Home Theater Forum and Systems banner
1 - 16 of 16 Posts

·
Registered
Joined
·
9 Posts
Discussion Starter · #1 ·
Hello shackers,

I have recently acquired a BlueSky 2.1 monitoring system, naively thinking that this would magically make me prdouce better mixes. I had, of course, completely neglected the importance of room acoustics.
I realized after having done my first mix that listening to it with headphones resulted in a horrible, completely distorted low spectrum (and other things).

So I borrowed a mic (EMX-7150) and did some measurements with REW.

You can find 2 graphs here :


Full range 1/3 octave smoothing.


Sub+speakers (green is without sub, to show the cutoff freq)
It's not exactly what you could call a flat response, I guess. Going from 74dB at 54Hz down to 47dB at 107Hz seems pretty massive to me, but I'm new to this stuff so please correct me if I'm wrong.

The room is a small student's room, 2.4m x (3.3m + 2.1m) x 2.4m (w x L x h).
I'm saying 3.3 + 2.1 m for the length because it is shaped this way :


Unfortunately the desk is stuck to the wall so I can't really change my setup.

So what do you think ? Is this a desperate case ? Can I eq, or is it going to be even worse ?

Oh and I forgot to talk about the nice, long flutter echo that I get because of the empty flat parallel walls ;)
 

·
Banned
Joined
·
516 Posts
The first thing you need to do is establish left-right symmetry of the speakers.

The resulting imbalance due to the lack of symmetry caused by the room interaction cannot be EQ'd.

If you are considering EQ, ONLY consider it for the sub.

Also, you might want to also reconsider your headphones if the result is as you say.
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #5 ·
Don't you think that the hole at 100Hz is weird ? It looks like the cutoff frequency of the sub is too low, but I can't change it (and I checked the polarity on the speakers, it's not that).
About the symmetry of the speakers, I thought it wasn't great to have everything perfectly symmetrical... anyway is it going to have an effect on the frequency response ? Because the stereo image is not that bad (or at least that's what I think :D).

Wouldn't it be good to give at least a little boost in the 1k-2k range and a high shelf after 5k ?
 

·
Premium Member
Joined
·
6,711 Posts
Strange to see the speakers response roll off so early, especially given that from your diagram they are close to the wall. From the Blue Sky info the high pass to the mains is at 80Hz, so shouldn't be the cause. Measure each speaker separately to see what the speaker responses alone look like, could also try connecting directly to the speakers to confirm what effect the sub high pass has on those speaker outputs. It would also be worth trying the sub midway between the speakers rather than off to one side (if that leaves any room for your feet).

Beware of using high frequency measurement roll-offs as an excuse for boosting the high end, there are many factors that contribute to a roll off when measurements are made at the listening position. Almost certainly best leaving that alone. Also worth measuring a few positions within a foot or so of your head position to see how uniform (or not) the response is.
 

·
Registered
Joined
·
9 Posts
Discussion Starter · #7 ·
Ok, I did a few more measurements.

Here they come :

Moving the mic up and down with L+R speakers (+30cm is the listening position).
The first graph here looks better than yesterday... I may have moved the speakers a bit, but not more than 10cm or so.


Right speaker only, +30cm and 0cm


Left speaker only, once with the speakers flat on the desk and then pointing up towards the mic


And finally sub center/off axis


Any thoughts ?
 

·
Premium Member
Joined
·
1,400 Posts

·
Registered
Joined
·
9 Posts
Discussion Starter · #9 ·
Thanks for the links !
I haven't read everything yet, but it made me want to try and change the distance between the speakers and the front wall.
Here are the results :


You can see that the impact on the high frequencies (5k+) is tremendous, and that the notch at 100Hz is moving too.
I read here and there that placing speakers against the front wall was a good way of diminishing the impact of SBIR, but it seems to produce the worst results for me.

I actually managed to get a graph that doesn't look nearly as bad as the first ones by placing the speakers at the very end of the desk, and setting them farther from each other.
It's not the most practical setup, but it does sound better...
 

·
Registered
Joined
·
8 Posts
I am a noob to this forum and to REW, and I am looking at posts by others to see what I questions I should ask once I run my own REW measurements.. now this last frequency response graph by cocoflunchy, does it look like the freq response that I should be shooting for? because that looks like a very flat response to me...
 

