The amplitude is flat because the voltages sum together and the power has a dip because they are uncorrelated and don't add in the same way. Your last sentence is the answer. My post above yours had the best explanation I can give as to the why.
There is a much greater difference between an LR2 and an LR4 than the difference in phase alone. The drivers are summed over a wider range and the power response can be very different depending on the directivity of the drivers. These factors would have a much more audible effect. To hear the effects of phase on a crossover and keep that the only variable is a better but slightly more difficult test.
No worries there, thanks for the clarification. Unfortunately I quit university (ee) way back then to become a software engineer, so I have to learn all these mathematical concepts on my own. Especially if you're developing your own crossover software that can be painful at times
Thx fluid,
I need to test the lack of correlation with acoustic level summations.
Is there any other counter for the lack of an efficiency gain from two sources back to level, other than the split signal being non-correlated?
The non-correlated bit just doesn't make sense still..
I'm not sure how to do such a test with something other than a sub, where on-ax is probably a decent proxy for total acoustic power.
I need to test the lack of correlation with acoustic level summations.
Is there any other counter for the lack of an efficiency gain from two sources back to level, other than the split signal being non-correlated?
The non-correlated bit just doesn't make sense still..
I'm not sure how to do such a test with something other than a sub, where on-ax is probably a decent proxy for total acoustic power.
I've read alot of Rod's excellent work. Pls point me directly to the info he gives that provides the context you refer to.
Put a two way speaker with a basic LR2 or LR4 crossover in a room and take a Moving Mic Measurement. You will see a dip in the response at the crossover point.
This is power response so one single axis will not necessarily reveal it. In room response has enough similarity to a power response to show it.
Ok, if it measures, that will be easier to detect than i thought possible...
I totally get it has to be about power response, cause I've seen too many on-ax measurements to know there is no dip on-ax.
Hey, if acoustic power response dip is really so.. it appears to be one more reason to go steep. Less width in the power response loss area.
Don, this is the longer quote from Self which seems much the same?
"The power response of a loudspeaker is the sum of all its off-axis and on-axis amplitude/frequency responses; it is the frequency response of the total acoustical power radiated into a given listening space. Because the signals are not phase-correlated, phase is ignored and summation is by RMS addition of the highpass output VHP with the lowpass output VLP: "
"The power response of a loudspeaker is the sum of all its off-axis and on-axis amplitude/frequency responses; it is the frequency response of the total acoustical power radiated into a given listening space. Because the signals are not phase-correlated, phase is ignored and summation is by RMS addition of the highpass output VHP with the lowpass output VLP: "
Attachments
Any direct quotes from in there, or directions to relevant info there, per this conversation/thread?
Thx.
I've seen it in so many measurements that I'm sure you will be able to see it too. The lower the order the more range it will cover.
I think this is a good way of stating it. The rest seems like getting hung up on semantics.
Plas, like this? I'm new to FIR but will test it later on, the bad thing, the memory flip is 2 sec silence on OpenDRC-DA8.
The best i think is to record the signal 2 times and flip between the same song.
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.