1. Do I need to get the sub in there before I do anything else? If so, how is it's position in the room determined?
I find the easiest way is to put the subon the floor, up against the front wall, but not centered: a few inches off to the right or left of the center line is normally a good starting point.
The first order of business is to set the cross-over point and the phase of sub. Try something like 85 to 95 Hz as your cross-over, and you set the phase very simply: Do two tests in REW, one with the phase at 0°, the other with the phase at 180°, and see which setting gives you the flattest response in the low end (one setting normally gives you a huge dip in the response, due to phase cancellation at the mix position). Then adjust the relative levels of the mains and subs so they match (no roll-off or increase in the bass level, with respect to the mains).
Then I normally do a bunch of listening tests and REW tests while nudging the sub around in all directions a few inches, to see if it improves or gets worse. In other words, mark the initial spot on the floor where you have the speaker at first, then move it a couple of inches to the right, test, a couple more inches, test, a couple more inches, test, etc. Mark each point as you go, so you can get back to it if you need to (stick masking tape on the floor, and scribble on that)
Then put the sub back at the original spot, and do the same moving it to left. Look at your graph for all of those tests, and also just listen with your head close to the mix position (but without bumping the mic!), and find the two smoothest ones, then do another series of tests starting from each of those positions, but now moving the speaker away from the front wall, out into the room, moving it just a couple of inches each time, same as before. You can keep on doing that as much as you want to find the "perfect" location, but it isn't necessary: You should find a really good location like that fairly quickly. Then do a final tweak of the cross-over point in the same way (adjust it just a couple of dB up, test with REW, a couple more up, test, etc, then repeat going down), then a final adjustment of the relative level, and you are done. It shouldn't take more than a few hours to do that, max.
Other people have other methods for locating subs, but I like this one because it is a repeatable, logical, consistent sequence, and you can usually see form the graphs very quickly if you are moving in the right direction or not.
2. When doing tests on individual mains, does the microphone remain pointing straight up?
Some people say to point it straight up, but I prefer to angle it at about 45°, so it is pointing more or less at the top of the front wall. I don't want any direct reflections from the ceiling doing nasty things at the tip of the mic: ceiling reflections will still show up on the IR graph, of course, but I don't want sharp artifacts from those messing up my readings. Yeah, the difference is subtle, but it is there.
Do ALL of your tests with the mic in the exact same location and orientation in the room. Do not change anything. And then measure the precise location of the mic in the room, so you can always get it back to the exact same point again in the future, for additional tests. You MUST always test with mic precisely in that location, accurate to within a quarter of an inch or so: if not, you won't be able to validly compare those REW graphs with the others. So make sure you measure the location of the tip of the mic very, very carefully, in all 3 dimensions, and note down those measurements some place safe. It is important that you get the mic back to that same point, every time. That should be exactly on the room center-line, exactly at ear height, exactly in the middle of your head, and angled 45° upwards.
3. You say I should increase the test level, where would I adjust that at? I believe the REW software has the level set to -12db...but the weird thing is, when it runs the test it says I only have about 5db of headroom...I must have something calibrated wrong? Could you elaborate a bit on how this is accomplished? I have a Universal Audio Apollo interface, which I calibrated using the info in the REW manual.
If you only have 5 dB of head room, then very likely the problem is that you did not calibrate REW to show the correct SPL level!
The procedure here is to run the test pink noise sound from REW, then adjust your speakers so that you are getting a level of 85 dBC on your hand-held sound level meter, then to use the "SPL meter" button inside of REW to tell REW what the real level is. If you don't do that, REW has no way at all of knowing what the actual SPL level is. All it gets is a signal from your interface at a certain level, but it does not know what the level means in the real world, unless you tell it. For example, if it is getting a level of "28,391", it has no idea if that refers to an SPL of 55 dB in the room, or an SPL level of 95, or whatever. The only way it can know, is if you tell it!
4. I have the Dayton EMM6, and I found out that I can get a calibration file for it...I assume that before any further testing I should do that and import it into REW?
Definitely! The differences are not major, but they are still very much useful. So get that file, and make sure you use it for all tests.
5. In reading here the last couple days, it seems that it's recommended to get an external SPL meter, which I don't have. I used the Dayton mic to calibrate the spl meter...but maybe that wasn't the right thing to do?
Definitely not the right thing to do, and very definitely you DO need a decent quality meter. Absolutely! You cannot calibrate your room, your system, nor REW, nor measure anything else accurately unless you have a meter. Get a good one: they go for around US$ 100 right now on e-Bay and Amazon. But do NOT get one of the cheap Chinese toys that are flooding the market, and go for under US$50. Get a respected name, such as Extech, Galaxy, or something similar.
You cannot use just a mic to calibrate the level in REW, because REW needs to know the actual sound level in the room, and as I mentioned above, there is no way that it can know that unless you measure it with something external to REW, and then tell REW what it was. If you try to use a mic and your ears, you will be waaaaaay off.
6. Sorry about the files, I didn't realize there was a "save all measurement in one file" option! I'll certainly do that from now on, with notes on each...
Yup! As soon as you do each new test, put a brief label in the box at the top right of the test, saying something like "L only" or "LR+Sub", then add your comments down below in the comment area, as you already did.
7. The closet. I was wondering if this was going to be an issue. I do use it for equipment storage, I have the superchunk in that corner and the one at the entry door with a pivot bracket that attaches to the floor, so I can just pivot them out of the way to get in/out and access the closet. I'm hoping not to lose it, and I can't put a door on it, but if it turns out to be a problem, I could put a hard surface (plywood/drywall) in the door opening to simulate there just being a wall there...? I could actually mount it to the back of the superchunk so it pivots out with it when needed. Just a thought, don't know if that would work or not acoustically...
Let's see how bad it is on the REW tests first, then figure out what to do about it.
8. What do I think needs fixing? Well I am concerned about that 345hz dip...it seems pretty massive
Yup, but it is fairly sharp, and there's no ringing on it, so I'm guessing that it is SBIR, and very probably associated with the back wall. To test that, do a series of tests in REW for several different mic locations. Test, move the mic back 4", test again, move another 4", etc. going back several feet. If it is SBIR, the frequency will change, if it is modal the amplitude will change: Send me that MDAT fill too, as it also reveals other stuff about the room.
I also did notice even before testing that there were certain bass frequencies that were seemed much louder, and looking at the graph I'm guessing it was right about 50hz. I'm not sure there's alot I can do about that though?
Yep, that's almost certainly a modal issues, since there is a long "tail" on it in the time-domain graphs. And it is associated with your 0,0,1 axis (front wall and back wall) so the treatment there is basically very thick bass trapping across the entire rear wall. But that will have to be done with care, in order to not make the room even more dead. Please post a couple of photos of the rear wall: That is the most critical wall in a control room. I suspect that whatever treatment you have back there is way insufficient.
The room sounds "different" to me. I've never had a treated room before, so as you say, it does sound somewhat "dead". I do notice a big difference in the way the low end sounds when moving away from my listening position. What I mean is, because of the depth of the room, it doesn't take much movement to get too close to room center and lose alot of low end there. If I keep moving backwards toward the back wall the bass picks back up again and about 2 ft. off the back wall it's much "fuller", more "enveloping"...sounds like the sub is in the room already back there! I realize this is probably due to not enough bass trapping on the back wall, but the room is only so big...may have to explore other options there...?
It might be possible to get some life back into that room by covering certain of your bass traps and other treatment in certain ways.
The reason you hear more bass as you move away from the mix position is because you are moving into modal response peaks and nulls, as well as other artifacts. It might sound "cool" to have more bass, but in reality it isn't cool at all! Turning up the bass is fine for when you just feel like listening to something for enjoyment, but for mixing, you want the flattest response possible in your room, with a slight bass rise and a slight treble roll-off. If you take a look at the specs for control rooms, such as those put out by the ITU, AES, EBU and others, you'll see that on their recommendations. But you don't want the type of overwhelming thudding huge deep bass that you hear from the cars of some idiots you see driving around with their pimped systems. Studio control rooms need to be neutral in both the frequency domain and time domain. They are not supposed to "sound good"! They are supposed to "sound flat". Because you need a flat, smooth, even "surface" to paint your acoustic "picture" on when you mix a song, just like an artists needs a flat, smooth, clean, white surface to paint his picture on. If his canvas dripping with mud, ketchup, and guacamole stains when he starts painting, and in addition he is wearing spectacles with one pink lens and one one lemon colored lens, and using lights that color half of his canvas blue and the other half green, then obviously the picture he paints will turn out like garbage. That's basically the same as what happens in a control room that "colors" the sound and does not produce smooth, flat response.
That's why the recommendation is to set up the room in a certain way, following certain geometric layout, in a certain shape/size room, and with certain treatment placed at certain places. All of that is designed to make the room sound as flat as possible...
at the mix position! And hopefully at other locations too, but what matters most is that you are hearing clean, smooth, uncolored sound coming direct from your speakers, without the room adding to that nor subtracting from it, while you are seated at the mix position, mixing. That's the goal. You need to be able to hear every detail of every sound, cleanly and clearly, in order to make your mix decisions. So you need that flat but slightly shaped response, because that has been shown to provide the best "white canvas" for you to paint on.
Some people who have never heard great control rooms are disappointment the first time the actually hear one: The are expecting to be blown away by earth-shaking bass and subsonics, and screamingly tinkling treble, but when they hear flat, smooth response instead, they don't get it: they don't understand that "boom and tinkle" is not the purpose of the control room. The purpose of the control room is to tell the truth to the engineer, so he can then make sure that you get all the "boom and tinkle" you want when you listen on your own system. But his system has to be flat in order to do that. If his system was exaggerating the bass so that it sounded pumpingly amazing in the control room, he would tend to pull down the bass in the mix to "flatten" it, so when you listened on your system, there would be no bass at all!
So it takes a bit of getting used to the first time you work in a "flat" control room. You need to spend quite a bit of time in there just playing around with mixes then listening to them elsewhere, to convince yourself that this is the right way to do it.
In short, if you mix a song so that it sounds good and well balanced in a control room that has flat response, then it will sound good and well balanced anywhere, even in places and on systems that have terribly "un-flat" response.
That's the goal, and your room is not too far from it right now! (Of course, perfection is impossible in a small room, but it can still be darned good!)
- Stuart -