Hello,
I started with using REW and UMIK-1 microphone to understand my room behaviour. On the way, I measured my LXminis. The Frequency Response is correct, but I'm puzzled with the Impulse Response measurement that does not looks at all like a pure dirac. See below:
- starts with a negative 100% (instead of expected positive),
- then a positive 80% (OK for a small "overshoot" but not that much).
Is this normal and related to the 2 drivers and the inverse polarity of the Full Range (by design inverse polarity to be in phase with the woofer at cross-over freq) ?
Is this abnormal and I should investigate ? In REW documentation, the picture for IR measurements shows a nice positive single peak, as expected, with only few ripples.
Nota, the Freq Response at crossover frequency is clean with no dip, which confirms that the woofer and the Full Range blend well and are in phase at that point.
This may be a quite stupid question, sorry for that. But willing to learn !
JMF
I started with using REW and UMIK-1 microphone to understand my room behaviour. On the way, I measured my LXminis. The Frequency Response is correct, but I'm puzzled with the Impulse Response measurement that does not looks at all like a pure dirac. See below:
- starts with a negative 100% (instead of expected positive),
- then a positive 80% (OK for a small "overshoot" but not that much).
Is this normal and related to the 2 drivers and the inverse polarity of the Full Range (by design inverse polarity to be in phase with the woofer at cross-over freq) ?
Is this abnormal and I should investigate ? In REW documentation, the picture for IR measurements shows a nice positive single peak, as expected, with only few ripples.
Nota, the Freq Response at crossover frequency is clean with no dip, which confirms that the woofer and the Full Range blend well and are in phase at that point.
This may be a quite stupid question, sorry for that. But willing to learn !
JMF
Attachments
The initial part of the impulse response is from the the highest frequencies, with the lower frequency contributions coming later on in the impulse (e.g. creating the longer tail). The initial negative peak just means the tweeter is in reverse phase compared to the input, and it looks like to the woofer as well. It's perfectly fine and normal and was probably done to get the system to sum properly at the crossover point, etc.
I'd be interested in knowing more about your setup: how large was the room, where were the speaker and microphone placed, how far apart were they, how high was the microphone off of the ground, etc. All those spikes in the tail are odd, but seem to soon in time to be reflections.
I'd be interested in knowing more about your setup: how large was the room, where were the speaker and microphone placed, how far apart were they, how high was the microphone off of the ground, etc. All those spikes in the tail are odd, but seem to soon in time to be reflections.
Thanks Charlie for the reply and the explanations.
I understand that even if it does not looks like a pure Dirac, it is a normal IR measurement for this type of design, aligned with the speakers design (reverse polarity of the tweeter compared to input and woofer).
Would this call for "reversed" input to have the first 100% peak positive ? Or is it fine as is (asked in a different way: is the wiring to the speakers OK or should it be reversed) ?
About the measurement and the room:
- Measurement was made at 30 cm of the speaker. Mic was at the height of the Tweeter.
- The room is 25 m2 about 6m x 4m. Speakers were moved about 1.5-2m from the wall. The ceiling is tilted and somehow low in that part of the room (about 2m ? not at home at the moment and can't measure).
JMF
I understand that even if it does not looks like a pure Dirac, it is a normal IR measurement for this type of design, aligned with the speakers design (reverse polarity of the tweeter compared to input and woofer).
Would this call for "reversed" input to have the first 100% peak positive ? Or is it fine as is (asked in a different way: is the wiring to the speakers OK or should it be reversed) ?
About the measurement and the room:
- Measurement was made at 30 cm of the speaker. Mic was at the height of the Tweeter.
- The room is 25 m2 about 6m x 4m. Speakers were moved about 1.5-2m from the wall. The ceiling is tilted and somehow low in that part of the room (about 2m ? not at home at the moment and can't measure).
JMF
To answer this question it might be instructive to view this impulse as frequency response, and also look at group delay?
There is no need or desire to change the wiring. What is most important in a loudspeaker is that the drivers are in phase at the crossover point (or at least no out of phase, which would cause cancellation). But as you move away from the crossover frequency, the phase of each individual driver become quite different, since a driver has its own phase rotation.
Unless the loudspeaker was designed to be phase coherent, meaning that it has flat group delay, then in my opinion looking at the time domain response such as the impulse response is not very useful. There is too much going on and there are multiple drivers contributing to the response. It's a little different for a single driver, but the frequency response has the same information in that case and is much easier to understand IMHO.
Could the wiggles in the tail of the IR be the reflections around the chassis, woofer just below and the pipe just behind it?
Thanks for the confirmation Charlie. So I will keep the wiring (as per design, and won't look too much to the Time Domain for normal usage). I may just confirm the phase between Woofer and Tweeter at crossover frequency by reversing the tweeter to check the inteerference notch and adjust the Tweeter delay in the DSP for deepest notch.