So here's the polar graph of my current design starting at 11.25 degrees off center, in 11.25 degree steps out to 90 degrees:
Modeling the low end gets bass to 100Hz.
Now here's what the room does to that response:
The mic is placed at 5 spots across the couch. So this is the average room response at the listening position(s). The bass boost on the receiver is turned up a bit or else that 200-300 Hz floor bounce suck out makes it sound too light. Both ends of the couch have a pretty smooth response, and the middle has the most bumps. Anyone care to comment on what going on with all the variation?
Thanks,
Dan
Modeling the low end gets bass to 100Hz.
Now here's what the room does to that response:
An externally hosted image should be here but it was not working when we last tested it.
The mic is placed at 5 spots across the couch. So this is the average room response at the listening position(s). The bass boost on the receiver is turned up a bit or else that 200-300 Hz floor bounce suck out makes it sound too light. Both ends of the couch have a pretty smooth response, and the middle has the most bumps. Anyone care to comment on what going on with all the variation?
Thanks,
Dan
Sure Dan. Since this is near and dear to my hear at the moment, I'll jump in. May get a ton of flack for it, but here goes:
Measuring in-room response with an omni-directional mic is just wrong.
I can see the omi's usefulness in free field or anechoic conditions, but not in room.
Why?
Because your ear is not omni-directional. It's closer to a supercardioid pattern, as best I can determine. I've spent a lot of time searching for polar response patterns of the human ear, and they are darn hard to find. (Need to go to the library). But the few I have seen look like a supercariod mic pointed about 55º off axis.
If this is the case, then an omi mic is going to be picking up a lot of room reflections that your ear will not. Or at least picking them up a lot louder from the rear and front than your ears do. We do have 2 ears (most of us) but we can at least get started right with the proper mic pattern. One that is closer to what the ear picks up. And not a mic that picks up an equal amount of reflected energy from all around the room. Ears don't work like that.
So very recently I've taken to measuring and EQing with a cardioid mic pointed about 52º off axis.. I have a cardioid and and a hyper-cardioid, but no super, so I'm making do. Still, for me the results have been much better.
Using an omni at the listening position and getting the FR flat or even with a bit of top end roll-off just never sounded right to me. Using the cardiod pointed off-axis had been much, much closer to how I would tune the system by ear.
This is a new approach for me, so still fine tuning it. But I think it has some bearing on your "at the couch" response curves. If you're using an omni mic, it's picking up a lot of stuff that your ear does not.
Luckily, most of us have two ears......
When I listen to a mono source and turn my hear around, the frequency balance dont change all that much - our two ears together are virtually omni.
Thanks for that post, Dr. Geddes. I had started thinking that you see your CD-waveguides as a thing of their own or just as an intellectual challenge. Time to withdraw my threadjacking accusation.
Do the axes of the speakers hit the side walls or the rear wall? And what is the difference with/without the ceiling diffusor sonically? Have you tested the Messiah from the BIS?
I don't believe it. More like 2 supercardioid mics in an X/Y configuration.
A dummy head microphone might be interesting to work with, but how do you take into account the 2 mics? Count it as an omni? I don't think it is.
That's not because our ears are omni, it's because of the processing power of the thing between our ears
See how much the HRTFs for different directions vary.
Hi Dan
This variation is standard and unavoidable. Your data has one of the clearest demarcation between the modal and statistical regions that I have ever seen. Just below 200 Hz it becomes clearly modal and above that clearly statistical. The peaks and dips above 200 Hz are not significant until you average out the five curves. Then this average has significance. the theory behind all of this is very complex and not very well documented.
I would not assume the hole is a floor bounce. There are lots of other things that could cause this. There is usually a jump in reponse when the room transitions from modal to statistical and this happens at that exact same point.
As to the mics being omni, I completely disagree with this. The correlation between what we measure and what we hear is a lot more complicated than simply the pattern of the microphone. As Markus says, its all in our heads!
I never said it is as simple as what type of mic to use. Please re-read my post. I did say that I think the omni mic is wrong. The problem is hugely complex, as we all know. 2 ears, a head, a brain. They are all working to make what we hear not sound like what a mic hears. Our eyes don't work exactly like a camera, either.
But I think there should be some steps in a better direction. At least with my fat head I do NOT hear all frequencies equally from all directions. Yes, I have tested this. My 2 ears, head and brain are not an omni-directional, point source microphone.
If we want to consider the room and loudspeakers as a system, we need to admit that the directionality of our hearing is going to play a large role. In a typical listening room a high percentage of what hits our ears is reflected sound. If we don't hear all frequencies equally from all directions, that can make a big difference to the tonal balance.
I have just started doing listening position measurements and EQ with a cardioid mic, so this is all fresh. But it has already gotten me to a much better tonal balance. And a lot of other things have cleared up as a result.
There is a lot of talk about a speaker's power response - how it spreads the energy around the room. It's important because we listen in reflective environments. The way in which our measurement mic picks up the energy at the listening position should be just as important.
But I think there should be some steps in a better direction. At least with my fat head I do NOT hear all frequencies equally from all directions. Yes, I have tested this. My 2 ears, head and brain are not an omni-directional, point source microphone.
If we want to consider the room and loudspeakers as a system, we need to admit that the directionality of our hearing is going to play a large role. In a typical listening room a high percentage of what hits our ears is reflected sound. If we don't hear all frequencies equally from all directions, that can make a big difference to the tonal balance.
I have just started doing listening position measurements and EQ with a cardioid mic, so this is all fresh. But it has already gotten me to a much better tonal balance. And a lot of other things have cleared up as a result.
There is a lot of talk about a speaker's power response - how it spreads the energy around the room. It's important because we listen in reflective environments. The way in which our measurement mic picks up the energy at the listening position should be just as important.
Cardioid mics do not have a linear frequency response in the far field at low frequencies because of the proximity effect. If you compensate for the mic's bass roll-off, you will most certainly get at different tonal balance - one that I personally do prefer.
My own preferred target, as measured with an omni mic, has a 10 dB rise in response between 150 and 20 Hz. That would probably measure quite flat with a typical cardioid mic.
That is very true for vocal mics. Almost all have that low end roll-off which, as you mention, would lead to a bass heavy EQ. Most supercardioids are vocal mics (and vice-versa)
But there are a few supercardioid mics out there that don't do that. The Schoeps mk41 is a shining example. Very good pattern and response. Like all Schoeps, it isn't cheap, so I'm not running out to buy one. I have used them for recordings tho. The mk41 is known in the biz as a very neutral sounding mic. And here is the interesting part. It's well liked as a mic that "sounds like what you hear" in far field recordings. Hmmm....
My biggest problem right now is that I don't know the frequency response of my cardioid mic. Bought it too long ago. I do remember it having a droop in the low end, but not as bad as a vocal mic. So I have to take it with a grain of salt. I hear room problems (or maybe speaker) circa 180Hz that the mic does not. Still a work in progress.
But there are a few supercardioid mics out there that don't do that. The Schoeps mk41 is a shining example. Very good pattern and response. Like all Schoeps, it isn't cheap, so I'm not running out to buy one.
I have used them for recordings tho. The mk41 is known in the biz as a very neutral sounding mic. And here is the interesting part. It's well liked as a mic that "sounds like what you hear" in far field recordings. Hmmm....My biggest problem right now is that I don't know the frequency response of my cardioid mic. Bought it too long ago. I do remember it having a droop in the low end, but not as bad as a vocal mic. So I have to take it with a grain of salt. I hear room problems (or maybe speaker) circa 180Hz that the mic does not. Still a work in progress.
This make a lot of sense.
Here's a large part of my problem:
An externally hosted image should be here but it was not working when we last tested it.
This is the woofers response. One box is fiberglass filled, the other is not. The box is well oversized and the bass output is just too low. You can also see that the cone break up isn't anything like Eminence's graph. That's giving me fits in the crossover. So we're back to the rim resonance problem as far as that goes. Any one know of a manufacturer's graph you can trust? I gotta put a notch filter in that woofer.
Anyway, I think this clear demarkation of modal behavior could be a good thing for me. Looks like adding in a nice eq and I should be able to dial flat at the listening positions easily, but we'll see how that goes.
Dan
Hi Dan
I'd have to say No, I don't know of any speaker manufacturers who are going to give you rock solid data. B&C is about the best that I have come across, but even they have their variations. And remember that one driver isn't going to be the same as another driver - production variances can be quite large. Lot to lot variations even more. But basically I would agree that the driver you posted does not seem to have very much usable bandwidth (at least not the top - unstuffed I presume - graph). The other one is pretty typical. How a driver handles the first few breakups is IMO the difference between a good one and a bad one.
Last edited: