Help interpreting waterfall graph of my room acoustics?

[enroper]

I'm hoping somebody can spell out what I'm seeing in this graph.

I used Room EQ wizard, with an SPL meter sitting on a tripod in the 'sweet spot' of my mix position. I calibrated everything as per the directions, and am using KRK RK-5 monitors.

My room is about 12x18, not exactly sure. I've attached both the waterfall and a rough layout of my room.

No treatment is in place yet. This weekend I plan on installing DIY basstraps with OC 703 equivalent in all 4 corners, and 2 panels of 2" directly behind my seat at the mix position.

So two questions

a. what is this telling me
b. when i treat my room, what should I be looking for on the new graph.

I know this is a loaded question, any help would be appreciated

[jwl]

Hi enroper,

A waterfall graph is basically a 3D graph of loudness (up/down, louder is higher) of a range of frequencies (left/right, left is lower) over time (back/front, front is longer). You can use it to see where the ringing in your room is occurring. You can see some pretty nasty ringing between 40-50Hz, another narrow ring at around 61Hz, 75-80Hz, and about 140-150Hz. These are likely related to the room modes of your room; for example, 45Hz has a wavelength of about 22', so the octave of that (90Hz) is close to 11', close to one of your room dimensions. The 61Hz peak you have relates to 18.5', which is about the length of the room. Etc. For more info on this, you can see Optimizing Acoustic Treatment using the ETF software, the article is geared toward different software than what you have, but the principles are the same.

You have the right idea, getting as many bass traps into as many room corners as you can will help clean this up. When you repeat your test you should see a less jagged waterfall graph, with ringing that doesn't take up as much "floor space" on the graph (ie, the ringing stops sooner in time).

In addition to the bass traps, if at all possible I'd reorient your room, this will clean up your graph significantly also. Ideally, you should swap the recording/mixing area with the bed. For best sound, you want a symmetrical setup in the room (particularly from the listening position forward) with the listening position 38% back from the front wall in the center of the room between the side walls. The speakers should make an equilateral triangle with your head, ideally with the tweeters at the same height as your ears. This setup will give you optimum stereo imaging, and the placement within the room minimizes the effect of room modes on the bass. For more detail, read How To Set Up A Room.

Good luck! Post back with more tests after you rearrange and/or install bass traps. Both will help.

[jwl]

Just to add more thoughts to this, your room is a good example of what happens to the low end in a small room without bass trapping. Look at your graph between about 90Hz and about 170Hz. There are lots of peaks and valleys, this represents a VERY uneven response in the room at your mix position, a difference of about 25-30dB over a VERY narrow range.

For instance, there is a large, narrow peak at around 150Hz, and another at around 160Hz (it's hard to tell exact frequencies, since the graph is not labelled between 100 and 200Hz). But between these 2, there is a severe null, a difference of about 25dB. This means, as you listen, you will hear a huge hump at around 150Hz, then 155Hz will seem to disappear, then another huge hump at 160Hz.

This is why acoustic treatment is so important. What good is audio gear (speakers, preamps, etc) with ruler-flat response when your room skews what you hear so much?

My guess is that it's difficult to hear bass instruments in your room, to get things sounding right. My guess is that different bass lines sound "one-note-ish" where no matter what note the bass is actually playing, it sounds like they're hitting the same note over and over (to a degree). You probably get mixes just right in your room, then when you play it back on other systems it doesn't sound quite right.

Fixing the room response, with bass traps and reorienting, will help you with this.

[enroper]

Thanks for your detail analysis! I've picked up some OC 703 compatible material today ( johns mansville 800 series, same acoustic properties as OC 703 ). I'm going to be installing at the minimum 4 4" bass traps, and 2 2" panels behind my head at mixing position. Hopefully that helps.

Its supposed to be a dreary weekend so I'm hoping to have some results by monday. Keep an eye on this topic if you care to see the results :)

[peeder]

Don't ignore the rearranging part of the advice.

[enroper]

Well, I got all my materials tonight, and already built one 4" bass trap. I hung it up to check the fit and finish. I can tell this is going to be interesting. When I'm up close to the trap I can notice a dramatic difference singing into it, vs. singing just a few inches over towards the bare wall. I'm curious to see how all 6 of them are going to change my dynamics. This time tommorrow night we should know..

[Glenn Kuras]

Originally Posted by enroper

Well, I got all my materials tonight, and already built one 4" bass trap. I hung it up to check the fit and finish. I can tell this is going to be interesting. When I'm up close to the trap I can notice a dramatic difference singing into it, vs. singing just a few inches over towards the bare wall. I'm curious to see how all 6 of them are going to change my dynamics. This time tommorrow night we should know..

Very cool to see you moving forward. Here is a setup that might help you set up things in the room a little better. This is not your set up to scale, but should give you a few ideas.

http://i227.photobucket.com/albums/d...s/102602-1.png

Glenn

[enroper]

Well, Installed 4 corner traps of 4" 4'x2', and 4 wall mounted panels of 2" 4'x2'.

Below are the new graphs. The first one is the before, the next is the after.

Now, I still dont know what the hell I'm looking at, so , is this a good improvement?

[jwl]

Hi Enroper,

Yeah, the graphs look better. Well done!

The graphs are a bit too zoomed-in to get a detailed reading (mostly I can't read the time axis (front/back)), but you can see that you have a much flatter response.

For instance, in the before graph you have serious "valleys" (or "dips" or "nulls") at around 105Hz, 125Hz, and 145Hz. In the new graph, these nulls are practically gone!

No room is perfectly flat, but yours looks TONS better now than it did.... the overall response looks much less jagged, there are fewer high, narrow peaks and deep valleys in your response curve.

Did you re-arrange the room at all, or is everything still where you left it?

Most importantly, how does it sound? You should hear a much flatter bass response, put on a good recording that you know well with a nice bass line. You should be able to hear much more detail in the bass.

Depending on where you placed the wall panels (did you make a RFZ?) you should have much clearer stereo imaging as well.

Looks like you have a very good start!

[jwl]

A couple more thoughts about reading the graphs, they can be tricky to wrap your head around.

Look at the very top of the graph where the green waterfall starts, right where the green curve "touches" the graph paper-ish calibration lines toward the "rear" of the 3d graph. if you look at this, that's the initial frequency response of the room, the moment that sound begins in the room. And you can see a swing in frequency response.... 105Hz is at around 56dB or so, and 140Hz is up at around 88dB or so. So that's a 32dB difference in volume between 2 freqencies less than a half-octave apart!

Then, you can see more curve lines, similar in shape to the first one, on the green graph in front of that. You have to think 3D to get this one. each of those curves represents a step forward in time about 10ms or so (kinda hard to read, the time calibration is on the right of the graph).

So again, what you are looking at is a frequency response graph of your room, plotted over time, with (approx) 10ms intervals. So you can see how sound decays in your room, where the resonances are, where the ringing is, and what the overall response in your room is.

The after pic shows much less difference between the loudest frequency and the softest frequency, you can see the curve is much smoother.

I hope I'm helping here, and not just being verbose..... ;-)

Tell us how your room sounds now....

[enroper]

Thanks for your analysis!

From a mixing/listening standpoint, I would say things sound "flatter". The bass isn't either too in my face or absent. I suppose thats the point? I guess I'm hearing what the monitors are really supposed to sound like.

I can tell a dramatic difference just walking into the room though, it sounds more like a "studio". A lot of the flutter echos are cut down. It passes the "clap test". I still have a couple problem areas, notably two closets with metal folding doors, and windows with blinds and no covering, but overall it sounds way better.

[jwl]

excellent, enroper. Glad to hear it.

Adding 4 bass traps and 4 panels to a 12x18 room should definitely make an audible difference, it's a great start.

I'd say the next step would be to add more bass traps into more corners; a room that big can definitely fit more, especially once you start treating the wall/ceiling corners. When you have the budget, I'd recommend repeating what you just did, and add the additional traps to the room. It will improve things even more. If you were able to rearrange the room (if you haven't already), that would help, too. Did you use the wall panels to create a RFZ? Also, as they say in acoustics/studio building forums.... PICS OR IT DIDN'T HAPPEN!!!!

I know just what you mean when you say the room sounds more like a studio. That was a very vivid moment for me, the first time I set up mics in a treated room. "Oh, THAT'S how the pros do it....."