Maybe unlike you, I use 1/6th smoothing for x-over work. I may use 1/12th, or 1/24th after a while, or just to show off on the forums. 😎 For me, I see nothing here to change my mind, or conclusions. Any differences you see in these measurements would have no effect on a x-over decision. In addition, I think someone said that once a x-over is connected, any issue of this kind goes away. Well, when I do a x-over, that's about 5 minutes after the first sweep. The next 100 sweeps will all have a x-over which is being changed at each sweep.
Last edited:
What software are you using for measurments? You're sure you're not accidentally bumping something?
So far you're the only one who has taken data to see any difference. I think you should take a larger data set. Take 3 sets of measurements. All the same way but going back and forth between shorting and not shorting. Not three back to back with it shorted then 3 with it unshorted.
Go unshorted, shorted, unshorted, shorted, unshorted, shorted. This should reveal if its the method.
This looks to me like a measurement where I have accidentally moved something. I know this, because I have done this. This is why my method of measuremening ensures everything stays fixed in position with no possibility of movement. That's why I never move the mic or the unit under measurement. If you're removing the unit to rewire it for the next driver measurement then your data is going to be thrown off. You'll never get it exactly in the same position again. If you are moving the mic up and down or side to side to line up with a driver then again, no possible way to get it exactly back in place agian.
Cute.
Hey, we all aspire to different degrees of precision and sound quality.
Btw, if I was trying show off.......
The same 4 raw sections I posted @1/24th shorted and open, .......but after xovers, EQs, delays, and levels....iow, fully processed.
Acoustic measurements. Low Purple. Blue Mid, HF Green, VHF Red.....Sum Black.
Along with each section's impulse response.
1/48th smoothing on all...that's showing off!...😛
The set I posted were with Crosslite+. I'm well practiced with it, REW, Smaart, and ARTA.
Yes, absolutely sure nothing was bumped, moved, altered, etc.
I've taken multiple tens of thousands of measurements over the last dozen years, indoors and out.
I'm a dedicated student of measurements and multi-way speaker processing, regular attending training classes on FFT measurement software and best measurement practices.
Not trying to tout myself at all...just trying give you a little conversational introduction...
No need for your credentials. I'm just trying to figure out why out of 3 us performing a similar experiment, yours shows so differently than ours do.
I would say it's likely you have far better equipment than us but I'm not sure that would even matter since you're showing large differences. Multiple decibels difference. In my limited experience you need a serious resonating object, like a driver within the same airspace, to get to the sort of the difference you are showing.
Did all of these drivers have their own air space in your test? What did this unit look like? I have other builds so maybe I can replicate your findings if we have a similar styled design.
The design I ran my test on is quite unique and is probably not the best test subject as far as science is concerned.
The only significant measurement difference was on his tweeter near 1k, and the mid near 2k. The tweeter is probably 80dB down at 1k after the filtering, so it would have had no effect whatsoever on his x-over.
The discrepancy with the mid near 2k could be tweaked later with a mouse click when the design is measured. So, in my opinion, the 2k discrepancy would make no difference to the end results either, because you would see the effect eventually, and tweak it.
It was nice of Mark100 to suggest a possible reason for the problem seen in your measurements. Like all of us, he's here to help, and kill time. I actually enjoyed testing the idea. It didn't change anything in my measurements, but that doesn't mean that it can't change them on someone else's.
I still don't think it's an issue to be concerned about, but he does. We all do things our own way. If I attempted to convince you how to do a woofer to mid x-over, you would probably think I'm doing it wrong. I'm sure he would, and that's OK.
The discrepancy with the mid near 2k could be tweaked later with a mouse click when the design is measured. So, in my opinion, the 2k discrepancy would make no difference to the end results either, because you would see the effect eventually, and tweak it.
It was nice of Mark100 to suggest a possible reason for the problem seen in your measurements. Like all of us, he's here to help, and kill time. I actually enjoyed testing the idea. It didn't change anything in my measurements, but that doesn't mean that it can't change them on someone else's.
I still don't think it's an issue to be concerned about, but he does. We all do things our own way. If I attempted to convince you how to do a woofer to mid x-over, you would probably think I'm doing it wrong. I'm sure he would, and that's OK.
Last edited: