Hi Camplo, sorry I can't help with wherever you seem to be headed....(which frankly I can't make out.)
Besides and more importantly because i don't think the subs are worth agonizing over,
I've come to the conclusion i don't think it's a good idea to take a CD as low as it can go..
I think we have to think of excursion having a negative effect (modulation), in terms of how many octaves above the max excursion freq does the excursion modulate. Your drive to take the axi lower seems to me, to only magnify potential adverse modulation..
Besides and more importantly because i don't think the subs are worth agonizing over,
I've come to the conclusion i don't think it's a good idea to take a CD as low as it can go..
I think we have to think of excursion having a negative effect (modulation), in terms of how many octaves above the max excursion freq does the excursion modulate. Your drive to take the axi lower seems to me, to only magnify potential adverse modulation..
I knew you were of this thinking, and I don't try to contest it. It gives me something to think about.
On the other hand, I've read @gedlee saying that THD would have been issue well before IMD becomes audible.
Excursion is really low for the compression driver, so THD isnt an issue.
I wonder is there a way to test accuracy while applying a broad spectrum signal.
That's easy for me. I'd never use anything with slots - I see no advantage only disadvantages.
Multiple sealed 15's is my opinion. I thought that I'd said this before.
You cannot compare THD to IMD because they are both the same things. Just different ways of viewing the same nonlinearity.
What I do believe is that it is LF modulation of HFs (from the nonlinear excursions) that will be the most objectionable aspect of driver nonlinearity. "Muddy" as we said. This is not going to happen in a compression driver because the excursion is low and the bandwidth not too large. Which is why I would not take a compression driver too low as this tends to aggravate this issue.
In a woofer the upper edge of the passband is the biggest issue, which is why inductance nonlinearity is such an audible problem. Lower bandwidth here helps the woofer but hinders the compression driver.
I guess that question is, will offsetting the entry point into the transmission line be better because it seems that shuffling the modal contributions in that way makes for a smoother response?
You'll find it's a popular approach because people are looking for a response which is easy to work with. I don't know that it has to be that way, after all the first mode gets relatively reduced, and that's the one you want. The other modes just seem to be in the way, so it seems to be popular as a practical compromise.
Are you suggesting that at low excursion, nonlinearity will be low and without nonlinearity the two won't have the condition to intermodulate... or something else?
That is the implication, but it has to be noted that low excursion is not a guarantee of low nonlinearity. That was a key conclusion of our study of nonlinearity perception. It is also the idea behind the "first watt" concept for amplifiers. High nonlinearity at low levels is by far (BY FAR) the most insidious type of distortion and absolutely has to be minimized.
DSP can flatten the phase, the amplitude, or both. In theory it is very flexible. In practice it has some limitations.
You really need to think through this question. That "jump" is not real, it is an artifact of the graphical presentation.
Well when I plug this response into VituixCad, its messing up the summing between it, and the 15"midwoofer....Or atleast I think it is...
I am not giving enough details I bet, give me second...
Last edited:
It's a "phase wrap", meaning the representation of phase stays within 360 degree, from +180 to -180, see the index on the right. Each time there's more phase rotation than either +180 or -180 you'll see a dotted vertical line and can trace it from there on. That's the "wrap" part. You can "unwrap" the phase in REW.
So that sudden jump you outlined isn't a phase jump, it just means you've passed the 180 degree line, you can continue following that graph at the bottom line at -180 degree.
You wouldn't want to use DSP to linearise a "phase trace" of a 500 ms measurement window though... at least, I wouldn't. You don't see the phase result of the driver by itself, you see the phase as generated by the total measurement, with the room noise and all reflections included.
So that sudden jump you outlined isn't a phase jump, it just means you've passed the 180 degree line, you can continue following that graph at the bottom line at -180 degree.
You wouldn't want to use DSP to linearise a "phase trace" of a 500 ms measurement window though... at least, I wouldn't. You don't see the phase result of the driver by itself, you see the phase as generated by the total measurement, with the room noise and all reflections included.
Last edited:
Ok, heres what summing looks like in the raw,
Sub inverted.
Thank god its alot better with the XO I used but I feel like summing drops off near 100hz (in the current xo design) and I blamed the phase....when I look at the phase I see what I showed in my last post....
Am I going to be able to create perfect summing or not? its just a matter of having everything in phase right? Is this normally or easily accomplished with my situation is what I am trying to research.
Sub inverted.
Thank god its alot better with the XO I used but I feel like summing drops off near 100hz (in the current xo design) and I blamed the phase....when I look at the phase I see what I showed in my last post....
Am I going to be able to create perfect summing or not? its just a matter of having everything in phase right? Is this normally or easily accomplished with my situation is what I am trying to research.
I don't like how I have some cancellation here, right before 200hz....I also am going to try to design a XO bringing the PPSL to 200hz....but I am saying that I would like it to sum perfectly with the 15".... in order to do that, phase alignment has to take place....Is that a big deal? Meaning, should it be easily accomplished or am I trying to perform brain surgery?
All my current measurements are on the pos laptop upstairs I need to send at least the measurement of the horn running to 200hz, to my main computer so I can work on a XO. But right now, If I can see the potential of DSP to create perfect driver summing, then I can plan my next moves regarding the woofers.
I would think that I need to make the PPSL phase look more like this (15m)
I pointed to the anomaly in the phase between 100 and 200 as culprit but coincidence does not equal causation, so I come to the board.
All my current measurements are on the pos laptop upstairs I need to send at least the measurement of the horn running to 200hz, to my main computer so I can work on a XO. But right now, If I can see the potential of DSP to create perfect driver summing, then I can plan my next moves regarding the woofers.
I would think that I need to make the PPSL phase look more like this (15m)
I pointed to the anomaly in the phase between 100 and 200 as culprit but coincidence does not equal causation, so I come to the board.
Last edited:
I have thoughts but I recognize that this whole IMD is more complicated than can be imagined at face value...I am pretty certain that there is IMD and then Excess IMD? I also think that FM is more of an issue than AM. FM happens at low level, and is easily induced by multiple frequencies that are near each other.
Any truth to those ideas?
"trying to including rear subs to show final excursion"
There are rear subs, planned for the final configuration. Those subs will be ran as high as possible 120hz-200hz is my guess. So, when judging excursion character of the mains I included the acoustical output of the planned rear subs. The green line represents 8 18's (forced flat FR). the mains together have 4 18"s, and another 4 18's to represent the rear subs. 4+4=8. So if I ran vertically stacked 18"s up front. The green line represents excursion required for 116db/1m. Running these 18"s to the XO at 200-300hz. Everything would be Vented in this scenario
Or I could just leave it in its current configuration, of 15" over dual 18's in PPSL. If I can create perfect summing using DSP, this configuration is much more efficient as long as the 18"s are vented, but even if they aren't the excursion never touches the 15" so headroom in the midrange is higher.
Headroom is not a problem for either scenario. Its been suggested that fixing phase to cause perfect summing between the drivers should not be an issue, so it kinda settles it for me. I can also leave everything sealed and still not worry about headroom due to the high xmax of the 18"s....and then it would be up to me to decide if I need the increase in lower frequency headroom. If you look at the light blue, that is 8 sealed 18"s.....this is what the excursion of the mains would look like in combination of the rear subs, flat FR at 116db/1m.... I think could leave everything sealed and be ok.
