Summing at crossover - LR24db/BU24db

[Cathul]

Don't know why. When comparing the measured distances transfered to ms to the measured impulse response in ms it's not far off, but i still have to flip the polarity to get a good summation at the xover.
And it wasn't just a small dip, it was a deep null when i measured with periodic pink noise of about 30db right on the crossover point with 1/48th octave smoothing before i flipped the polarity.
Maybe the drivers are out of polarity (although i checked this)? Maybe i need to check polarity again with everything defeated, crossovers, equalizers, time delay, everything.
The installer used BW 12db slopes on the initial tuning and didn't flip the polarity of any speakers and did only a minimum of equalization, so maybe he switched the physical connections on purpose? I don't know.

After all, the alignment tool is too far off in the xover region between midbass/tweeter to be correct imho. 2ms of delay on the tweeter where measured distances to the microphone tip in listening position are 71cm for the left tweeter and 104cm for the left midbass cannot be correct. Jazzis spreadsheet gives a delay relative to the right midbass of about 1.25ms for the left midbass 2.21ms of the left tweeter, 0ms on the right midbass and 0.47ms for the right tweeter. Put an additional 2ms delay on the tweeters on top of that... nope... too much delay. This would mean the left tweeter is 1 cm away from your head.

Good think is that now i'm only "missing" 2db at the midbass/tweeter xover which is the best i've ever achieved i think.
And it sounds oh so good now. Listened to "Chocolate Chip Trip" from Tool after i was done for today. Oh my, that drum really kicked when Danny Carey went crazy on the drums.

[jtrosky]

Thanks - and don't take me the wrong way - I'm not doubting you - I'm just trying to understand it a little better. To me, phase is some magical thing that I have a hard time grasping for some reason. :-)

On the subject of drums - something else I've noticed when playing with different crossover frequencies lately (xover between 6x9 door midbass speakers and 3.5" dash speakers) - the higher I set the xover freq, the "fuller" everything sounds. I guess the smaller speakers in the dash just don't have the "uumph" of the 6x9's in the 300hz-500hz range (even though they measure at the same volume level). It's like I also "feel" the 300hz-500hz area freqs from the 6x9's whereas I just "hear" them from the 3.5" dash speakers - and it just makes everything sound more "powerful" (again, even though those freqs play at the same volume level regardless of which speaker plays them).

I *was* trying to get the dash speakers playing *lower* - but it seems that I actually get better results the *higher* I set the xover (to a point, obviously). It's strange since the measurements will *look* the same...

[mauian]

It fun to see all this discovery going on, but please remember your results are really good. You're well into splitting hairs territory now so don't get too frustrated if you can't fix every last little thing!

I cannot overstate how thankful I am for this perspective from you. My OCD (along with OP’s probably) would kill me, so to hear that from you goes a long way to keeping me sane


Sent from my iPhone using Tapatalk

[mauian]

Hmmm.... I've been seeing people "flip" the polarity of a single speaker lately, in order to correct a small dip in the overall response. If time alignment is set correctly, we shouldn't need to flip the polarity of any speakers, should we? I thought that the whole purpose of time alignment was to get all of the speakers 'in phase" at the listening position. If time alignment is set by measurements, how could the time alignment be *so* wrong for a speaker that we need to do a 180 degree phase shift for an entire single speaker? Wouldn't that suggest that the time alignment is way out of whack for that speaker?

I could understand maybe needing an allpass filter to correct a phase-related issue at one small area of the response (if it truly is caused by a phase issue), but having to do a 180 degree flip of the polarity of an entire speaker? That just doesn't make sense to me. I thought that was the best we could do before advanced DSP's were available, but that it wasn't really a common practice now that we have very powerful DSPs to help with timing and phase?

I'm sure that I'm just misunderstanding something though. Although, I've never had to flip the polarity of an entire speaker in my car, which is why I'm kind of curious...

Thank you!

In my car I’ve had to flip the polarity, I think, because they were actually flipped in the wiring. A lot speakers don’t clearly mark positive vs negative and the terminals can be the same size making it difficult to tell on sight, also the wiring can get flipped at the amp or dsp by accident.

I’ve verified polarity with a click/pop test but I don’t think these are very reliable.

Also, I think the car environment, with reflections, can alter the polarity or phase of the sound waves.

As I’ve learned recently, always good to verify this with measurement.

But you’re right, the time alignment with measurement is important to get right.


Sent from my iPhone using Tapatalk

[mauian]

Don't know why. When comparing the measured distances transfered to ms to the measured impulse response in ms it's not far off, but i still have to flip the polarity to get a good summation at the xover.
And it wasn't just a small dip, it was a deep null when i measured with periodic pink noise of about 30db right on the crossover point with 1/48th octave smoothing before i flipped the polarity.
Maybe the drivers are out of polarity (although i checked this)? Maybe i need to check polarity again with everything defeated, crossovers, equalizers, time delay, everything.
The installer used BW 12db slopes on the initial tuning and didn't flip the polarity of any speakers and did only a minimum of equalization, so maybe he switched the physical connections on purpose? I don't know.

After all, the alignment tool is too far off in the xover region between midbass/tweeter to be correct imho. 2ms of delay on the tweeter where measured distances to the microphone tip in listening position are 71cm for the left tweeter and 104cm for the left midbass cannot be correct. Jazzis spreadsheet gives a delay relative to the right midbass of about 1.25ms for the left midbass 2.21ms of the left tweeter, 0ms on the right midbass and 0.47ms for the right tweeter. Put an additional 2ms delay on the tweeters on top of that... nope... too much delay. This would mean the left tweeter is 1 cm away from your head.

Good think is that now i'm only "missing" 2db at the midbass/tweeter xover which is the best i've ever achieved i think.
And it sounds oh so good now. Listened to "Chocolate Chip Trip" from Tool after i was done for today. Oh my, that drum really kicked when Danny Carey went crazy on the drums.

Yes! Danny Carey is why I started playing drums. Amazing.

Thanks for sharing your process above too. Very helpful to see how you worked through that.


Sent from my iPhone using Tapatalk

[ckirocz28]

In the past few days i tried a lot of things to get a better summing at the crossover between midbass and tweeter.
Acoustical crossover is at 2300 Hz.

I tried nulling with phase switching, impulse response and "brute force" by going 0.4ms (should be a bit more than one full cycle on the tweeters and a xover frequency at 2300Hz) up and down in the smallest steps possible with the Mosconi D2 100.4 DSP amplifier (0.02ms) from the calculated time delay values from Jazzis spreadsheet after measuring the distance again from the tip of the microphone to each speaker. After measuring i checked the polarity of the drivers with a polarity checker. Polarity was uniform on all drivers.
Never, not even once was i able to get a 6db sum on either side between midbass and tweeter. Most i could get were the current 4db i have right now.
Now i start wondering if adding a dedicated midrange would help me as the midrange is much nearer to the tweeter than the midbass.

Current locations of the drivers in my S550 Mustang are the original locations midbass down low in the doors and tweeters in the stock A-pillar location.

Don't get me wrong. It sounds great right now. Very good staging, strong bass/midbass integration and overall sound, but the missing 2db really nug me.

Like Justin Zazzi said, 4 db summation at 2300 hz is pretty good. 6 db summation generally only occurs when the speakers are within a quarter wavelength of each other, at 2300 hz that is slightly less than 1.5 inches. From your lower door to your A-pillar is considerably further than that. You probably could tweak that up a little with an all pass filter, but again, like Justin said, your splitting hairs at this point. If it sounds great, I'd just leave it alone.

[Cathul]

I know, i know... it is hair splitting... but (!) it's always microscopic things like these that keep you going forward and try something else and (hopefully) learn something in the process.
My wife always gives me the dirty look when i go to the car, again(!), just to tweak it some more. The first few times she always yelled at me, by the time she doesn't say a word anymore, but still gives me dirty looks.

[jtrosky]

My wife always gives me the dirty look when i go to the car, again(!), just to tweak it some more. The first few times she always yelled at me, by the time she doesn't say a word anymore, but still gives me dirty looks.

That's too funny - my wife is the same way. I'll say that "I'm going out to the car to take a measurement" and I'll usually get some kind of remark like "oh, going out to see your girlfriend again?". Obviously, the "girlfriend" that she is referring to is my car. :-)

So you are definitely not along, my friend!

Although, I honestly do most of my tuning in the house now. I have things setup where I really only spend enough time in the car to take measurements - then I "analyze" them and modify EQ in the house, on my full-size home PC, in the comfort of my office chair - with a mouse, keyboard and a large screen. :-) Obviously, I do have to spend a little extra time in the car after I get things really dialed in for listening tests and minor fine-tuning, but I get most of the EQ work done in the house now using REW. Better that futzing with it in a hot car while draining the car battery.

I actually typed up the process I use for another user that was interested in "in-home tuning" - I will gladly share it with the forum if there is any interest. Works really well - especially for the bulk EQ work.

[Justin Zazzi]

I don't think it's a new thing. If you have a set of measurements go to All SPL, click Controls at the upper right, then click Alignment Tool. A little window will open which is the Alignment Tool. Select the 2 measurements you want to align from the drop-downs. In the SPL window of the measurements, click your cursor where you want to align. In my screenshot I have it around 68Hz. Make sure you have Show Phase Traces checked in the Alignment Tool then click the Align phase at cursor button. It will calculate the amount of delay you need on one of the measurements and the white line in my screenshot shows what the response would be with that delay. You can play around with the delay slider and watch the phase and response change. It's a neat tool. I don't know everything about it but I've been playing with it to see the effects when I adjust delay. If you click the Aligned Sum button it will generate a new measurement.

Justin, you're the expert here being a sound engineer. Perhaps you can share with us how to best utilize this tool when tuning.

Attachment 11612

Okay, I tried this feature today while tuning a car with only REW (no smaart, no multi-mic array, no fancy tools).
This feature is amazing!

Before following B518's instructions above, try looking at the impulse responses to ensure you have the polarity correct for all of your channels first. Then use this tool to do time alignment. In addition to B518's instructions I would also recommend:

When you select the two traces in the drop-down menus to align, try also displaying those two curves on the graphs on the top half of the screen. It's not automatic and you have to do this manually. The black line that shows up on the top and bottom graphs (spl and phase) is the expected result of the alignment tool. The two colored lines on the phase plot will match the two colored spl traces on the SPL graph up above. These are the two measurements that the align tool will be working with.

Place cursor at the crossover frequency where the two speakers overlap the most. Then press the button "level phase at cursor" to remove the common delay of both measurements. For me, this made the phase plot much easier to read in that frequency range. Then press the button "align phase at cursor" which will attempt to time align the two measurements in that vicinity of your cursor on the phase plot. You should end up with the two colored phase traces very close to eachother and overalpping at and near the cursor, and the black line should be fairly smooth and flat in that region too.

The tool will adjust the delay sliders by some amount and that is the delay that you need to enter into your DSP. If the delay for one channel is zero then don't change that one. If the delay is positive then add it to whatever delay you ALREADY have on that channel. If the delay on one channel is negative, then either subtract it from what you already have on that channel or add it to the OTHER channel.

If you don't get results that look useful, try adding smoothing, like a LOT of smoothing up to 1/3rd octave.

This tool is super freaking cool and certainly take REW into a whole new realm of possibilities. I'm so impressed! This is very much like SMAART and it has a few new tricks that I wish SMAART had too!

[Cathul]

Did you try to align left/right or only one side drivers only?