Need help For subwoofer alignment. Best someone who understands a tittle German

[Foly83]

I am looking for help with the connection of my subwoofers (2 pcs.)
Unfortunately, my English is not very good.
Could just write everything with the Google translator.
Is there a professional who can help
Messequip:
UMIK, Mini DSP 2x4HD

I already have experience with measurements, but 2 subwoofers make it difficult for me to bring them perfectly in phase.
What does the perfect phase look like?
what do I have to look at and how.
How do I proceed in my constelation?


Hallo
Ich suche Hilfe in der Anbindung von meinen Subwoofern (2stk.)
Mein Englisch ist leider nicht sehr gut.
Könnte alles nur bei dem Google Übersetzer schreiben.
Gibt es einen Profi, der helfen kann
Messequip :
UMIK,Mini DSP 2x4HD
Ich habe schon erfahrung mit Messungen, aber 2 Subwoofer breiten mir doch schwierigkeit sie perfekt in Phase zu bringen.
Wie sieht die perfekte phase aus? was muss ic h wie betrachten.
Wie gehe ich in meiner konstelation vor?

 

[jtalden]

The translator works reasonably well at least on this end. It does take more time.

Step-1 is to setup your 2 subs to work together favorably as group.
Your 2 subs are already time aligned and working well together. This is the result of; their locations are equidistant from the LP and they are reasonably favorable room locations relative to your LP. Your posted measurement file confirms this per the attached charts. There is always a possibility that a different delay setting or polarity settings in the MiniDSP will slightly improve the SPL smoothness and this depends on many room setup factors. The REW alignment tool (AT) is a good way to virtually investigate the impact of those changes. In this case, I would not expect significant improvement given these sub and LP location choices.

There is almost always improvement to be found with sub and LP location changes. Many of us do not have very many practical location options for subs or the LP nor do we have the willingness to investigate very many options.

Step-2 is to get the correct measurements to determine a favorable XO handoff from the 2 subs to the left and right main speakers.
Are you ready to start step-2 or do you intend to do further work on Step-1?

 

[Bernard]

Hallo Foly83,
Sie können DeepL verwenden, das von Ihren Landsleuten erstellt wurde und nach ISO 27001 zertifiziert ist.

DeepL Translate

N.B. : Es ist keine Werbung, es ist Information!
Is my answer in German correct?

 

[Foly83]

Can you please explain to me why the step response is helpful here?
Is the first rash important? Or the biggest rash? I've read a lot that it is too imprecise.
When I look at the phase of subwoofer left and right, I would guess that at an Xo of 80 probably the interaction is no longer so good
Where I ask myself again the question, where do I put the Xo best. I would tonight as soon as I'm home. Measurements of different Xo make if necessary. Just let me know and I will get it done quickly

Otherwise we are welcome to start with the next step. I would do this now.
1. Align sub left & right
2. Sub's Eq
3. Align front left & right and Eq, the volume etc everything as good as possible the same.
4.Xo find out which one is the best.
5. Adjust Front & Subs Timing.
Correct me if I'm completely wrong.

I would also be happy if you could show examples with the phase.
It is best that everything stays on the temporal level.
Which is sometimes very difficult to implement.

Thank you for taking so much time :Sprung:

I just noticed how big the time difference is. It's 7:49 for us right now. Have a nice evening and see you later

 

[Foly83]

Hallo Foly83,
Sie können DeepL verwenden, das von Ihren Landsleuten erstellt wurde und nach ISO 27001 zertifiziert ist.

N.B. : Es ist keine Werbung, es ist Information!
Is my answer in German correct?

works really well. I usually always use the google translator. But if this one works better, of course I'll use this one.
Thanks a lot

 

[Berry 76]

The translator works reasonably well at least on this end. It does take more time.

Step-1 is to setup your 2 subs to work together favorably as group.
Your 2 subs are already time aligned and working well together. This is the result of; their locations are equidistant from the LP and they are reasonably favorable room locations relative to your LP. Your posted measurement file confirms this per the attached charts. There is always a possibility that a different delay setting or polarity settings in the MiniDSP will slightly improve the SPL smoothness and this depends on many room setup factors. The REW alignment tool (AT) is a good way to virtually investigate the impact of those changes. In this case, I would not expect significant improvement given these sub and LP location choices.

There is almost always improvement to be found with sub and LP location changes. Many of us do not have very many practical location options for subs or the LP nor do we have the willingness to investigate very many options.

View attachment 42090

View attachment 42091

Step-2 is to get the correct measurements to determine a favorable XO handoff from the 2 subs to the left and right main speakers.
Are you ready to start step-2 or do you intend to do further work on Step-1?

I'm also integrating 2 subs with a mini dsp. I have a lot of room issues. How do you go to "step response"?

And if im correct you have some dutch roots ? because you named your left en right sub with te dutch words.

Greeting from te Netherlands

 

[Foly83]

@Beere 76 Hello
My measurements are labeled in German
You can get to the Step Response via the overlay

 

[jtalden]

Can you please explain to me why the step response is helpful here?

It's not particularly more important than other graphs. I could have posted; Impulse, ETC and phase as well. All show the timing of those 2 subs to be timed similarly. We expect that when the subs are at equal distances from the mic or when the difference in distance is adjusted using delay to the sub that is closer to the mic.

Is the first rash important? Or the biggest rash? I've read a lot that it is too imprecise.

The rise is one good thing to look as it represents the first arrival of the direct sound. It does not establish the most favorable timing however as the room will influence the bass range differently for the 2 subs in many cases. I use this time aligned situation as the reference condition or starting point and then use the REW AT to determine if a better delay offset can be found. We are looking for smooth SPL through the bass range. The Impulse peak of Impulse is much more likely to be misleading by several ms than the initial rise. I personally use the phase with a 4-6 cycle FDW to establish the direct sound timing as shown in the REW AT. That starting point is used to assure we do not change the delay too far away from that timing. We should stay as close as possible in the search for the best SPL with the FDW disabled. I try to stay within 1/4 cycle of the XO frequency, but there may be exceptions that are still favorable.

I do understand why the initial rise could be too imprecise as a starting point in some cases. In this case it is perfectly clear. It is best to look as all the evidence.

When I look at the phase of subwoofer left and right, I would guess that at an Xo of 80 probably the interaction is no longer so good
Where I ask myself again the question, where do I put the Xo best. I would tonight as soon as I'm home. Measurements of different Xo make if necessary. Just let me know and I will get it done quickly

We can work on optimizing the XO for years as a hobbyist willing to spend lots of time for learning and fun. The REW AT can usually quickly identify one or two favorable settings where we can see and understand why the settings are likely the best choices. There are room setups where there is no favorable setting that can be found and then the only choice is to change the setup or the room or just accept the situation.

Adjusting the XO frequency is often helpful, but the interaction between the subs and mains makes predicting this in advance difficult. Yes, there appears to be the likelihood of a sag in the SPL response around 80-90 Hz using an 80 Hz XO. It is difficult to be sure how bad that will be. The mains my help or worsen the issue and the house curve you chose will impact it. A lower XO may turn out to be more favorable, but only investigation of the various choices will determine that.

Otherwise we are welcome to start with the next step. I would do this now.
1. Align sub left & right
2. Sub's Eq
3. Align front left & right and Eq, the volume etc everything as good as possible the same.
4.Xo find out which one is the best.
5. Adjust Front & Subs Timing.
Correct me if I'm completely wrong.

Directionally good, but don't get too far ahead. One step at a time so we are on the same page.

I would also be happy if you could show examples with the phase.
It is best that everything stays on the temporal level.
Which is sometimes very difficult to implement.

I am not sure what you are looking for. Can you clarify what the the phase question is.

 

[Foly83]

Hello Jtalden
Thank you for the informative text and also still so much.:Redgrin:
I have created myself the measurements with 2M to eliminate influences in the room something. For this I also use FDW on mostly on 5 or 6.

On the subject of phase.
I have made a pair of pictures.
Please have a look at the phase example picture.
I just set the delays as the phase is relatively close to 0. But when I look at my other two examples. (delay 7.77 & delay 1.66) they are not close to 0, which is also impossible.With normal measurements you can hardly get the phase at 0, it always moves downwards.
With an FIR equalized system almost everything is close to 0.
This confuses me a lot. FIR is again another big topic, which I would like to consider only if I have the necessary basic knowledge
(which can last still for a long time :unschuldig: ).

I made an extra measurement with Xo80 and Eq's
With this I have set both Bespiele (7.77 &1.66).
Which of the two delays is right and wrong? Or are both correct?
With example delay 7.77 the sub is additionally inverted.
I am very curious which delay is wrong and why.


Additionally I have created several measurements
With
Xo70 ,
Xo 80,
Xo90
Without Eq's without further settings.
So we have a clean start how you would proceed.

 

[Foly83]

I also looked directly at my measurements.
I would put Xo70 completely in the trash.
For 80 & 90 I have attached the two best results as a picture

 

[jtalden]

I have created myself the measurements with 2M to eliminate influences in the room something. For this I also use FDW on mostly on 5 or 6.

256 kHz is plenty for this type of work. 2 MHz is fine also, but brings no particular advantage.

On the subject of phase.
I have made a pair of pictures.
Please have a look at the phase example picture.
I just set the delays as the phase is relatively close to 0. But when I look at my other two examples. (delay 7.77 & delay 1.66) they are not close to 0, which is also impossible. With normal measurements you can hardly get the phase at 0, it always moves downwards.

I think you misunderstand. Step-1 is time alignment of the 2 subs so they work well together. This is what we were discussing until now. Step-1 is then completed. We have time aligned the 2 subs favorably and assured the SPL without FDW is favorable.

The AT delays shown in the charts of -1.60 ms and -7.98 ms with the SW inverted are regarding Step-2, i.e., the timing alignment of XO. There is no relationship to 0 ms as a known starting point for that process. The starting point for the XO timing is found just as you did here in the 2 charts. The FDW is applied and close phase tracking is found. This shows you know how to use the AT to find a good starting point.
The problem is that you did it for only the left channel. It is best to use the AT to sum the 2 subs and to sum the L and R mains. Then use those 2 new measurements to determine the starting point timing for the XO timing.

With an FIR equalized system almost everything is close to 0.
This confuses me a lot. FIR is again another big topic, which I would like to consider only if I have the necessary basic knowledge
(which can last still for a long time :unschuldig: ).

Yes, it is a big topic that is best set aside until the basic setup is completed.

I made an extra measurement with Xo80 and Eq's
With this I have set both Bespiele (7.77 &1.66).
Which of the two delays is right and wrong? Or are both correct?
With example delay 7.77 the sub is additionally inverted.

Additionally I have created several measurements
With
Xo70 ,
Xo 80,
Xo90

I suggest the following course of action since I am not willing to evaluate numerous possible scenarios for you.
We work through the 80 Hz XO for one of the 2 favorable starting delays. I think after one XO option is done you will be able to investigate others on your own if you like.

Suggestion
[Step 2: Find favorable SPL for the XO range by evaluating delay, gain and polarity of L+R vs SW1+SW2.]

The following 4 measurements provides all the measurements needed to complete this analysis.

Setup:

• Mic at LP
• 80Hz XO active in the AVR (main speakers set to small)
• REW Acoustic timing: active
• Sweep full-range for all 4 measurements (20-20k Hz for example)

Measurements:

1. Left main (alone)
2. R main (alone)
3. Sub1 (alone)
4. Sub2 (alone)

Analysis:

1. REW AT sum left and right mains
2. REW AT sum sub1 and sub2
3. REW AT find one or 2 starting point delays using a FDW

Post the measurement file results and your delay finding here for review before further analysis. I will also do this analysis to confirm your selection.

 

[Foly83]

Hello
The time difference of 7 hours is annoying

Thanks for the information
(256 kHz is plenty for this type of work. 2 MHz is fine also, but brings no particular advantage.)

I hope I have understood it correctly, the translator works with some words somewhat wrongly

I had created the other measurements on occasion with, that I do not have to make measurements again.
I was unfortunately too fast, please excuse
An intermediate question, which is perhaps embarrassing.
Why does every measurement have to be made from 20-20,000 when I separate the sub with Xo80 anyway.
i have compared measurements from 20-300 and 20-20,000 and it looks identical to me

So, I recreated a complete measurement
Xo80 and from 20-20.000
All channels individually
left and right together
sub's together
FDW I have set with 5 at analysis 3.

Here are my results:
Analysis

1.
front left -2.32 / front right 0
2.
sub left -1.97 / sub right 0
3.
front left & front right 0

version 1. sub left & right -2.48
Version 2. sub left & right -8.57 inverted
But I can't tell myself if version 1 or version 2 is better.

If my theory does not fit and you have the time you can also explain it to me on my new measurement? I hope I do not cause too much problems.

 

[jtalden]

First let's start with getting good measurements for this process. There appears to be 2 problems with the current ones. See the impulse chart below. this is the left speaker with XO 80Hz of Post-9 Vs the same speaker in Post-12. I vertically offset the Post 12 trace about -20% so that it is easier to see the differences in the 2 traces.
1 - The acoustic timing is not working correctly: t=0 is located on the first peak on Post-9 and the second peak in Post-12.
2 - It appears L and R speakers have been measured together rather than alone: There are 2 peaks of almost equal size on both traces. measurements.

The 2 right speaker measurements has these same problems (not shown). Please investigate and correct this before we discuss all the other questions.
After correction it is best to confirm the acoustic timing on the 2 sub measurements are working properly also.

 

[Foly83]

I have checked it again
The second deflection is a reflection
The mic is very close to the back wall, unfortunately there is no other way in this room.
New measurements with absorber at the back wall

But I really have to say. front left to sub left no problem, but right and left side together I just do not get. I just do not know how to proceed
I entered it once and made measurement. Results was disastrous. For the reason I had also written.

 

[jtalden]

I have checked it again
The second deflection is a reflection
The mic is very close to the back wall, unfortunately there is no other way in this room.
New measurements with absorber at the back wall

Okay - good. The measurements looks normal now. [REW picked up the louder reflection of the timing sound instead of the direct sound in one case of Post-12.]

But I really have to say. front left to sub left no problem, but right and left side together I just do not get. I just do not know how to proceed
I entered it once and made measurement. Results was disastrous. For the reason I had also written.

You added delays to the R sub and to the L main. That destroyed the good timing between them. We only are summing them at this time to get the combined outputs. You may also summed them with the FDW filter active. The summations need to made with no FDW active.
Here are my summation results. See if you can get these results.

 

[Foly83]

Did I understand that correctly, with the summing you just have to turn off the FDW?

I get a similar result, but I have to push on the delay. Something I do wrong or forget something.

I have created Aligned sum from front left and right
the same with sub left and right
then juxtaposed them in TA. But I do not get the same result
Too bad I don't have a video available.

Have saved the two test with

I correct myself.
I have now checked different data
If I set Window Ref time to 0 for FDW and disable FDW for all measurements, I get almost the same results without changing the delay to TA.
I call the new result (correction).

I had always assumed that I have to align the phases before I can put left & right together.
The same with the sub left & right. Since at time Ref measurements the subwoofers always showed a different distance, also at each subsequent measurement.
What about the time difference of both front speakers and both subwoofers?
Does the Time Ref automatically compensate for this?
But where do I get the delay data from, which I have to enter into my Mini DSP 2x4HD?
dealy everything in TA is now set to 0
I'm probably already asking questions that will only come in the next steps. Sorry if that is so

 

[jtalden]

Did I understand that correctly, with the summing you just have to turn off the FDW?

Yes, FDW must be off for summation with this process. When FDW is on, the 2 filtered impulses are summed. We don't want that. We want the 2 unfiltered impulses to be summed so that the entire room response is included. We are only using the FDW temporarily to remove much of the room influence so that it is easier to see the direct sound phase response. That helps identify the most favorable phase tracking starting point. It is the starting point for determining the final delay setting. We then turn off FDW to see the resulting SPL response. Then we can change the delay small amounts to see if the SPL is significantly improve with a slightly different delay setting. I try not to change the delay more than ±1/4 wavelength from the starting point. For a 80Hz XO a ±1/4 wavelength is about ±3.1 ms. I have occasionally adjusted up to 1/2 wavelength or even a little more in some cases. Unless the SPL is significantly improved just use the starting point as the final delay.

The starting point for the sub1 to sub2 delay is zero in this case, I just checked to see if SPL (without FDW) is improved with a delay change from that point. I found a negligible improvement, i.e., not enough to be significant.

I correct myself.
I have now checked different data
If I set Window Ref time to 0 for FDW and disable FDW for all measurements, I get almost the same results without changing the delay to TA.
I call the new result (correction).

Good!

I had always assumed that I have to align the phases before I can put left & right together.
The same with the sub left & right. Since at time Ref measurements the subwoofers always showed a different distance, also at each subsequent measurement.

The left vs right and sub1 vs sub2 are perfectly aligned now (as measured) without needing any delay change. Look at the overlay impulse chart. Look a the overlay phase chart with FDW set to 2 cycles and set 1/48 smoothing. It does not get better than this for direct sound phase tracking. Don't look at the notes in the REW measurement. Those notes are not relevant to this particular process. They are maybe used for other setup procedures.

What about the time difference of both front speakers and both subwoofers?

A time difference between the 2 subs or between the 2 mains? There is none. See above.

Does the Time Ref automatically compensate for this?

No, normally in this process the reference time setting does not need to be adjusted. REW will show different values on different measurements, but that does not seem impact this process. I am just in the habit of setting that to 0 ms due to an older process. I don't think it applies here.

But where do I get the delay data from, which I have to enter into my Mini DSP 2x4HD?
dealy everything in TA is now set to 0

The MiniDSP is normally used when it is necessary to change the delay between 2 or more subs to account for their different distances from the LP. You don't have that situation. Many people also is use it to add additional PEQ filters for the subs either before or after an AVR's automated EQ process.

For the next step of XO delay;
Note all the distance settings in the Marantz that were active for this testing in case we need them.
If you now intend to use Audyssey in the Marantz for distances and PEQ, I would have told you run that first as it will set distances (delays) automatically. We could just then measure to confirm they are correct. We could still readjust the SW XO delay manually if needed. I should have asked about your overall process plan initially before I started answering your questions about manual setup procedures. I am normally working on full manual setups or multi sub setup settings. There is not much to do if Audyssey is run unless the sub delay is set wrong or additional PEQ to the sub is needed. What is the overall plan here?

 

[Foly83]

Hello
I have a separate room in which is a complete home theater, as in the info on my person.
Which cost a lot of money to set and build. Over time I have also started to make measurements.
I would like very much to be able to do it myself someday.
The system around what it is here stands in our second living room.
mini dsp 2x4HD, B&W M1 for left and right, 2x Cambridge subwoofer Minx X201, Thomann Tamp 2 channel.
The whole thing is connected to the Tv via Toslik.
In German-speaking countries there is very little that you can find how it all works well.
I have already tried many ways to connect the front to the sub. But the only thing that has worked so far was this kind of connection that I found here in this forum. But unfortunately only left speaker and left subwoofer. Both have not worked so far.
That's why I contacted you. I can imagine that it is a little annoying when I ask stupid questions. But with us there is hardly someone who wants to explain it or can.
My goal is to be able to adjust my system well by means of this measuring method. So that I can also change everything. Or help friends, like you help a complete stranger. I would like to understand it.

Previously I am on the way to work with the phase on this site audiosciencereview . c o m under (Manually time-aligning subwoofer(s) to mains - how to). What then made me search at AVnirvana
This worked well, but as already written. Only with a subwoofer and a front speaker. As soon as I set up several, the GD became very bad, also the phase

So in short there is no receiver only the mini dsp 2x4HD and the components as already written.

 

[jtalden]

Okay, good info. Thank you.
I can use the existing 4 measurements and summations in 'ja14a - Xo80.mdat' to continue explaining the final steps needed to determine the XO delay needed. Will that fulfill your needs?
Do you first need answers or clarification on what we have done so far? I think there were a few questions above that I did not answer yet so if you are not clear on the process thus far let's address those questions before we do the last series of steps. I think I already mentioned the needed steps to complete the process, but I will add detail and post some charts to illustrate the rest of the process.

Just some notes on the process:
We are summing the subs as they will both work together a mono subGroup. That is, they both work to provide the bass for the left channel and the right channel.
We really wanted an average of the left and right channels and could have done that directly using REW 'trace arithmetic' rather than the REW AT summation. To adjust the L+R summation back to average, we can just reduce the L+R summation by 6 dB to get the average. We want the average to make it easier to find a compromise delay setting. If we found the delay needed for left and right independently then we would have 2 different delay values and we then need to discover the best compromise delay. Using the average of the left and right instead of doing it individually provides the best compromise delay in one step. The individual impact on left and right channels with that delay can still be calculated in REW and then measurements can be made with the new delay setting to confirm it matches the calculation.

 

[Foly83]

Hello jtalden

QUOTE]I can use the existing 4 measurements and summations in 'ja14a - Xo80.mdat' to continue explaining the final steps needed to determine the XO delay needed. Will that fulfill your needs? [/QUOTE]
Yes I would be happy if you explain the steps in more detail.
You can also do it by video, if that is easier and faster for you.
Possibly I understand it so much better with a fluid process.
Can give you via PN one of my email for video if not desired here.

To my questions that I still ask myself .
I create times points that interest me
1.
I was told when my cinema was set that deviations are bad even if so small. Because timing between front and subwoofer is everything. He explained it this way. ("If everything doesn't play exactly on the dot, you lose a lot of dynamics. And dynamic = perceived pressure) Therefore, I wonder if I / we should not still adjust the front two subwoofers, even if it is minimal? So that the wavefront of both subwoofers fit perfectly.
2.
As written above, if I do a Time Ref on front left and right, I get a good answer where the T=0 is. With the subwoofer I noticed that the delay between T=0 and sub is always minimally different. No matter how many measurements I make it is always a different delay.
3.
Original question:
(Why does every measurement have to be made from 20-20,000 when I separate the sub with Xo80 anyway. I have compared measurements from 20-300 and 20-20,000 and it looks identical to me.
That's why I wonder if you can believe this ref point at TA at all. Or is it related to the range of the measurement and by measuring up to 20k instead of 300 it is easier to determine. Because low tones are more difficult to estimate in time.
4.
I have now understood it so that we put the front left and right together to a whole.
So that both play at the same time and amplify in sum around 6db (in the best case). Wouldn't it make sense then to process the left and right speaker beforehand with Eq's, so that there are no big differences between the frequencies of both speakers? The same also with subwoofer left and right via TA. (Just a question for understanding, as it would prolong everything here unnecessarily).
So that's all the questions I could think of right now
Once we go through those, I think we can start with the details
Looking forward to it.

If you have explained everything, I'll try it step by step.
And upload it incl. pictures etc.

 

[jtalden]

1.
I was told when my cinema was set that deviations are bad even if so small. Because timing between front and subwoofer is everything. He explained it this way. ("If everything doesn't play exactly on the dot, you lose a lot of dynamics. And dynamic = perceived pressure) Therefore, I wonder if I / we should not still adjust the front two subwoofers, even if it is minimal? So that the wavefront of both subwoofers fit perfectly.

The main objective for subs/L+R XO range:

• Obtain a smooth SPL response to follow the chosen house curve.

Additional concerns:

• The final delay timing chosen should not deviate too far from the delay timing that would be chosen if the room influence was removed, i.e, close phase tracking of the direct sound through the XO range.
• Minimize the amount of PEQ required.
• Follow PEQ best practices.

Regarding the sub to sub timing:
There is nothing wrong with closely aligning the 2 subs in timing. That is exactly what we did here. They are perfectly time aligned with a 0 ms delay adjustment. Please tell me why or show me a chart to explain why you think they are not! Where are you getting that idea from?

That said, there nothing wrong with choosing an offset delay for the 2 subs so long as the main objective and the additional concerns just stated above for XO range is equally or better served. A offset timing of the 2 subs can sometimes enhance the main objective in a problematic room.

2.
As written above, if I do a Time Ref on front left and right, I get a good answer where the T=0 is. With the subwoofer I noticed that the delay between T=0 and sub is always minimally different. No matter how many measurements I make it is always a different delay.

Are you looking at the measurement notes instead of the various charts. I already said that the measurement notes do not apply to this process. How may degrees of phase difference is this variation at 80 Hz? A 1 ms variance in timing is about ±14° at 80 Hz. This is not enough to have a significant impact on the main objective.

3.
Original question:
(Why does every measurement have to be made from 20-20,000 when I separate the sub with Xo80 anyway. I have compared measurements from 20-300 and 20-20,000 and it looks identical to me.
That's why I wonder if you can believe this ref point at TA at all. Or is it related to the range of the measurement and by measuring up to 20k instead of 300 it is easier to determine. Because low tones are more difficult to estimate in time.

My thought process is the reverse. Why do I need to keep change the measurement range when when measuring a; woofer, midrange, or tweeter?
There is no reason that I can think of. I think in some trace arithmetic operations the accuracy or resulting range my be limited. I think full range measurements may provide slightly better REW estimates of the timing in some case. So, I consider it easiest and safest to just measure full range - less buttons to push - less mistakes. This applies only to this process where the XOs are active and thus no damage to the drivers.

4.
I have now understood it so that we put the front left and right together to a whole.
So that both play at the same time and amplify in sum around 6db (in the best case). Wouldn't it make sense then to process the left and right speaker beforehand with Eq's, so that there are no big differences between the frequencies of both speakers? The same also with subwoofer left and right via TA. (Just a question for understanding, as it would prolong everything here unnecessarily).
So that's all the questions I could think of right now
Once we go through those, I think we can start with the details
Looking forward to it.

Trace arithmetic: It's exactly -6 dB in all cases of converting 2 vector summed measurements into the vector average of the 2 measurements.

EQ is another topic with lots of different approaches. We have not discussed this. Regarding this XO setup process; yes, PEQ can be helpful in clarifying the response and assuring the choices we make in the XO setup are helpful to the overall house curve. If this case was my setup, I definitely would have done at least an preliminary EQ to get a much better idea where the house curve needs to be set. PEQ can often be effectively done after the XO is setup as well. Some advanced DIY Hobbyists EQ the drivers to a acoustic LR4 XO filter shape. This is a popular DIY approach on the 'rePhase' thread at diyAudio.com. That is an idealized approach that is probably not critical in terms of sound quality. Possibly that is what the consultant was advocating. I would be very surprised however it they can afford to take the time needed to create that condition. Some of the advanced automated setup processes are very good at doing this so far as I can tell.

Whatever your choices, good PEQ practices should be followed. Prominent experts stress is that no EQ is used above 500 Hz (some say 200 Hz or 300 Hz) unless it is equally applied to both speakers to shape the direct sound to the house curve or address driver resonance. Below 200 or 300 individual channel PEQ is done address the room response.

If you have explained everything, I'll try it step by step.
And upload it incl. pictures etc.

Good - the practice will help you. Details may not have been mentioned yet. I will also start to do the XO analysis and detail the steps.

 

[Foly83]

Thank you for the detailed answer to my questions.

(There is nothing wrong with closely aligning the 2 subs in timing. That is exactly what we did here. They are perfectly time aligned with a 0 ms delay adjustment. Please tell me why or show me a chart to explain why you think they are not! Where are you getting that idea from?)

When my cinema was measured. Two different delays had to be set for the two front subwoofers on the left and right, although the room is built completely symmetrically. Subwoofers are all identical. Therefore, I had asked again, since my room in which the system is what we write about on the right side is open. That's why I suspected that the right, although he removed the same must be set differently.


(Details may not have been mentioned yet)

That's what I meant when you explained all the other details. Then I will try to reconstruct everything myself. I would be happy if you could also explain the subwoofer. (Assuming the subwoofers would not be placed the same) There are many rooms where these can not be placed symmetrically.

And really many thanks again that you take so much time

 

[Foly83]

So I still send four pictures today before I go to sleep.
I have times your two SUM A/B calculated, even if I do not have the last details yet.Since I am in the forum still stumbled on some information.(REW Alignment tool - Guides or Manual?)I hope I understood your explanation there correctly. I came to this result, what do you think of it? For me it also looks good at 80Hz in the phase and GD area, also the summing at Xo view good. Since the LS are not yet Eq I ignore the other deflections below and above.
Delay information etc. are visible in the picture on the bottom left.

Well then, until tomorrow / later
have a nice day

Aligned Sum Front L&R and Sub L&R

 

[jtalden]

When my cinema was measured. Two different delays had to be set for the two front subwoofers on the left and right, although the room is built completely symmetrically. Subwoofers are all identical. Therefore, I had asked again, since my room in which the system is what we write about on the right side is open. That's why I suspected that the right, although he removed the same must be set differently.

I think you understand now, but I will say it all again a slightly different way is case it is helpful.

The initial starting delay for 2 subs regardless of their distances is the delay needed bring them into time alignment of the direct sound. This is found by analysis of; the Phase, Impulse or Step overlay charts when using a heavy FDW filter. Optionally we can just measure the difference in distance of the 2 subs from the LP and adjust the delay accordingly. If sub1 is 4 m from LP and sub2 is 6 m from LP the difference in distance is 2 m. Therefore, sub1 must be delayed by 2 m or about 5.8 ms delay. It does not matter if there are openings, or reflections or other room conditions as the direct sound is not impacted by them. Their impact arrives later.

So, sub1 = 5.8 ms is the starting point for the delay analysis. Now the FDW is then removed so the resulting SPL at the LP can be seen. This includes the full impact of room openings, walls, etc. We then adjust that sub1 delay a little as needed to improve the SPL through the XO range. We should stay as close to the starting delay as possible and avoid changing the delay more than about 1/2 wavelength of the XO frequency. If no significant improvement is found then use the 5.8 ms delay setting. If for example, we find a significantly better SPL with a setting of 7 ms we would just that as the final setting instead of the 5.8 ms.

Regarding the cinema setup, the consultant may have used this process, but probably did not. Note that the second paragraph leads us to an adjusted delay time on sub1 so that final delay adjustment for SPL may have result in the same relative sub delay that the consultant found. There are many different methods that can lead to to favorable results however. This method is more complicated than most, but has the advantage that we can often see in the charts the reason our settings are favorable or unfavorable. It can provide confidence that there are no other settings that will provide a technical advantage. It also often helps to disclose the reasons for XO range SPL issues and that helps identify ideas for remedial action to; the room, the XO frequency, or to position changes of the subs or LP. This info is helpful in deciding what type of corrective action that is more likely to be beneficial. Evaluating random changes to these items would likely to take longer.

 

[jtalden]

I came to this result, what do you think of it? For me it also looks good at 80Hz in the phase and GD area, also the summing at Xo view good. Since the LS are not yet Eq I ignore the other deflections below and above.

Yes, It looks favorable. My comments/thoughts on the charts.
Chart 1 - Alignment tool.

• This looks like FDW was was probably still active. If so, it is the intermediate step to find the initial or starting delay setting.
• You found a favorable delay. Phase crosses near the XO frequency and SPL support across the XO range is very good.
• Notice that the subs vs mains phase diverge from each other a little more than necessary after crossing. Considering this chart, I would expect to find closer phase tracking with normal polarity selected on the subs. If so, that would have identified a better starting point for delay.
• This crossing phase setting does reduce the group delay by 1/2 wavelength so it is near one of the popular favorable delay settings.
• The only question is; will the SPL still hold up favorably or will some additional adjustment be needed when the FDW is removed.

Chart 2 - Predicted SPL

• FDW still active so limited value for SPL evaluation.
• The same info is present on Chart 1.
• This chart adds value when showing comparison to other optional delay settings.

Chart 3 - Phase

• Phase charts should always be 1/48 smoothing otherwise there is too much shape distortion in them to interpret correctly.
• This would have been more informative without the FDW and Psy smoothing. It would have looked very messy, but would have shown the phase problems created in the XO range due to room effects.
• Can use this chart to compare the impact of other delay settings.

Chart 4 - GD

• There are offsets here that make this impossible for me to interpret.
• Scaled and filtered correctly (heavy FDW an 1/48 smoothing) the chart would show the impact of the phase rotation on the GD.
• The GD handoff of the subs to the mains in range of the XO range is primary interest.

I will try to do my analysis of the data tomorrow.