|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
70.210.138.26
Cheapy pyle on Dayton H45e x@ 3k to Eminence beta8A in .4 cu'
Follow Ups:
How about this one?
Well that's a real loudspeaker!
You have two measurements of the speaker, the top is magnitude (and possibly phase), the other is the impulse response which is the "time view" of the resistive portion of (energy time curve or envelope time curve) of magnitude and phase. The impulse and magnitude and phase are two different views of the exact same things, with one, you can calculate the other.
Observations;
The flat looking response below about 50Hz is probably not real, the response would normally continue to fall off below the lf corner at about -12dB / oct for a sealed and -24dB/oct for a vented box.
The appearance of "free bass" in impulse based measurement systems can be from not using a long enough impulse, not averaging several impulses and or reaching the system noise floor (for example see if that changes when you increase the mic gain). The lf corner appears to be around 70Hz (this is a smallish speaker?) and the in band response appears to wiggle about + - 6 or 7dB or maybe a bit with whatever smoothing was applied. For marketing loudspeakers + - 3dB is what people are conditioned to expect and this would probably "fit" if it were smoothed "enough". Recall that while we can hear a 3dB change in level, generally it is a +10dB change (10 times more power) that is subjectively "twice as loud".
Without phase being visible and time of flight removed, I can't tell where crossovers are or if it's a single full range driver.
> The flat looking response below about 50Hz is probably not real, the response would normally continue to fall off below the lf corner at about -12dB / oct for a sealed and -24dB/oct for a vented box.
I'm more inclined to think that's the result of cabin gain kicking in around 50Hz, off-setting the acoustical roll off of the speaker. OP, you'll get a much more useful chart if you change the scale, making the vertical lines 3 to 5 dB apart rather than 10dB.
Hi Bill
It's possible he measured in a bunker or basement but i have never seen a "normal house" measure more slope than about 9dB/oct and one was as low as 4dB/oct.
What i see often enough though is people using sequence based measurement systems that show extra an lf plateau like is shown.
When one see's the lf response curve rolling off and then flattening out, it is most often been the noise floor, not a long enough gate time or changes when say 5 measurements are averaged. Each time you double the number of measurements being averaged, you reduce the noise uncertainty by 2. Remember these sequence based systems do not have good noise immunity compared to TDS.
I've seen that result more than a few times, where the acoustic roll off kicks in quite a bit higher than where cabin gain starts. I've also seen guys measure low frequency peaks that shouldn't be there, which usually ends up being ambient noise ranging from heating/cooling equipment to road traffic, but those tend to be narrow bandwidth.
I've measured 20-25Hz content at major concerts that was at the same level whether the band was playing or not. It was audience foot traffic, road traffic, refrigeration equipment and generators. Taking a control measurement when the venue was empty it wasn't there.
It's a sealed .4cu' cab so anything below about 80 is the room I would suppose. I am still wrapping my head around the 1st reflection gating thing. Quite crude setup I have on the living room floor. Jh
It's kinda hard to know where to put this - here or under Tom's post."so anything below about 80 is the room I would suppose. I am still wrapping my head around the 1st reflection gating thing."
Making loudspeaker measurements in "regular" rooms is fraught with difficulties, mostly due to room reflections and resonances. But with a gated system, you're creating a window of time, much like the shutter of a camera - it's open for a certain period and then closes, so your measurement system is "listening" to the sound for only the moment you want. Back in 1974, at the acoustics lab of the Institute for Advanced Musical Studies in Montreux, Switzerland, John Curl, John Meyer and Bob Minor developed a system to do exactly that. The computer program also allowed them to average a number of tests, to reduce the influence of ambient noise. This was a big deal in those days. Now, 40 years later, gated systems are readily available.
Now, remember that sound travels at about 1129 feet per second at typical room temperature and humidity. So that's about 1.1 feet per millisecond. So, if your first reflection from a wall, the floor, an object, etc., arrives 5 milliseconds after the direct sound, you've got to close your window before that happens. But wait, it gets even better... You don't want to start the measurement time window until "just" before the desired sound reaches the microphone. So, if your microphone is 10 feet from the speaker, you'll want the measurement to start roughly 9 milliseconds after the sound left the speaker, AND, you'll want to close the window before the first reflection reaches the mic (in this example, 5 milliseconds later).
While all this sounds very cool, there is a limitation: The technique doesn't work below a certain frequency range, due to wavelengths, room size, and first reflection delay time. I've long since forgotten what the "rule of thumb" is for determining where that frequency range is, but IIRC, for "average" sized home living rooms, I think it's in the neighborhood of a couple hundred Hertz. Maybe Tomservo or someone else will refresh my memory on the details.
Hope this helps.
:)
Edits: 02/12/16
Thank you for the explanation. "Fraught with difficulties" like my 4 and 10year old stomping down the hallway :-). JH
Being sealed the acoustic roll off is 12dB/octave, the same as cabin gain, so that meshes with why response flattens below 50Hz, the likely cabin gain knee frequency.
Bad connection, non-functional come to mind as that doesn't look like a loudspeaker measurement
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: