
Chris

[Soundman2020]

Well, clearly that cheap sub is NOT doing you any favors at all! It looks rather like a typical "one note sub", that has a tuned port on it that turns practically any note into whatever frequency the port is tuned to. If you do get a sub, then be sure to get a really good one that produces a good range of frequencies at similar levels. BUt it's also hard to tell exactly what the sub did, since the "after" graph shows the result of both the sub and the treatment. It would be far more meaningful to have a pair of "before" and "after" graphs where the ONLY difference in the room, is the sub.

Also, you seem to be using 1/3 octave smoothing on your graphs, so it is impossible to see what the sub or treatment really did, in any detail. Smoothing hides all the modal response information, as well as a lot of other detail. If you post the actual MDAT file then we can do the analysis in more detail and figure out what is really going on.

But to answer your basic question: Yes, a good sub installed correctly and adjusted correctly will extend the low frequency response of your system, and fill in the details of what is missing right now. However, it might also excite low frequency modal response more strongly than at present, so you might need to beef up your bass traps too. Your waterfall plot shows that there is still a lot of modal stuff going on in the bottom end, so beefing up yor bass traps would be a good idea anyway!


EDITED TO ADD: Actually, looking at your photos, I don't see any bass trapping at all in that room! So you definitely need to fix that ....

- Stuart -

[axemanchris]

Hi Stuart!

Thanks! You were a huge help to me in my other thread here: http://www.johnlsayers.com/phpBB2/viewt ... =3&t=17090

Well, clearly that cheap sub is NOT doing you any favors at all! It looks rather like a typical "one note sub", that has a tuned port on it that turns practically any note into whatever frequency the port is tuned to. If you do get a sub, then be sure to get a really good one that produces a good range of frequencies at similar levels. BUt it's also hard to tell exactly what the sub did, since the "after" graph shows the result of both the sub and the treatment. It would be far more meaningful to have a pair of "before" and "after" graphs where the ONLY difference in the room, is the sub.

I expect you're absolutely right about the cheap sub I've got. It is the sub that went with a setup like this:


I'd be expecting much greater things from either the Yorkville YSS10 http://www.lamusic.ca/product-p/yss10.htm or the new Mackie MR10Smk3 http://www.lamusic.ca/Mackie-MR10Smk3-1 ... 10smk3.htm

I'll see if I can run a quick test in my room without the sub. I can't really re-run it with the sub, as the power amp I am using is a bit weird. I can't balance the sub with the mains to save my life with this thing. (another advantage of getting a proper sub that will power my mains on its own...)

Also, you seem to be using 1/3 octave smoothing on your graphs, so it is impossible to see what the sub or treatment really did, in any detail. Smoothing hides all the modal response information, as well as a lot of other detail. If you post the actual MDAT file then we can do the analysis in more detail and figure out what is really going on.

I'll see if I can find the original file. It might be gone, though. In any case, when I run another test without the sub, I'll upload the file and post the graph with maybe 1/6 octave scaling.

But to answer your basic question: Yes, a good sub installed correctly and adjusted correctly will extend the low frequency response of your system, and fill in the details of what is missing right now. However, it might also excite low frequency modal response more strongly than at present, so you might need to beef up your bass traps too. Your waterfall plot shows that there is still a lot of modal stuff going on in the bottom end, so beefing up yor bass traps would be a good idea anyway!

EDITED TO ADD: Actually, looking at your photos, I don't see any bass trapping at all in that room! So you definitely need to fix that ....

Bass traps don't really seem to be an option for me. I have, like almost NO corners.

SE corner - The distance between the rear-most wall and the front of the build out (hides the water meter, etc.) that contains the built-in shelves is about 15 inches. I could put something in there. There is only about five inches at best between the shelving and the other wall, though. (pic was taken before I added the panels as shown below)


NW corner - The closet door opening is about 6 inches from the corner, all the way up... and then there's the window...


NE corner - the swing door opening is about 3 inches at best from the corner.


SW corner - could use this corner.


Thinking of placing traps horizontally:

Across the west wall there is the window in the corner. I suppose I could mount something in front of it.

Across the east wall is the sloped 45 degree bevel that comes down to cover the beam across our basement. Below that leaves about 2 inches to the top of the door.

Across the north wall is the closets. I have no more than about 10 inches at the deepest above either of those closets.

Across the south wall is the top of the fruit cellar door, and then the build-out for the duct work and stuff. Not much room there either.

... and if symmetry is a desired feature, well.... haha.

It seems from my limited knowledge that I'm pretty much rooked for any kind of corner traps.

Thanks for your help! I'll post those results later on.

In the meantime, any other suggestions? Would I be doing well to take advantage of a $50 savings on one of those subs and figuring out the details as I go, or should I best hold on?

Chris

[Soundman2020]

I expect you're absolutely right about the cheap sub I've got. It is the sub that went with a setup like this:

ummm......

I'd be expecting much greater things from either the Yorkville YSS10 or the new Mackie MR10Smk3.

I think those would be somewhat underpowered for your room and a decent set of mains. If your budget can stand it, I would suggest moving upscale a bit, such as to one of these:

http://www.lamusic.ca/JBL-JRX118SP-18-P ... x118sp.htm
http://www.lamusic.ca/Yamaha-MSR800W-15 ... sr800w.htm
http://www.lamusic.ca/Yamaha-CW115V-15- ... cw115v.htm
http://www.lamusic.ca/Yamaha-DXS12-12-A ... /dxs12.htm
http://www.lamusic.ca/Electro-Voice-ZXA ... a1-sub.htm

Any of those should be more in line with what you need. That store doesn't stock them, but you could also consider an Adam SUB10, or maybe something from Genelec.

It's not that you you'll be using all the power that those guys can put out, but the thing with subs is that if it can pump out high levels before it starts straining, then it can handle lower levels easily, half asleep, blindfolded, and with one hand tied behind its back! A small sub might be able to do the job, but you'll be running it at the limit all the time, it will be straining, and you'll hear the difference. A speaker being pushed to the limit really does sound like it is being pushed to the limit, and the sound is ugly!

It's sort of the same as with a car: A car with a small 1400 cc engine really might be able to get up to 100 MPH, but the poor engine will be screaming and complaining all the way, whereas the exact same car with a 2200 cc engine can move it at 100 MPH without even breaking a sweat, all day every day, up and down hills, without any problems, and therefore it does so smoothly and forever.

So if you can stretch your budget a bit, go for something a bit heftier.

In any case, when I run another test without the sub, I'll upload the file and post the graph with maybe 1/6 octave scaling.

It's best not use any smoothing beyond about 1/24th octave. Modal issues are normally pretty high-Q peaks and nulls, so smoothing can easily hide them, or merge a couple together unnoticed.

Bass traps don't really seem to be an option for me. I have, like almost NO corners.

Bummer! I see what you mean....

OK, have you considered the BOTTOM of the room? The least-used last-resort is the wall-floor corners. People don't normally think of those, but they are just as effective as any other corner! It just looks ugly, though... Perhaps behind the desk, under the window, or along the bottom of the rear wall?

The front left corner looks like a good option, and also the front right corner, at least up to the bottom of the window, as well as across the top of the front wall, between the window and the left wall.

The top of the left wall also looks usable.

I have no more than about 10 inches at the deepest above either of those closets.

10" is way better than 0"!!!! It's not huge, but pretty much anything would be better than what you have now.

... and if symmetry is a desired feature, well.... haha.

Symmetry isn't critical for bass trapping, especially if is behind you. It would be good to keep things symmetrical at the front of the room, though, if possible.

In the meantime, any other suggestions? Would I be doing well to take advantage of a $50 savings on one of those subs and figuring out the details as I go, or should I best hold on?

If you can afford to go for one of the better subs right now, then I'd do that, since there are some pretty good discounts on those that might disappear after Christmas, but 703 will always be about the same price.... So shop around and see if you can get an even better price on one of those subs, then with what you save you can go buy a few panels of 703 and throw those in the corners too!

- Stuart -

[axemanchris]

ummm......

I think those would be somewhat underpowered for your room and a decent set of mains. If your budget can stand it, I would suggest moving upscale a bit, such as to one of these:

http://www.lamusic.ca/JBL-JRX118SP-18-P ... x118sp.htm
http://www.lamusic.ca/Yamaha-MSR800W-15 ... sr800w.htm
http://www.lamusic.ca/Yamaha-CW115V-15- ... cw115v.htm
http://www.lamusic.ca/Yamaha-DXS12-12-A ... /dxs12.htm
http://www.lamusic.ca/Electro-Voice-ZXA ... a1-sub.htm

Any of those should be more in line with what you need. That store doesn't stock them, but you could also consider an Adam SUB10, or maybe something from Genelec.

It's not that you you'll be using all the power that those guys can put out, but the thing with subs is that if it can pump out high levels before it starts straining, then it can handle lower levels easily, half asleep, blindfolded, and with one hand tied behind its back! A small sub might be able to do the job, but you'll be running it at the limit all the time, it will be straining, and you'll hear the difference. A speaker being pushed to the limit really does sound like it is being pushed to the limit, and the sound is ugly!

It's sort of the same as with a car: A car with a small 1400 cc engine really might be able to get up to 100 MPH, but the poor engine will be screaming and complaining all the way, whereas the exact same car with a 2200 cc engine can move it at 100 MPH without even breaking a sweat, all day every day, up and down hills, without any problems, and therefore it does so smoothly and forever.

So if you can stretch your budget a bit, go for something a bit heftier.

Yowza.... they sure do get expensive. Now, the Yorkvilles that I'm using for mains are 70W. I really don't push them at all, really, to get up to comfortable mixing levels or even fairly loud levels. If I can't go up to doubling the price point I was looking at, then the 250W Yorkville would be a much better choice than the 120W Mackie? (hard to say how they measure it, though, because the Yorkville also says 95W under the program power rating, 20W HF Program power and 75W LF program power.... ???!) It's funny that you mention "underpowered for your room" when my room is so small. Or should I really, really consider saving up for one of the more beastly ones? (also given that I have limited space to put something like that.... Even the 15" one at $539 will consume the entire space under my desk except for where my computer is, pumping out bass about 8" from my kness...) I mean, I will if I will be really glad I did in the end, but if it is only a case of "ideally" vs "perfectly usable", then I might hedge on the side of less money for more space.

It's best not use any smoothing beyond about 1/24th octave. Modal issues are normally pretty high-Q peaks and nulls, so smoothing can easily hide them, or merge a couple together unnoticed.

Okay, so here is my SPL graph from today, with no subwoofer, smoothed to 1/24 octave.

dec 9.jpg

Here is a waterfall from the same measurement:

dec 9 waterfall.jpg

And here is the .mdat file.
https://drive.google.com/file/d/0BwMDmd ... sp=sharing

OK, have you considered the BOTTOM of the room? The least-used last-resort is the wall-floor corners. People don't normally think of those, but they are just as effective as any other corner! It just looks ugly, though... Perhaps behind the desk, under the window, or along the bottom of the rear wall?

As long as it can be broken up to make room for the side panels of the desk, I could put one along most of the bottom of the front wall underneath that window and behind the desk. I'd have to stop it about a foot from the corner under the window to allow the closet door in the corner to open. If that's all okay, I could make one that would come out about a foot from the wall and go up, I guess, a foot up the wall.

The rear wall is where I put my guitar amp, and there is not really anywhere else for it to go. Besides, there is only 33" between the build-out and the door.

The front left corner looks like a good option, and also the front right corner, at least up to the bottom of the window, as well as across the top of the front wall, between the window and the left wall.

Front left corner is fine. Front right corner only has (measured it this time) 9" from the corner to the closet opening. Top of the front wall between the window and the left wall is fine.

The top of the left wall also looks usable.

There is about 20" from the corner to the fruit cellar door (the first reflection absorbers are mounted to the door), at which point, there is only 6" clearance between the top of the door and the ceiling.

10" is way better than 0"!!!! It's not huge, but pretty much anything would be better than what you have now.

As it turns out, the right wall is more like 7" from the ceiling to the top of the first closet door, and 3.5" from the ceiling to the top of the second closet door.

Here is a "nearing construction completion" photo of that wall...

If you can afford to go for one of the better subs right now, then I'd do that, since there are some pretty good discounts on those that might disappear after Christmas, but 703 will always be about the same price.... So shop around and see if you can get an even better price on one of those subs, then with what you save you can go buy a few panels of 703 and throw those in the corners too!

- Stuart -

Good point. I can get the Yorkville or the Mackie today. Anything else will have to wait a while... end of February maybe.

Thanks for your help! This is really useful!

Chris

[Soundman2020]

Okay, so here is my SPL graph from today, with no subwoofer, smoothed to 1/24 octave.

There you go! See all the nasty stuff the smoothing was hiding? You have a big modal dip at just under 96 Hz, and a peak at 148? They were nowhere near as obvious as they are now, and what looked like a single peak at around 60Hz is now clearly a few different things, at 40, 46 and 65 Hz.

You also have a problem with grounding: There are two "tails" in the waterfall plot, at 60 Hz.and 120 Hz. that do not decay over time. They are at a constant level, so they are not acoustic issues. That is "mains hum" getting into your system somewhere. Probably lack of grounding, or a faulty mic cable, or some such. Worth tyring to find and fix that.

Now take a look at the RT60 graph, and you'll see that you have a bass problem in your room. The decay time rises from less than 200 ms for the rest of the spectrum, to over 400 ms below 90 Hz.

Also take a look at the IR graph: You have some pretty large reflections hitting your listening position within 2 ms of the direct sound, so those are most likely coming off your desk or your video screens. To check that, pile up a few thick towels on your desk and repeat the test: I'm betting those will go away, and the mid range will smooth out a bit....

So much more info, just from removing the smoothing!

But you need to calibrate REW properly, using a sound level meter: According to the file you posted, you did those tests at a whopping 140 dB, peaking at 150 dB! If that really were the level you used, the you would be a red smudge on the floor, and your house would be splinters spread around the neighborhood! That's the level of sound you normally find inside the tailpipe of a 747 jet engine, at maximum takeoff thrust.... So I think your REW calibration is just a tad off! You should be measuring at about 83 dB for each speaker by itself, or 86 for both together.

You need a sound level meter so you can calibrate REW properly to get those levels right. Right now I can't tell you much about the actual response of the room, since I've never tried to measure one inside a jet engine before... ... even though the graphs are showing the right information, proportionally, it is totally wrong in absolute terms. So I know there's a modal dip at 96 Hz, and I know it is big, but I don't know how big, because the reference point is way off...

So please calibrate REW using a hand held sound level meter, then repeat the measurements. Also, please to one for the left speaker alone, one for the right speaker alone, and one for both together, all in the same file. Label each test right after you do it, so you don't get confused as to which is which.

As long as it can be broken up to make room for the side panels of the desk, I could put one along most of the bottom of the front wall underneath that window and behind the desk.

That would be fine.

I'd have to stop it about a foot from the corner under the window to allow the closet door in the corner to open. If that's all okay, I could make one that would come out about a foot from the wall and go up, I guess, a foot up the wall.

That's fine too. Right now, anything at all will be useful to some extent. It's not going to tame your modes entirely, or course, but it will make some difference.

Front left corner is fine. Front right corner only has (measured it this time) 9" from the corner to the closet opening. Top of the front wall between the window and the left wall is fine

OK. So just do as much as you can get in.

Here is a "nearing construction completion" photo of that wall...

Yup, that's a problem that needs fixing, for sure! Symmetry....

If I can't go up to doubling the price point I was looking at, then the 250W Yorkville would be a much better choice than the 120W Mackie? (hard to say how they measure it, though, because the Yorkville also says 95W under the program power rating, 20W HF Program power and 75W LF program power.... ???!)

I think I'd trust the Mackie rating more than the other one! Mackie isn't known for exaggerating their specs. They tend to be pretty accurate, and the math works out for their speakers!

It's funny that you mention "underpowered for your room" when my room is so small.

Well, yes, but bass needs a lot more power than mids and highs, simply because it is low frequencies. Are our ears are much more sensitive to mids than lows, so it takes more power to get the same perceived level. Plus, modes, treatment and furniture can all suck out a lot of bass energy from the room, so the speaker needs to be able to put enough energy in to deal with that. There's also the issue of cone size: it simply isn't possible for an 8" or 10" cone to produce low tones as accurately and as smoothly as 15" or 18" cone. Laws of physics, with things like impedance matching and such like.... And it takes a lot more power to move an 18" cone than it does to move a 10" cone, simply because the 18" cone is pushing and pulling on an awful lot more air. So even if both are putting out the same tone at the same level, the 18" jobbie is working harder to do it, and needs more power. And because it has more power, and more surface area, it does a much better job of transferring the energy into the room. Notice that the 12" Yamaha DSS12 has 950 watts power (peak) behind it and it can produce 131 dB, while the 15" Yamaha has 1000 watts of power (peak) yet it only produces 94 dB... Very similar speakers, very similar power, but one is a 15" and the other is only 12"... And other fun things like that... Lots of factors to consider.

But if your budget only extends to the Mackie for now, then that's what I'd go with. It might do the job just fine, and if not then you can always sell it in a few months and upgrade to one of the bigger boys when you have more money. ...

- Stuart -

[axemanchris]

"You have a big modal dip at just under 96 Hz, and a peak at 148? They were nowhere near as obvious as they are now, and what looked like a single peak at around 60Hz is now clearly a few different things, at 40, 46 and 65 Hz.

I guess I just smoothed it out because any spec sheet for anything doesn't have all those jagged points and stuff, so I figured (in other words, yes, assumed...) that smoothing was just what was done to some extent.

I do certainly see that, and have most definitely seen the huge dip around 96 and the neighboring peak at 148. Those are my biggest concerns, I think. (that, and the funny hump in the bottom.... even without the sub... maybe it won't be a funny hump once I address the big dip around 96hz.)

You also have a problem with grounding: There are two "tails" in the waterfall plot, at 60 Hz.and 120 Hz. that do not decay over time. They are at a constant level, so they are not acoustic issues. That is "mains hum" getting into your system somewhere. Probably lack of grounding, or a faulty mic cable, or some such. Worth tyring to find and fix that.

I've gotta do some more reading on those waterfall graphs. I'm not exactly sure what to make of them. I mean, I do - more or less, but only more or less. I do see the tails that you are talking about, though. Good observation that they are probably grounding issues... 60, and then the multiple at 120. Amazing what those little ECM8000's pick up! I sure as hell don't hear it. haha. In any case, I'm not even sure where to begin with that. Unplug things one at a time - power cables, mic cables, etc. - until it improves?

Now take a look at the RT60 graph, and you'll see that you have a bass problem in your room. The decay time rises from less than 200 ms for the rest of the spectrum, to over 400 ms below 90 Hz.

Also take a look at the IR graph: You have some pretty large reflections hitting your listening position within 2 ms of the direct sound, so those are most likely coming off your desk or your video screens. To check that, pile up a few thick towels on your desk and repeat the test: I'm betting those will go away, and the mid range will smooth out a bit....

So much more info, just from removing the smoothing!

So bass traps will help with the decay time in the bottom end? And adding a sub won't just make it worse?

I'll try the towels or a blanket or something and see if that addresses the early reflections. If it does, I guess I could just make a pad for my desktop to control this? Or would you recommend something else?

But you need to calibrate REW properly, using a sound level meter: According to the file you posted, you did those tests at a whopping 140 dB, peaking at 150 dB! If that really were the level you used, the you would be a red smudge on the floor, and your house would be splinters spread around the neighborhood! That's the level of sound you normally find inside the tailpipe of a 747 jet engine, at maximum takeoff thrust.... So I think your REW calibration is just a tad off! You should be measuring at about 83 dB for each speaker by itself, or 86 for both together.

You need a sound level meter so you can calibrate REW properly to get those levels right. Right now I can't tell you much about the actual response of the room, since I've never tried to measure one inside a jet engine before... ... even though the graphs are showing the right information, proportionally, it is totally wrong in absolute terms. So I know there's a modal dip at 96 Hz, and I know it is big, but I don't know how big, because the reference point is way off...

So please calibrate REW using a hand held sound level meter, then repeat the measurements.

Okay, will do. I figured (yes, assumed again...) that it was more the frequencies relative to each other that were what really counted - not so much the actual levels too. I don't have an SPL meter, but there is an iPhone app that, I think, is at least in the ballpark. I'll try that first.

Also, please to one for the left speaker alone, one for the right speaker alone, and one for both together, all in the same file. Label each test right after you do it, so you don't get confused as to which is which.

Okay, will do.

As long as it can be broken up to make room for the side panels of the desk, I could put one along most of the bottom of the front wall underneath that window and behind the desk.

That would be fine.

I'd have to stop it about a foot from the corner under the window to allow the closet door in the corner to open. If that's all okay, I could make one that would come out about a foot from the wall and go up, I guess, a foot up the wall.

That's fine too. Right now, anything at all will be useful to some extent. It's not going to tame your modes entirely, or course, but it will make some difference.

Front left corner is fine. Front right corner only has (measured it this time) 9" from the corner to the closet opening. Top of the front wall between the window and the left wall is fine

OK. So just do as much as you can get in.

Great! Thanks!

Here is a "nearing construction completion" photo of that wall...

Yup, that's a problem that needs fixing, for sure! Symmetry....

Actually, the construction is done. The door on the left is the one that has one of my early-reflection acoustic panels with the other one mounted on the wall directly beside it. The closet on the right has hollow-core bifolds on it. It's just the photo was taken a year ago.

If I can't go up to doubling the price point I was looking at, then the 250W Yorkville would be a much better choice than the 120W Mackie? (hard to say how they measure it, though, because the Yorkville also says 95W under the program power rating, 20W HF Program power and 75W LF program power.... ???!)

I think I'd trust the Mackie rating more than the other one! Mackie isn't known for exaggerating their specs. They tend to be pretty accurate, and the math works out for their speakers!

...

But if your budget only extends to the Mackie for now, then that's what I'd go with. It might do the job just fine, and if not then you can always sell it in a few months and upgrade to one of the bigger boys when you have more money. ...

I think I was leaning toward the Mackie one too. I guess I'll put in an order.

It's funny that you mention "underpowered for your room" when my room is so small.

Well, yes, but bass needs a lot more power than mids and highs, simply because it is low frequencies. Are our ears are much more sensitive to mids than lows, so it takes more power to get the same perceived level. Plus, modes, treatment and furniture can all suck out a lot of bass energy from the room, so the speaker needs to be able to put enough energy in to deal with that. There's also the issue of cone size: it simply isn't possible for an 8" or 10" cone to produce low tones as accurately and as smoothly as 15" or 18" cone. Laws of physics, with things like impedance matching and such like.... And it takes a lot more power to move an 18" cone than it does to move a 10" cone, simply because the 18" cone is pushing and pulling on an awful lot more air. So even if both are putting out the same tone at the same level, the 18" jobbie is working harder to do it, and needs more power. And because it has more power, and more surface area, it does a much better job of transferring the energy into the room. Notice that the 12" Yamaha DSS12 has 950 watts power (peak) behind it and it can produce 131 dB, while the 15" Yamaha has 1000 watts of power (peak) yet it only produces 94 dB... Very similar speakers, very similar power, but one is a 15" and the other is only 12"... And other fun things like that... Lots of factors to consider.


- Stuart -

So much to learn! Ahhhhh!!!! haha

Thanks again for your ongoing advice!

Chris

[axemanchris]

Okay, ran the tests as suggested. The difference was surprisingly significant, but I'm not entirely sure what to make of it.

So many permutations and combinations, I don't know if posting a graph is all that useful overall, but I'll just post the one for sake of reference for others who might follow this later.

There are some background noises - there always are. I'm wondering how much of a role these play in the results. (dishwasher on upstairs, furnace on, car driving up the street, etc.) You can hear them if you listen for them, but would not be heard if I had music on and was mixing something. Should I wait until nobody is home and turn *everything* off?

Red = today with my desktop clean
Green = today with three thick towels on my desktop.

dec 11.jpg

Here is the mdat file:
https://drive.google.com/file/d/0BwMDmd ... sp=sharing

I'm curious as to what you'll make of this.

Chris

[Soundman2020]

You can hear them if you listen for them, but would not be heard if I had music on and was mixing something. Should I wait until nobody is home and turn *everything* off?

That's the best way of dealing with those noises, if you want really accurate results. Another way that helps a bit is to tell REW to do several sweeps for each test, and average them out internally. On the "Measure" window, there's a field for setting that. That helps for intermittent noises, such as foot steps and doors closing, but not for constant noises, such as the dishwasher or a car idling.

The difference was surprisingly significant, but I'm not entirely sure what to make of it.

Yup! There's a notable difference all over, and now that we have a full set of measurements, there's an awful lot more that can be seen, and needs working on!

First is symmetry, and I suspect the "nook" over on the right. If you look at just the left and right speakers, for both sets (with and without towels), you'll see that in both cases there's a large dip at 95 Hz for the left speaker, which just isn't there for the right speaker, bot for both cases of the right speaker there's a large dip at 300 Hz that isn't there for the left speaker. So something in the room is sucking out energy asymmetrically, with the frequency depending on which speaker is firing. It could be two entirely different things, or it could be two aspects of the same basic problem. I'm inclined to go with the second option: two aspects of those "nooks" over on the right wall. How did you deal with those? What did you do to them, to close them off? One of the reasons I'm thinking of those is because in the "both speakers" graphs, those artifacts disappear, making me think that they cancel each other out perfectly on the room center-line, but since your ears are not on the center line (even though the mic is), your ears will be hearing the individual un-balanced signals. So it would be good to try to trace that issue and fix it.

Next, is the effect that just a few towels is having on the entire room: Those are actually reducing your overal decay time, believe it or not, and are even affecting the modal response: acting like a bass trap, and fairly effective, too, give the small volume. So imagine how effective true bass traps would be, located correctly...

Then there's the original reason why you tried those towels: the reflection. The reflection at about 1ms has gone, so my theory was right there, but the reflection at 2ms has not gone: it is still there. Meaning that it is coming form something else. So here's how you fine it. Get a piece of string, measure the distance from the acoustic center of your speaker to the tip of the measurement mic, and mark that distance plus 25" on the piece of string. Tape the string to the front panel of your RIGHT speaker such that one of the marks is exactly at the front panel, as close as you can get it to the acoustic center, and tape the other to a mic stand set up exactly where the tip of the mic was (don't tape it to the mic itself! Just set up a stand in the right place), such that the second mark on the string is exactly where the mic tip was.

So now you have a piece of string that is exactly 25-1/4" longer (in distance) than the direct line from speaker to mic, and that works out to exactly 1.87ms longer in time, which is the time delay of the reflection, according to REW. So now carefully move that dangling loop of string around to see what surfaces or objects in the room it touches. There will be only one spot in the room where you can get that string to touch perfectly: perhaps on the ceiling, or a wall, or one of your monitors, or a part of the desk that wasn't covered by the towels. Whatever it is that the string touches, is casing that reflection and needs absorption on it. If you want to double-check, then get another piece of string marked at 25-7/8", and use that on the left speaker. The loop should end up touching the exact same spot as the string on the right speaker. Both of them are seeing the same issue, with and without the towels.

So I would start with those items, and try to figure out how to fix them.


- Stuart -

[axemanchris]

That's the best way of dealing with those noises, if you want really accurate results. Another way that helps a bit is to tell REW to do several sweeps for each test, and average them out internally. On the "Measure" window, there's a field for setting that. That helps for intermittent noises, such as foot steps and doors closing, but not for constant noises, such as the dishwasher or a car idling.

I'll have the house to myself on Sunday - maybe even for a bit tomorrow night. I'll kill everything on my next set of tests.

I actually did do four sweeps for each measurement. Maybe I'll try eight just for kicks.

Yup! There's a notable difference all over, and now that we have a full set of measurements, there's an awful lot more that can be seen, and needs working on!

Totally not being sarcastic when I say, "Yay!"

First is symmetry, and I suspect the "nook" over on the right. If you look at just the left and right speakers, for both sets (with and without towels), you'll see that in both cases there's a large dip at 95 Hz for the left speaker, which just isn't there for the right speaker, bot for both cases of the right speaker there's a large dip at 300 Hz that isn't there for the left speaker. So something in the room is sucking out energy asymmetrically, with the frequency depending on which speaker is firing. It could be two entirely different things, or it could be two aspects of the same basic problem. I'm inclined to go with the second option: two aspects of those "nooks" over on the right wall. How did you deal with those? What did you do to them, to close them off? One of the reasons I'm thinking of those is because in the "both speakers" graphs, those artifacts disappear, making me think that they cancel each other out perfectly on the room center-line, but since your ears are not on the center line (even though the mic is), your ears will be hearing the individual un-balanced signals. So it would be good to try to trace that issue and fix it.

I took a video that shows a 360 degree tour of the studio, in case that helps. It's about 1:30. The two closets are closed over with hollow-core bifold closet doors. There is a fair bit of stuff in both closets. The smaller of the two is on the right of the mix position and is where one of my early reflection absorbers is attached. This, possibly, is counterbalanced on the other side by the fruit cellar to my left behind the wall with the wood paneling (not flimsy stuff.... it's about 3/4" thick or more of actual tongue-in-groove wood), seen at the 9:00 position in the video. The door is basically never accessed and the fruit cellar is empty, save for some bare shelves. The other things that significantly (I suppose) impact the symmetry of the room are the window in the corner above and behind the right speaker (seen at 0:23), and the build-out that contains the water meter, where there is a shelving unit built into the corner, as seen in the video at 1:00-1:07.

http://www.youtube.com/watch?v=GaB3dPs83EE

Next, is the effect that just a few towels is having on the entire room: Those are actually reducing your overal decay time, believe it or not, and are even affecting the modal response: acting like a bass trap, and fairly effective, too, give the small volume. So imagine how effective true bass traps would be, located correctly...

Yeah, I couldn't believe how it impacted the top end. I'm thinking that it had some desirable effects, but in some ways, took arguably too much away from the top end?

Then there's the original reason why you tried those towels: the reflection. The reflection at about 1ms has gone, so my theory was right there, but the reflection at 2ms has not gone: it is still there. Meaning that it is coming form something else. So here's how you fine it. Get a piece of string, measure the distance from the acoustic center of your speaker to the tip of the measurement mic, and mark that distance plus 25" on the piece of string. Tape the string to the front panel of your RIGHT speaker such that one of the marks is exactly at the front panel, as close as you can get it to the acoustic center, and tape the other to a mic stand set up exactly where the tip of the mic was (don't tape it to the mic itself! Just set up a stand in the right place), such that the second mark on the string is exactly where the mic tip was.

So now you have a piece of string that is exactly 25-1/4" longer (in distance) than the direct line from speaker to mic, and that works out to exactly 1.87ms longer in time, which is the time delay of the reflection, according to REW. So now carefully move that dangling loop of string around to see what surfaces or objects in the room it touches. There will be only one spot in the room where you can get that string to touch perfectly: perhaps on the ceiling, or a wall, or one of your monitors, or a part of the desk that wasn't covered by the towels. Whatever it is that the string touches, is casing that reflection and needs absorption on it. If you want to double-check, then get another piece of string marked at 25-7/8", and use that on the left speaker. The loop should end up touching the exact same spot as the string on the right speaker. Both of them are seeing the same issue, with and without the towels.

So I would start with those items, and try to figure out how to fix them.


- Stuart -

Okay, I found this a little confusing, but what I ultimately wound up doing (which seemed to be the same thing) is measuring 25-1/4" and 25-7/8" from the microphone and seeing what I could touch with it.

There seems to be two possibilities, with a remote third. My thinking, in order, is:

- my computer monitors (no idea what can be done about that... not like you can cover them... haha) were about exactly that distance, depending on what point on the monitor you measured to.
- my chair, as I recall, was a little further behind me than usual when I measured last night. It could easily have been that distance from the mic. (it's a padded swivel office chair, leather faced, that comes up to just below my shoulders, but the sides stick out... a little...)
- my cloud - a remote possibility, given that it angles away from the mix position, and is actually more like 27" from the top of the measurement mic, and is not really a reflective surface. (fabric, covering 6 mil vapour barrier, covering 4" of semi-rigid fiberglass, mounted on 1/2" plywood)
- the surface of my rack case that houses my interface, etc. - but it's more like 27" from where the top of the mic would have been at its closest point, which is even below (though only by a few inches) the level of my desk top.

I'm starting to think about bass traps.... are you thinking of superchunks (or at least as super as I can make them... hehe), or am I totally getting ahead of myself here?

Thanks!

Chris

[Soundman2020]

Okay, I found this a little confusing, but what I ultimately wound up doing (which seemed to be the same thing) is measuring 25-1/4" and 25-7/8" from the microphone and seeing what I could touch with it.

Actually, that isn't the same thing at all. The reason for the string is that we are talking about a DIFFERENCE in flight time of about 2 ms, which works out to a DIFFERENCE in distance of about 2 feet (sound moves at roughly 1 foot per second). So you can't do this by measuring 2 feet from your mic, or 2 feet from the speaker: it is a DIFFERENCE of 2 feet that we are looking for.

In other words, there are two paths that sound is taking from your speaker to your mic. One is the direct straight-line path, and the other is a path that is 2 feet LONGER, since the sound wave hits something then bounces. That's why you need a piece of string that is 2 feet longer than the direct path. That piece of string represents that second path (with the bounce it it). That's the path that some sound waves are taking: they leave the speaker at an angle that is DIFFERENT from the direct path to the mic, then they bounce of "something" in the room, and that "something" then redirects them to the mic. That path is longer than the direct path, by 2 milliseconds, which is 2 feet.

So the piece of string will be exactly as long as that path that bounces off "something", which is the direct distance plus 2 feet.

If you attach one end of that string to the speaker and the other end to the mic, then the extra 2 feet of slack will sort of hang down in the middle. If you then use your finger to lift up that slack until the string is stretched taught again, there will now be two straight sections of string: one going straight from the speaker to your finger, and the other going straight from your finger to the mic. If you move your finger along the string and around the room, keeping the string taught, then at some point your finger will touch an object or a surface: that's the item that is causing the reflection. It will be located at some point between the speaker and the mic, perhaps on a wall, or the cloud, or the ceiling, or the desk. There will only be one point where this happens.

This is very different from just measuring 2 feet from your mic. Not the same at all. It's a difference in path length of 2 feet.

Not sure if that made it any clearer!

I'm starting to think about bass traps.... are you thinking of superchunks (or at least as super as I can make them... hehe),

Yup!


- Stuart -

[axemanchris]

That clarified it perfectly, thanks! Right, not the same thing at all.

Okay... so the results are... ready for this.... (two likely candidates, in order....)

1. My f***ing forehead.

Yep. I was sitting right behind that mic while I was doing the tests. The mic, set at about ear level, means the tip is at about the bridge of my nose. About a foot behind the mic is probably *exactly* where I was sitting. Should I wear a hat while mixing? Would you suggest a toque, or a Stetson, or something different? I expect a baseball hat will be too reflective. Would the resonating chamber in the Stetson be problematic, or would it act as a bass trap? Oh, my.... so many questions.

2. The computer monitors again. Now, because the computer monitors sit about 2" in front of the front face of each monitor (but actually overlapping the speaker cabinet), the sound would have to travel at an angle of about 160 degrees, just *barely* clearing the edge of the right computer screen, travel over to the left computer screen and bounce off it and to the mic. Now the point where the string touched the left computer monitor would mean the reflection from the computer monitor to the mic would be exactly perpendicular.

Given the placement of my speaker monitors, computer monitors, and me sitting directly behind the mic while doing all of this, that if I repeated this with the 27" extra length on the string from the other speaker, that the two points of possible reflection would be identical.

I don't know if that helps you to help me much or not, but hopefully it gave you a bit of a laugh anyways.

Now, I've moved the computer screens backwards so that they are behind the front edge of the speaker monitors, and so that they are not angled back to me. That way, if sound does travel at 160 degrees from the speaker and bounce off the computer screen, the reflection will be less likely to come back to the mic, and instead be redirected past the mic. Just in case that was my problem.

It does make the monitors less easy to see/read, as they are now further away and at kind of a funny angle. Moving the speakers forward was the other possible solution, but the dust caps are only about 29" from my ears as it is. My perception is that moving them closer is not a desirable option. (and it will, of course, change the way they behave in the room)

I'll run more tests and see if that helped or hindered later tonight.

Thanks!

Chris

[axemanchris]

Okay, I turned off the furnace (-10 celsius here!!) for a little bit, got rid of all the noise I could find except for the computer internal fan.

I ran several tests, and have notated them as follows:

L = Left speaker
R = Right speaker
B = Both speakers
T = 2 heavy towels placed on desk (not three this time)
xt = no towels on desk
P = me just behind mix position (like yesterday, about a foot behind the mic, letting all the joyful noise bounce off my forehead to see if the mics would pick up the reflections.)
U = me under (and sort of in front of) desk (but the chair in position... a mostly uninformed look at the IR graphs is suggesting to me that there may be more reflections probably coming off my chair (about 2' behind the mic) than there ever was off my forehead or any other part of my body before.
MO = monitors in original position (as described above)
MB = monitors back somewhat behind the monitors and flattened out - not angled towards my eyes

I did most permutations and combinations of these. In fact, I did all 12 of the possible combinations with the towels in place, and I did six possible combinations with the towels not in place. I did 8 sweeps

The link to the UPDATED mdat file is here:
https://drive.google.com/file/d/0BwMDmd ... sp=sharing

Hopefully this will lead us to some starting conclusions.

Thanks!
Chris

[axemanchris]

Another thought....

I'm currently using a 23" monitor (2048 x 1152) set at maximum and a 19" monitor (1440 x 900) set at maximum. My video card supports digital resolution to 2560 x 1600, which represents the real estate (pixel-wise) of three times my 19" monitor set to its current resolution.

Acoustically, would I be better off getting, say, a single 27" monitor that will display 2048 x 1152, and placing it in the middle?

Currently, with the two monitor screens in front of me, the dust caps of my speaker monitors are 49" apart (absolutely as close as they can possibly get right now), with my ears about 30" from them. If I move my listening position backwards to the 49" mark from the monitors, I will be almost exactly dead center of my room.

Going to a single video monitor will allow me to reduce the distance between the dust caps to as little as 42". My desk is 32.5", and my speaker monitors are sitting on stands on either side (703, plywood, two concrete cinder blocks, plywood, screws - from top to bottom, then wrapped with fabric).

Just an idea... maybe I should try it with just my 23" monitor at the resolution it is now, and run some tests, but that will really only give me just over half the pixels that I would get by upgrading...

What do you think?

UPDATE: I tried it, and there seem to be some trade-offs, but for the most part, a more moderate curve overall. I've re-uploaded the mdat file with the last two measurements (with and without towels) with the video monitor as far back as I can get it, and with the speaker monitors moved in about as close as I can get them. Instead of the dust caps being 49" apart, they are now 42" apart. I also backed the speakers as close to the absorbent panels behind them as they could go (about an inch on either side between the corner of the speaker and the absorber). The imaging seems better already.

I stood the power amp on its edge, which allowed me to get the monitor even another inch and a half back towards the middle absorber. (the stand sort of wedges underneath the edge of it)

photo.JPG

It looks neater, but I'll just have to get used to using Alt+# keys to change between window sets. That's fine.

Thanks!
Chris

[Soundman2020]

Finally found time to download your REW data and take a look at it!

It's not the screens. Or rather, they do have an effect, but that's not the big issue.

It is also not where you are standing, even though that also does have some effect.

It is the desk.

Here's the issue, taken from your last two tests, which are identical except for one thing: absorption on the desk.

Here is the IR without the absorption :

axemanchris-ir-before-01.png

Now with the absorption:

axemanchris-ir-after-01.png

See the huge difference where the cross-hairs are? There's a big reflection at 1.2ms that is barely 7 dB down from the direct sound when there is no absorption on the desk, and in the second one it is GONE! Not there at all. The first graph shows a big problem, the second shows a useable room.

So let's try this: pull both speakers about 3" or so away from the front wall, towards you, then re-angle the speakers a bit more inwards so they are still pointing correctly (just past your ears, not AT your ears!), and also raise them about an inch or a bit less (put a book under each one for now, just for testing.) Do a REW test like that for both speakers, and post it here.

One comment: you could benefit from having a good sub-woofer in that room!



- Stuart -