·
Super Moderator
Joined
·
9,309 Posts

Hey longhorn,

First, welcome to the Forum!

The reason the “75 cm” graph “looks” so good is because it has a huge 100 dB vertical scaling (20-120 dB), where the “Influence” graph had a tighter 60 dB scaling (40-100 dB). Large scaling has the effect of “flattening” out response, at least visually, and making it “look” better than it really is. But look closer at the 50-90 Hz range in both graphs. Note that the blue trace in the “Influence” graph fully spreads the 10 dB span between 60 and 70 dB – about the same as in the “75 cm” graph.

For general purposes, we typically recommend using a 60-dB vertical scaling – see this post:
Getting Graphs Ready to Post


Regards,
Wayne
 

·
Registered
Joined
·
8 Posts
Hi Wayne,
Oh ok I see it now. Thanks.

I'll post my graphs when I get my EMM-6 and Art USB Dual Pre. I just realized by reading this forum that my UCA202 and RS meter won't cut it for full spectrum measurements...
 

·
Registered
Joined
·
29 Posts
Hello shackers,

So I borrowed a mic (EMX-7150) and did some measurements with REW.
Hi cocoflunchy,

This is a bit OT but I am curious what mic preamp you used for this measurement? How was the gain?

I reviewed the overall specs for that mic are very good. The sensitivity spec. of the EMX-7150 is only 6mV/Pa . I would like to know how much gain was needed to adequately run REW with that mic.

Thanks
 

·
Registered
Joined
·
110 Posts
Don't you think that the hole at 100Hz is weird ? It looks like the cutoff frequency of the sub is too low, but I can't change it (and I checked the polarity on the speakers, it's not that).
About the symmetry of the speakers, I thought it wasn't great to have everything perfectly symmetrical... anyway is it going to have an effect on the frequency response ? Because the stereo image is not that bad (or at least that's what I think :D).
The "hole" at 100Hz may be because there is a room mode (standing wave) at that frequency, and your measuring position is at or close to one of that standing wave's nodes (cancellation point, SPL or volume low-points). This is why EQ cannot correct room issues, the amplitude response at that frequency is very likely to be very different in other parts of the room.

As an experiment, and to open your eyes (ears) a little, set REW's Generator to sine and the frequency to track the graph cursor. Watch your levels and press "play"; drag the cursor over that hole (if your head is where the mic was, you'll hear it diminish). Now move around your room and listen to the amplitude of that sine wave -I bet that it will not be at the same level in all parts of the room (if it is quiet everywhere, you do have a monitoring or crossover issue!).

Move the cursor to one of the peaks and try the same thing -if its a primary axial mode, you'll find it's quiet halfway between the walls creating it and loud at the walls, if it's a seconday axial mode, it will be loud at the middle and the walls, but quiet at the two places in-between. Non-axial modes get complicated, and other things in your room will start to diverge the calculations and the reality at this point (usu ceilings and floors tend not to reflect the really low frequencies, or plasterboard walls start geting "transparent" ;) ).

Playing sines at the peak and trough frequencies can tell you things about your room. Putting broadband bass absorption in the loud areas can increase the level at a quiet zone in your room at that "trough" frequency. It can do your noodle in, but it's very interesting once you get a handle on it, and once you add acoustic treatment the difference is very marked...


>
 

·
Premium Member
Joined
·
2,116 Posts
One thing that helped me make sense of my graphs was to use a room mode calculator (there are multiple on the web).

If the graph is decent, sure enough you'll find that some of the peaks/dips correspond pretty well to the room modes.

It helps to understand what's the room, and what's not. (at least in my experience)
 

·
Registered
Joined
·
110 Posts
I've asked JohnM if he could add the room modes as overlays to response curves, simulated curves and Spectrograms. If you create a simulated curve, you enter the room dimensions anyway, and it would be great to see the mode lines overlaid to scal on the REW traces.

John was reasonably receptive to the idea so I have my fingers crossed :)

The only concern is the validity of rectangular room theory in real rooms, and the danger of leading some users to apply too much weight to the theoretical. If you've worked with mode calculations before, you'll know that they don't always predict what you see in reality, and for those that do tally, usually only the strongest seem to be relevant ;)

Would be a good tool and extra to the program though, I hope it happens.


>
 
1 - 16 of 16 Posts
Top