Interpretation of measurements

[dominique-tanguy]

Good morning to all,

I am not a sound engineer and I need some help to interpret measurements. I obtain reliable results when I place my microphone close to the loudspeaker (blue curve), but when I place it about 1 meter from the cabinet, in order to measure a full range curve for the 3 speakers, I get a severe dip close to 50 Hz (green curve), which is exactly matching the frequency of the second bump of the impedance measurement.

Is there any relation, like phase issue ?

Thank you for your help

Dominique

[Panomaniac]

It is interesting that it comes right at the same place as the higher peak. I don't remember seeing that any any measurement I've done.
Maybe it's a phase cancellation with the port. What happens if you mask the port or move the mic a little bit?

BTW, is that your new impedance plot with the current port? I see that the second peak is higher than the first.

[dominique-tanguy]

I am doing some testing with an extended (deeper port). The higher peak to the right is not conventional but the FR measurement and the listening are a bit better. See the incredible low resonance value for a 515-8G... (normally 37 Hz free air).

Moving the microphone doesn't help and I had the same problem with previous port. I will try tomorrow masking the port.

We are invited by friends this evening...

[RonSSS]

Room response? Or are you measuring this outdoors?

Ron

[dominique-tanguy]

i Ron,

This is room response, 3 ft in front of the loudspeaker cabinet (VOT) for the green curve and just at the end of the front horn for the blue curve.

[GM]

How tall is your ceiling? It looks higher than average, so could be the floor/ceiling eigenmode that would require significant ceiling modification to decay it away. Most rooms [~ 8 ft high] have this notch around 80 Hz where you have a smaller one, so is the exposed beam ~8 ft up?

GM

[dominique-tanguy]

Good morning GM,

The ceiling is 9 ft high. I must admit that I will need more explanation to understand the issue... I have wooden beams across the ceiling every 2 feet, could this play a role?

Dominique

[dominique-tanguy]

Dear GM,

Further to your remark concerning the room, I suddenly remembered that I did measurements in the same room with another loudspeaker, two years ago, when I was just discovering ARTA and the measurements. I was lucky enough to find a copy of a still picture and guess what...

[GM]

Good morning GM,

The ceiling is 9 ft high. I must admit that I will need more explanation to understand the issue... I have wooden beams across the ceiling every 2 feet, could this play a role?

Dominique

Greets!

OK, not high enough to cause the floor-ceiling eigenmode [dip] all by itself, but all those beams may be causing a delay due to the eigenmodes [standing waves] between them, which if true would lower the dip's frequency. At 9ft, the dip would normally be around 63 Hz. Then again, if there's a ~11.5 ft span between two hard parallel surfaces in the room or a combination of different eigenmodes that can sum just right..............

Regardless, it's pretty obvious now that it's a strong room mode, so can't really do anything about it without a lot of effort and/or a lot of $$$$. Ditto all the peaks/dips higher up in BW until above ~2 kHz where the WLs have become small enough to be easily damped. Even if I hadn't seen pictures of it, your measurements tell me you have a very 'lively' [resonant] room indeed.

GM

[Steve Schell]

My nephew and I had a situation like this when setting up his system in his new house. It is one of those designs with most of the living room open to the second floor. High above the rear wall there was a large flat area angled at 45 degrees to wall and ceiling. When we fired up his speakers (Karlson enclosures with Stephens 103LX for bass) the low bass was missing altogether at the listening chair. We finally figured out that the sound was bouncing off that angled section and causing a near perfect half wavelength cancellation centered on 50Hz.

Dominique I'll bet you have a similar situation centered on about 48Hz. the wavelength of 48Hz. would be 1,130 divided by 50, or 22.6 feet. A half wavelength would be a little over 11 feet. It is a certain distance from the speaker to your listening chair or measurement position. The sound will radiate omni-directionally at those low frequencies, so if there is a surface that will bounce the sound back to your chair with an additional eleven foot path then you have found the problem.