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

[Foly83]

Thank`s for the information.
I will read it carefully after work and also look forward to their details

 

[jtalden]

1. Offset L+R SPL by -6 dB (add to data)
2. Offset sub1+sub2 by -15 db (add to data)
3. 

 

[jtalden]

XO range is about 50 to 125 Hz based on a 30 dB drop. We could choose 40-150 for a 35dB drop or any other range we chose. The SPL is not impacted significantly once the drop is 30 dB down so that is a practical choice. This is the range of the frequencies we are concerned with for adjusting the XO settings.

 

[jtalden]

Opening AT without the FDW there is a very clean phase trace through the entire XO range, The first room effect appears in the phase of the sub1+sub2 (incorrectly labeled L+R in the chart note) at around 100 Hz so we can adjust the delay to see if the best setting can be found without the aid of the FDW. We find very good phase tracking and SPL support with delay setting around -4.5 ms as shown. Because there are not major room effects in the XO range this represents a favorable setting. It is not really necessary to use the FDW.

Applying a 5 cycle FDW anyway see the impact we find that in is necessary to shift to about -2 ms delay to find the best tracking. Depending on FDW value chosen the best delay will differ a little as expected. It does clean up the general phase trace however to give us confidence that we are in the correct range.

So, The -4.5 ms value found above without FDW will provide the best SPL for mono sounds. I saved that new sub1+sub2 location and also the aligned sub. There still may be difference in sounds panned left or right. So we can investigate that next...

 

[Foly83]

I just read and construct it times in REW that I understand and see better


I just brought my daughter to the bed. Directly on the sofa and entered the setting in the 2x4HD. Bass comes very powerful.
Why do you have the front and the sub - regulated? why the subwoofer not simply on -21?

as soon as I have gone through everything I'll report directly

But I notice very much that the right subwoofer is much louder. Although it is equally loud in the mini DSP

 

[Foly83]

Can it be because the subwoofers are not yet eq`t? I have to make the right subwoofer 8db quieter, that it sounds the same.

What I wonder very much, I get the same result only if I enter a completely different opposite value ?
All without FDW

 

[jtalden]

Why do you have the front and the sub - regulated? why the subwoofer not simply on -21?

I moved L+R down 6 dB because we vector summed them instead of vector averaging them. It was to properly place the L and R traces on the chart . Thus if we use this same measured data in the next step to look at the L and R channel individual responses they will graph at the same level as the original measurements.

There is no change needed to the L and R gains in the DSP to implement this setup. Sub1+sub2 gains should be reduced by 15 dB. Note that the -15 dB change was just a very rough estimate of what is needed bring the bass down to a level that can be EQ'ed to a typical house curve. The actual reduction needed depends on many things. It was done to recognize that the bass was way too high to expect to be able to adjust it using PEQ alone. You may need to adjust sub gain again when applying PEQ.

But I notice very much that the right subwoofer is much louder. Although it is equally loud in the mini DSP

I don't know why that would have happened unless the -15 dB sub gain was changed in only the left SW. That would also explain why the bass still seemed so strong.

 

[jtalden]

What I wonder very much, I get the same result only if I enter a completely different opposite value ?

That XO time alignment in your screenshot is in fact the same just as you indicated. Your AT shows 9.57 sub delay and we previously found this same alignment at -4.5 sub delay. I notice you used copies of the L+R and subs for this AT analysis. That would not be a problem unless one or both of the copies had a change to the t=0 location. That would explain why a different delay value was found. You can use the overlay impulse chart of the original sums vs the copied sums as see if there is an offset. Post the file if this does not reveal the issue.

 

[Foly83]

Unfortunately I can not find the error
here the file

I must also unfortunately say that I do not understand why with FDW is evident that -2 is worse than without FDW with -4.5. I suspect a wrong translation. Because the written for me no sense

(I notice you used copies of the L+R and subs for this AT analysis. That would not be a problem unless one or both of the copies had a change to the t=0 location.)

Yes I did, how did you connect the measurements? Don't see it on the picture because of the FDW window. If your way makes less problems because of T=0 I would like to know it to cause less errors.

I don't see on the Impulse overlay where sub and front start unfortunately. But when I use your L&R & Sub1+Sub2 as reference it looks the same.

 

[jtalden]

I do not understand why with FDW is evident that -2 is worse than without FDW with -4.5. I suspect a wrong translation. Because the written for me no sense

It's because -4.5 results a slightly better phase tracking at the LP due to the impact of the room. The change is hardy noticeable with only a 2.5 ms difference. We can see the small improvement in the SPL chart, but it is small enough to be trivial in terms of PEQ settings and sound quality.

Yes I did, how did you connect the measurements? Don't see it on the picture because of the FDW window. If your way makes less problems because of T=0 I would like to know it to cause less errors.

We don't move t=0 on any subset of the original measurements (like the subs). The acoustic timing that we took care to apply is lost if we do that. [We can move the entire group of 4 measurements if we want because the relative timing is retained, but there is rarely a need to do that.]

This impulse chart you posted is wrong. The L+R is still at 0 ms where they were measured. Both sub impulses are shifted to the left (less delay) from the original measured sub impulse locations. That shift will result is an erroneous delay result when using the AT. It will result in reporting more positive delay change than needed.

When we use the AT, we find the delay change that is needed from our current delay settings. So, if the AT result is +4 ms for sub1+sub2 we add +4 ms into the current DSP setting for the subs. So, if the current delay on the subs is 2 ms the new setting will be 6 ms.

I am assuming in our case the DSP setting on the subs was 0 ms when we did the measurements. We found a the delay change needed of -4.5 ms using the AT (see post-29). We cannot set a 0ms - 4ms = -4ms setting into the DSP. Instead, we can increase the delays in the L and R channels by +4.5 ms (assuming they are also going through the DSP). That achieves the same relative change in timing between the subs vs the L+R. We can make that setting change and remeasure. The actual measurement will then match the AT prediction.

 

[jtalden]

Attached is my file as discussed in Post-29.

 

[Foly83]

Yes the current setting was at 0 on the DSP.
Did you find my error in the file why my result was different? I would like to know what I did wrong.

I'll set the DSP like this now and then do a real measurement as soon as I get around to it today. I am curious about the result.

Also very interesting. So if the sub channel is - I have to enter the delay for the front channel at the DSP.
If that is so I have always understood it wrong and entered






Thank you so much for having such a long stamina with me. It sure is hard to teach someone something that is so far below your level of knowledge

 

[jtalden]

Yes, I found the error. The 2 subs impulse locations were moved from the original position. that is, the t=0 location was changed in your file as shown in the chart you posted. This indicates you changed it somehow either, on purpose, or by accident. I thought you agreed you moved t=0 for the 2 subs. You said 'yes' but maybe that was not in response to my statement. I also explained why the measured position must not be changed when using the AT to determine how much delay change is needed. I also posted my file so you could see the steps I took to do the analysis. You should try to create a similar file starting with the original file of 4 measurements - not the file you modified.

Regarding your DSP settings:
I indicated the mains should remain at 0 ms delay and the subs would be set to -15 dB.

I assume you changed the delay on the mains by +4.5 ms as I suggested?

If this is not done the results will differ.

 

[Foly83]

Yes my change was unintentional.
I could not find the error. I will try to get the same result with your file. With us it is just 1 o'clock at night. Will try it then tomorrow.
I wish you still a nice day.
A small question still. I have understood the delay correctly. Have made both measurements.
The result shows me that I have to enter the delay on the subs in the DSP and thus the TA window looks like this if that is so

(We cannot set a 0ms - 4ms = -4ms setting into the DSP. Instead, we can increase the delays in the L and R channels by +4.5 ms (assuming they are also going through the DSP).

This sentence was written very difficult by the translator and it is confusing.
If you just confirm where I have to enter the delay I will remember it for the future. Picture of the TA delay for reference.
The language bariere makes it unfortunately somewhat difficult, translators are just you only machines.
I would enter it at the subwoofer for my understanding

 

[jtalden]

A small question still. I have understood the delay correctly. Have made both measurements.
The result shows me that I have to enter the delay on the subs in the DSP and thus the TA window looks like this if that is so

View attachment 42281

Okay, your DSP allows you to enter a negative delay! I have not seen that before. That makes it less confusing for you. [So long as the DSP sends the signal to the L and R channels 4.5 ms after it sends the signal to the 2 subs then we are good.]

(We cannot set a 0ms - 4ms = -4ms setting into the DSP. Instead, we can increase the delays in the L and R channels by +4.5 ms (assuming they are also going through the DSP).

This sentence was written very difficult by the translator and it is confusing.

Well, my formula error did not help your understanding. I intended the formula to read: 0ms + -4.5ms = -4.5ms. I was putting my statement in words into a generalized formula. In different words: Original DSP delay setting (0 ms), plus the AT change in delay (-4.5 ms) equals the new DSP delay setting (-4.5 ms). Your DSP allows this settings so the part about increasing the L and R channels instead does not apply.

If you just confirm where I have to enter the delay I will remember it for the future. Picture of the TA delay for reference.
The language bariere makes it unfortunately somewhat difficult, translators are just you only machines.
I would enter it at the subwoofer for my understanding

The AT delay change needed is applied to the DSP. It appears that is what you have tried to do.

I don't know the operating manual for your DSP so I cannot tell you how to enter the values. It is better to ask someone who knows that device. I explained what it must do the the input signal in bold type above.

Regarding 2 charts:
I cannot tell if they are correct or not as there is no comparison to the target. Post the file you feel is correct and also post your new measurements with the correct -4.5 ms delay in the DSP and will check them for accuracy.

 

[Foly83]

So, I took the basic file. (Xo80) From which we started at the beginning when the measurements were usable for you.

Was able to construct everything exactly as you showed me this time.

-4.5 without FDW

FDW on -2

Also found the error. In the last alignment there was still a time offset stored in the IR window. When I corrected this to 0 it was one to one like yours.


When I was done I superimposed your finished analysis. Except for the bass increase, everything is identical. I left the bass a little louder, so I consider that normal.

To be on the safe side I adjusted the volume to your preset and saved it via offset to data and didn't make a copy like last time to avoid further mistakes.
Here are at the SPL on top of each other and you do not recognize them at all (SPL -6 & -15 / Aligned Sum -4.5ms)

I can not enter - delays, it was a cutout from the TA. Sorry for the poor display

I have also made an echo measurement for safety, because in the translation and in English it is unfortunately called a so and once so. In the German translation it makes a " , " where it means exactly the opposite.
The dip in the transition from the SPL tells me that I am right.

I must admit that I can now reconstruct everything that you have shown me. But it is still very difficult for me to see the same in the phase what you see. But that is 100% your experience!

I am curious whether I have not built a mistake in my explanation without noticing it.

 

[jtalden]

Looks Good!

 

[Foly83]

Hi
Have tried a little different settings the last two days. Everything works quite well
Would still have a thousand questions, but that should be enough for now.
A few .mdat for practice would not be bad now

And I would like to say thank you for your patience

 

[jtalden]