[aceinc]

I was just thinking (scary...) it might be useful to be able to adjust sub integration in real time. What I mean is to have the mic at the listening position, your computer running REW near the sub, send 10-250hz pink noise through the appropriate channel(s) covering the speakers you are trying to match with the sub and the sub. REW would then have a real time graph, perhaps a bar chart with 24 bars per octave that you could monitor as you fiddled with the phase control on the sub.

Does something like this exist in REW and I am not aware of it?

Is it a reasonable idea?

I could do something similar with a sound analyzer app on a tablet, but the mic in the tablet isn't calibrated and I couldn't easily have the mic in the listening position while fiddling with the phase control. Perhaps I could cast the app to the TV, but still I have no idea how good the mic in the tablet is.

Right now I run a REW analysis pass, adjust the phase and run again, until I get tired of fiddling or it looks good.

One of my issues is I actually have a sub, and then my DIY mains each have an OB woofer cabinet with a "sub amp" driving the woofers so I have 3-8 times the fiddling depending on how pedantic I want to get.

Interested in what others think of my half baked idea.

 

[JohnM]

Sure, use pink periodic noise on the generator and monitor the response on the RTA with these settings:

 

[aceinc]

Thanks. It worked like a champ.

My mains use a GR Research OB powered woofer cabinet with two 12" drivers powered by a servo controlled plate amp. The mid & tweeter assemblies are of my own "design" which has an L-Pad for the tweeter, so there are a lot of controls which impact the sound. Using the twiddle - test - twiddle method really was not getting things dialed in very well. Using this method I spent about 20 minutes per speaker and voila, it was done.

Now I need to use the same methodology on the Rythmic Audio 15" Main sub dialed to dial it in with the rest of the system.

Thanks again.:T:grin2:

 

[nathan_h]

Sorry to ask a newbie question but what does a good RTA graph look like when doing this? When do I know my sub phase knob is in the right spot?

 

[aceinc]

Remember my use above is a little different than traditional sub integration.

I saw a dip in the frequencies in the 140-200 hz when the phase was "wrong" and as I changed the phase (rotated the phase knob) I was able to see the dip flatten out. I also used the gain control on the plate amp to bring the entire 15-140hz area up to the level of the rest of the "speaker."

When I use it for the actual sub in my system, I expect to see the same results but at lower frequencies. If I can get to it in the next few days, I'll "detune" one of my speakers, do some screen grabs of the graph before & after tuning.

 

[nathan_h]

That makes sense. Feels too simple!

Would be fascinating to see the screen grabs.

 

[aceinc]

Here is an example the first image is out of phase, the second in phase.

https://www.hometheatershack.com/forums/attachment.php?attachmentid=148996&stc=1&d=1559865243

 

[nathan_h]

Awesome. Almost TOO easy to see. Thanks.

 

[carlthess40]

When you are doing this test. Where do I set the mic at? Where I sit at when I’m watching a movie? I have two subs and one is next to the left speaker and the other one I want to place it next to the sofa as a end table or should I place it next to the right speaker?


Sent from my iPhone using Tapatalk

 

[ths61]

I was just thinking (scary...) it might be useful to be able to adjust sub integration in real time. What I mean is to have the mic at the listening position, your computer running REW near the sub, send 10-250hz pink noise through the appropriate channel(s) covering the speakers you are trying to match with the sub and the sub. REW would then have a real time graph, perhaps a bar chart with 24 bars per octave that you could monitor as you fiddled with the phase control on the sub.

Does something like this exist in REW and I am not aware of it?

Is it a reasonable idea? .... .

FWIW, I did a sweep of the sub by itself and a sweep of each individual main.
I then went into the OVERLAY Phase window and placed the vertical cursor at the SUB/Mains XO point (my case 40HZ).
I then enabled FDW (Frequency Dependent Windowing) and removed smoothing.
I then wrapped and unwrapped all phase plots (mains and sub) so they would be replotted and aligned at the Sub/Main XO point.
I then read the phase of each speaker at the XO point and then used that information to adjust the phase on the sub.
I put the sub's phase half way between my left and right main's phase (split the difference).

This gave me more control than just +/- 180 degrees. The sub is now within @ +/- 8 degrees of each of the mains.

You can then verify the results with a combined sweep to see the summing/nulling/smearing at the XO point.

You can also repeat the procedure to make sure the sub's phase moved in the direction and amount you intended it to (since the sub's phase knobs are usually coarsely marked).

There maybe better ways of doing it, but this is the procedure I used.

