|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
124.168.141.104
I've got some large midrange horns here that I'm measuring. I'm finding that an on-axis FR measurement (it's gated to eliminate room effects) near the mouth is significantly different from measuring on-axis 2 feet away. Further away, the LF response rolls off earlier (higher in frequency) and the HF response rolls off more slowly. Is this behaviour to be expected?This is a 200Hz horn loaded with a compression driver. Measuring at the mouth, the LF is approx -3db at 300Hz. 2 feet away, -3db is 400Hz, 6 feet away, -3db is approaching 500Hz. I can certainly hear the hole in the response there from my listening position if I cross over too low.
The HF, on the other hand seems to roll off less steeply when measured from further away.
I'm just wondering if this is standard horn behaviour? I expected the measurment to be different, but I didn't expect to see those general effects.
Follow Ups:
Always like to hear what Bill F. & John S. say, they know of what they speak.
In this case, I'm more on Bills side, as I do give a F about what happens @ 2m even when listening @ 1 meter because any relatively small room I have ever been in (other than EV's anechoic chamber) is going to present enough reflections to skew the response.
And WTF would you listen to large midrange horns @ 1 meter? I shudder to think...
Hi Jeffmai .
I just set up the Altec A5 and I have been told you know alot about Horns stuff . could you mind to come over for a listen and tell me what you think .
email me off line if you could . I'm a friend of Danny . I'm not far from Danny place.
Cheers
Peter
I have measured several horns and this is normal behaviour. At least for the horns I have measured, mainly tractrix and exponential horns. The reason is, I believe, is that at the low frequencies the radiation is from the mouth of the horn. Then you are measuring in the nearfield of the aperture where you have different components of the acoustic field contributing to the sound pressure not only the ones surviving to the farfield. At higher frequencies the radiation takes part inside the horn where the radius of the radiating aperture is smaller and thus the nearfield is closer to the horn.
You have to measure from at least 2 meters, and do it outside, gating doesn't cut it. Less LF extension and higher HF extension than predicted is to be expected.
If you don't gate the measurement, how do you eliminate the ground bounce even outdoors without a lot of trouble?
I employ a highly absorbant non-reflective medium on the ground plane, ie., a four inch lawn. It's a good excuse not to mow too often.That is a good question, though. Take the ground bounce null frequency out of the loop. With woofers take the measurement with the mic a couple of inches above the ground, the path differential between the direct and reflected waves will be too short to show up in the passband. With mids and tweeters lift the box and mic high enough off the ground to render the reflected wave moot. With a full ranger measure both ways, then combine the low frequency response with the former method with the high frequency response of the latter.
Option two is to dig a hole in the ground, put the speaker in facing up, backfill to get a true 1/2 space mounting and suspend the mic via a boom. A bit more trouble than it's worth.
Do you remember the Sherwood ads, back when they made loudspeakers, that showed them in baffles in the ground, just for that purpose?
> You have to measure from at least 2 metersNo you don't. What if I listen at 1 meter (or any other arbitrary distance)? WTF do I care about what the device does at 2m (unless there happens to be a reflective surface there and I'm interested in what's being bounced around the room)? You should measure in a way that gives you good data relevant to your situation.
> gating doesn't cut it
Gating works just fine. If you measure in a tiny closet of a room and want to go down to 20hz, then no, gating won't 'cut it' because you are applying the concept incorrectly. But as long as you apply the theory correctly (limit your gate time based on where your first reflection occurs), it works just fine.
No need to have poor Ivan schlepping cabs and all that gear into the parking lot so he can measure at 10 meters anechoic. Unless he needs the exercise anyway.
Are you being sarcastic? I'll assume so.It seems you're confused on several different things. Real companies designing large prosound equipment have very different measurement requirements than most people around here trying to design their own hi-fi systems. A prosound company doesn't know where their design will be listened to from, but can usually assume it's further away than 1m. I (and most others here) know exactly where I will listen to my speakers from, so I can measure at that location. That may be 5m. It may be 1m. I know someone who set up an awesome nearfield system where you listened from about 10 inches. So saying "You have to measure from at least 2 meters" is not good advice.
And if you read what I wrote about gating, you'll notice I said you won't be able to use it to measure subwoofers at low frequencies in a small room. But that wasn't what Jeff was asking about - he was asking about midrange horns, which could be measured indoors using a gated measurement depending on the mic / speaker / room combination.
Hi BillWell it depends.
The 1W 1M spec is sometimes used to estimate how loud the speaker will be at another distance and power. For that use, it is desirable to have an accurate measure.The inverse square assumption for a point source assumes the source is small and occupies none of the volume in the space. Digging a big hole to drop the cabinets into is not practical but having enclosures which occupy a significant fraction of that 1 meter radius with an inflexible obstruction or have a large source area, is also not desirable so far as accuracy.
In addition, close to an lf source, one can measure “psudo sound” error which is a microphone signal caused by kinetic reaction of the moving air hitting the element and NOT radiated acoustic pressure.
All of those concerns are essentially eliminated when measuring at a larger distance, say 10 meters where 100Watts drive equals the 1W 1M level.Also, in answer to the original post, yes one normally does see some distance related changes in response. Up close, one has more low end than farther away.
For example, an SH-50 measures a gradual rise in low end response at one meter where as by the time one measures past about 6 meters, the low end is slightly depressed.
In general, the more a sources DI changes with frequency, the greater the response changes with distance.
Best,
This post is made possible by the generous support of people like you and our sponsors: