Measurement Length Changes Impulse Response? And a Sub Integration Question

I have no better insight as to what is best than you do. The advantage of this process is that you can now see exactly what is going on in this range and decide what alignment(s) you want to try out.

Here are my thoughts however in case you find them useful:

> Sub Polarity - Yes, the Sub is negative polarity. [FYI - The impulse peak reflects the highest frequency being measured; in this case 500 Hz. If the Sub is still in its passband then using the impulse chart is fine for this decision. If the Sub is well into its stopband and has a steep lowpass filter active then it may be misleading. It is better to look at the step response or the phase response.]

> Polarity to use - Your choice. Good alignments can be found either way. With an acoustic LR-24 or LR-48 XO setting both polarities positive is the alignment that is correct. As the acoustic filters often deviate from the ideal rolloff either setting may work as well. With enough deviation a better fit may be found with one driver reversed. The electrical filters chosen often do not provide the target acoustic shape.

> Direct sound or steady-state alignment - I initially find the best direct sound alignment and see how well it works for SPL. I also reverse polarity on the Sub and the reduce the delay of the sub (increase mains delay in this case) ~1/2 wavelength to find the best direct sound timing (~5 ms delay shift in this case). If the SPL support is good, I would use one of these 2 alignments. With room effects impacting phase at the LP and L vs R differences there are often other settings that in this range that provide as much
(or possibly more?) SPL support. These would all be good settings in my opinion. I am reluctant to shift the delay outside of this 5 ms range. We could find good SPL support with additional 1/2 wavelength increments, but the group delay offset starts to be excessive.

> Sub Position - Always a good consideration. Also a second sub to match the one you have may help a lot.

> XO frequency - Don't be reluctant to try a 120 Hz XO and see for yourself if there is distraction. With an acoustic LR-24 100 Hz XO I have no distraction. Others indicate they do. They may be more sensitive or may be they have more leakage higher into the mid bass, or possibly their room acoustics are different. It really only matters what you find in your case.
 
^Thanks for all your helpful comments @jtalden :T.
 
This is an old thread I know but I'll briefly revive it as it as I'm curious if the relatively new approach of having two acoustic timing reference signals (one at the start and one at the end of measurements) may be intended to have resolve my initial issue? I'm now routinely using a miniDSP SHD rather than the 2x4 HD I was when I started this thread, but testing again now and I get consistent impulse response shapes independent of measurement length.
 
This is an old thread I know but I'll briefly revive it as it as I'm curious if the relatively new approach of having two acoustic timing reference signals (one at the start and one at the end of measurements) may be intended to have resolve my initial issue? I'm now routinely using a miniDSP SHD rather than the 2x4 HD I was when I started this thread, but testing again now and I get consistent impulse response shapes independent of measurement length.
Yes, my understanding is that the trailing reference is added to correct for any clock difference.
 
Since your first post, REW has gotten a lot better in every way. There are many new settings.
 
Since your first post, REW has gotten a lot better in every way. There are many new settings.
Indeed, many of which I've used and appreciated. It's not that I've only just used REW for the first time in three years :). This relatively recent change reminded me of this issue as I did suspect the change was precisely for the reason @jtalden mentioned. Plus if it was I thought it would be good to tidy up this thread as a now solved issue should anyone else have been following it, or subsequently come across it.
 
Back
Top