HTH

 

[aceinc]

FWIW, I did a sweep of the sub by itself and a sweep of each individual main.
I then went into the OVERLAY Phase window and placed the vertical cursor at the SUB/Mains XO point (my case 40HZ).
I then enabled FDW (Frequency Dependent Windowing) and removed smoothing.
I then wrapped and unwrapped all phase plots (mains and sub) so they would be replotted and aligned at the Sub/Main XO point.
I then read the phase of each speaker at the XO point and then used that information to adjust the phase on the sub.
I put the sub's phase half way between my left and right main's phase (split the difference).

This gave me more control than just +/- 180 degrees. The sub is now within @ +/- 8 degrees of each of the mains.

You can then verify the results with a combined sweep to see the summing/nulling/smearing at the XO point.

You can also repeat the procedure to make sure the sub's phase moved in the direction and amount you intended it to (since the sub's phase knobs are usually coarsely marked).

There maybe better ways of doing it, but this is the procedure I used.

HTH

I suspect your methodology is a much more accurate way to do this. However, the patience to execute the iterative process the number of times required to optimize things is beyond my ability.

One note, I did not just flip between 0 & 180 to tweak each woofer cabinet to associated mid/tweeter cabinet, I slowly swept the phase knob through its entire range, monitoring the RTA looking for the flattest response curve. My mains are open baffle with an open baffle amplified woofer cabinet. I did each main independent of the other with the mic about 1 meter (39 inches) from the main. This is what i showed in the graph above (one of the two mains).

I also have a 15" sealed servo sub which I tweaked by putting my measurement mic at the main listening position, sending a full range pink noise signal through the sound processor with all Audyssey off and the "speaker selection" set to fronts & sub. I then fiddled with the various controls on the sealed sub to mesh it with the fronts.

Once everything was complete, I reran Audyssey XT 32 on my sound processor to set everything back up. Sounds pretty good.

Again, your approach may achieve superior results for those with the patience to execute it. My approach seems to yield reasonable results with more immediate gratification.

 

[aceinc]

For a single or dual sub system, I would do my testing with the mic at the main listening position.

If you have a sub per main and they are located near the mains I may try the testing both at 1 meter & at the listening position leaving the settings so the sound at the listening position seemed flattest. But this would be mainly to satisfy my curiosity more than anything.

 

[aceinc]

One last note, I played with the volume and crossover frequency knobs to get the best results on all 3 plate amps. I found on my OB Mains I needed to fill in some frequencies in the 100-120 HZ range and adjusting the crossover frequency allowed me to do this.

Using the RTA approach really sped up the process, from making a change doing a sweep, analyze the results, and repeat. Doing it the iterative way, my personality is such that after a few passes, my eyes would glaze over and I would finally say, "OK, close enough, Audyssey will fix it."

 

[aceinc]

I recently did the tricks above to integrate a pair of Magnepan MMGs with my woofer cabinets. Here is the result;

It is played on an NEC CD Player directly attached to Emotiva XPA-1 (gen 1) Mono blocks. I set up this configuration when I want to hear speakers with the least amount of electronic wizardry "improving" things. I am playing at a rather loud volume and recording straight on to my Samsung S10 phone. The notebook in the foreground is running the RTA from REW. The scale is from 0hz to 20Khz horizontally and 25 DB to 90 DB vertically.

 

[cineburk]

There maybe better ways of doing it, but this is the procedure I used.

Not sure if this is better, worse, or just different, but here is what I came up with (after reading tons of different article and threads). After aligning the subs, I set the crossover on my receiver to 80Hz for the mains and the subs. I disconnected the subs and took a measurement of the center channel at the MLP from 65Hz - 95Hz. This was my reference.

Then, to align the subs to the center channel, I disconnected the center channel and took measurements (also from 65Hz - 95Hz) of the subs and adjusted the phase using the sub distance setting in the receiver.

In the charts below, you can see the phase graph, and the impulse response. The red trace is the center channel, and the other traces are the subs at different distance settings (you can see the distance in feet in the measurement name).

You can see in the charts that the closes phase seemed to be at 43.0', but the closest impulse response was at 41.0', so I split the difference and went with 42.0'.

 

[Matthew Burton]

How do you do the Frequency dependant windowing FDW? Can't figure it out.

 

[cineburk]

Have you checked the settings here?

https://www.roomeqwizard.com/help/help_en-GB/html/analysis.html

 

[JohnM]

How do you do the Frequency dependant windowing FDW? Can't figure it out.

Just select it on the IR windows dialog.