Odd REW measurements

[sir sanders zingmore]

I may be doing something wrong but I measured my room (at the listening position) and got the following. 

The purple and reddish lines are L and R as measured individually

The Green line is L+R measured together.

I don't understand why the combined response is lower in the bass region

 

I've triple checked and I'm pretty sure everything is wired up in phase correctly

 

 

But just for fun, I switched the polarity of the cables on one speaker and I get this. 

Now the combined L+R is louder in the bass region than the individual 

 

 

 

As I said, I'm sure everything was wired correctly for the first measurements. 

Can anyone shed some light on the reason(s) for what I'm seeing?

[Ittaku]

One of your bass panels was wired backwards in the factory?

[Hytram]

I am going for weird room acoustics.. 

 

But I really don't know. 

[sir sanders zingmore]

12 minutes ago, Hytram said:

I am going for weird room acoustics.. 

 

But I really don't know. 

 

I was wondering if that was possible, but how?

I've recently moved my setup into a different room which is actually a much more 'normal' shape than my previous room. Everything is symmetrical. Speakers both the same distance from walls and from the listening position

 

 

 

 

 

 

[Tony M]

I think the measurements are telling you there is a phase error somewhere.  Does listening confirm this?  I guess you've rechecked the wiring many times.

 

How did the system sound in the previous room?

 

What speakers do you have now, Trevor?  That peak at 2K suggests you still have the AG's - I've modified mine to eliminate that and been a happy camper since.

Edited December 27, 2021 by Tony M

[frednork]

1 hour ago, Ittaku said:

One of your bass panels was wired backwards in the factory?

what he said

[sir sanders zingmore]

Not sure whether this helps, but this is the phase plot of the original chart (the one with the lower summed output).

 

 

[sir sanders zingmore]

3 hours ago, Tony M said:

What speakers do you have now, Trevor?  That peak at 2K suggests you still have the AG's - I've modified mine to eliminate that and been a happy camper since.

 

Hi Tony, the AG's have been sold (I do miss them, wish I could have justified keeping two sets of speakers)

I've gone back to the Sanders panels

[tripitaka]

4 hours ago, Ittaku said:

One of your bass panels was wired backwards in the factory?

+1

 

It's only an issue below about 180Hz, does that tally with the panel Xover Freq?

Or is there an introduced phase error in any Digital processing that you might be incorporating? 

 

Edited December 27, 2021 by tripitaka

[Ittaku]

35 minutes ago, tripitaka said:

+1

 

It's only an issue below about 180Hz, does that tally with the panel Xover Freq?

Yes, and not only that, you can see when he does try measuring them out of phase, there's cancellation and sound loss above 180Hz compared to the in-phase summation graphs. You can also see a new peak at 2k suggesting another crossover point with bad phase alignment again. Things are pretty screwy all round...

Edited December 27, 2021 by Ittaku

[sir sanders zingmore]

30 minutes ago, Ittaku said:

Yes, and not only that, you can see when he does try measuring them out of phase, there's cancellation and sound loss above 180Hz compared to the in-phase summation graphs. You can also see a new peak at 2k suggesting another crossover point with bad phase alignment again. Things are pretty screwy all round...

 

There's only one crossover (at 172 Hz)

When I put one speaker out of phase, I changed both the panel and the woofer inputs. So I guess that's evidence that the panel is correctly wired but the woofer may not be.

 

Edit: the only other thing I've changed in the new room is that I'm now vertical bi-amping ( ie  one amp per speaker) whereas I was horizontal bi-amping previously ( one amp for both panels and the other amp for both woofers) 

[frednork]

if you show the step responses it should be obvious but pretty compelling thus far

Edited December 27, 2021 by frednork

[sir sanders zingmore]

10 minutes ago, frednork said:

if you show the step responses it should be obvious

 

Initial setup (ie the one that corresponds to the bass apparently cancelling)

 

 

Setup where I deliberately inverted the speaker cables

 

 

[frednork]

ok, the intial graphs I saw had labelling that came up was confusing but the ones above are very clear. bass is wired out of phase on one. Now you need to figure out which one. I reckon right (red) if the colours match up

[frednork]

35 minutes ago, sir sanders zingmore said:

Edit: the only other thing I've changed in the new room is that I'm now vertical bi-amping ( ie  one amp per speaker) whereas I was horizontal bi-amping previously ( one amp for both panels and the other amp for both woofers) 

its possible the amp for one of the bass panels is out of phase as an alternative

[sir sanders zingmore]

7 minutes ago, frednork said:

ok, the intial graphs I saw had labelling that came up was confusing but the ones above are very clear. bass is wired out of phase on one. Now you need to figure out which one. I reckon right (red) if the colours match up

 

Admittedly I don't really know how to interpret the impulse graph, but it looks to me that the first graph has the initial impulse (at time zero) in the same direction and then after about 10ms they seem to go in opposite directions 

 

The second one (in which I switched the black and red inputs to one speaker), the initial impulse seems to be in opposite directions and then things line up as time goes on

 

So wouldn't that mean the second one is out of phase (as the initial impulses are in opposite directions)?

[Ittaku]

2 minutes ago, sir sanders zingmore said:

 

Admittedly I don't really know how to interpret the impulse graph, but it looks to me that the first graph has the initial impulse (at time zero) in the same direction and then after about 10ms they seem to go in opposite directions 

 

The second one (in which I switched the black and red inputs to one speaker), the initial impulse seems to be in opposite directions and then things line up as time goes on

 

So wouldn't that mean the second one is out of phase (as the initial impulses are in opposite directions)?

In your second impulse, your main panels are now out of phase, but the bass panels are in phase. In the first one it's the other way around.

Edited December 27, 2021 by Ittaku

[frednork]

the initial peak is high frequencies and as you go along the lower frequencies kick in until around usually 15ms or so in a normal domestic room. Everything after that is room reverberation.

[frednork]

identical speakers will give very similar step responses in the first 15ms or so

this info from here https://audiojudgement.com/loudspeaker-step-response-measurement/

 

Typical step response plots

Before I make the step response measurements, let’s analyze some charts that you will typically bump into.

This is a typical chart you will see on a 2-way speaker. There is an initial sharp positive peak in the step response, followed by an undershoot, and then, a larger, more slow rising, positive peak. The first peak is produced by the tweeter, and the second is produced by the mid-bass. Although the drivers are connected with the same polarity, and the speakers are placed flush on the same baffle, their acoustical centers are not aligned. The speakers are not time coherent, and the mid-bass lags the tweeter by a few fractions of a millisecond.

This 2-way loudspeaker step response is very common. Although there is a short delay between the two drivers, don’t be surprised if you will find such response on expensive loudspeakers. Achieving a perfect step response, might compromise the frequency response. As a result, most manufacturers tend to not shoot for perfection in this area.

 

Edited December 27, 2021 by frednork

[sir sanders zingmore]

7 minutes ago, Ittaku said:

In your second impulse, your main panels are now out of phase, but the bass panels are in phase. In the first one it's the other way around.

 

5 minutes ago, frednork said:

the initial peak is high frequencies and as you go along the lower frequencies kick in until around usually 15ms or so in a normal domestic room. Everything after that is room reverberation.

 

Ok that makes sense. But….. how can the room reverberations be so loud compared to the impulse ?

[sir sanders zingmore]

12 minutes ago, frednork said:

the initial peak is high frequencies and as you go along the lower frequencies kick in until around usually 15ms or so in a normal domestic room. Everything after that is room reverberation.

 

Actually, if I zoom in to show the first 18ms or so, the first impulse is like this:

Which looks like it's in phase (the lines seem to move in the same direction)

 

 

whereas the second one (with the speaker cables swapped) looks out of phase for the whole time

 

 

So could it be that the entire thing is caused by room reverberations rather than by any phase issue with the speakers?

[tripitaka]

15 minutes ago, sir sanders zingmore said:

 

Edit: I see what you mean from above, I'll await the brains trust 

Edited December 27, 2021 by tripitaka

[frednork]

11 minutes ago, sir sanders zingmore said:

So could it be that the entire thing is caused by room reverberations rather than by any phase issue with the speakers?

I doubt it but simple to check by inverting the right woofer cables and remeasuring

[sir sanders zingmore]

8 minutes ago, frednork said:

I doubt it but simple to check by inverting the right woofer cables and remeasuring

 

Do you mean invert only the right woofer cables but have everything else connected correctly

[frednork]

Just now, sir sanders zingmore said:

 

Do you mean invert only the right woofer cables but have everything else connected correctly

yep