REW Alignment tool - Guides or Manual?

[jtalden]

It's on the measurement dialog. Use REW 'help' and find or search for 'Making a Measurement'.

 

[Harrycr]

I have measured sub 1 with a delay of 6 meters and measured the same sub with delay of 7 meters (altered in AVR) and done the "All SPL" then "Alignment Tool" and the phase trace is exactly the same, over each other is this supposed to be like this?

Thank you

 

[JStewart]

I have measured sub 1 with a delay of 6 meters and measured the same sub with delay of 7 meters (altered in AVR) and done the "All SPL" then "Alignment Tool" and the phase trace is exactly the same, over each other is this supposed to be like this?

No. Not when measured with a timing reference and there are no user errors in the measurement process.

 

[Harrycr]

With phase aligning sub front & sub back do I measure both of them full range (10-120hz) or with xover active in the AVR (10-80hz)?
Also do I use the manual sliders for the alignment tool as the align at cursor doesn't do anything?

Thank you

 

[jtalden]

With phase aligning sub front & sub back do I measure both of them full range (10-120hz) or with xover active in the AVR (10-80hz)?

Step 1 (aligning sub to sub timing) could be done either way, but since Step 2 (SWs to mains timing) should have the AVR XO active. I recommend the AVR XO to be active for all measurements. That way the same SW measurements can be used for both Step 1 and Step 2 and the AVR XO remains active for all measurements.

Also do I use the manual sliders for the alignment tool as the align at cursor doesn't do anything?

For Step 1 the 'align phase at cursor' is not used as there is no XO. We are just using the slider to find favorable timing to create strong and smooth SPL support across the whole SW frequency range.

If 'align phase at cursor' didn't work in Step 2 then please post the mdat file containing only the 2 measurements being timed (L+R and SW1+SW2) when 'align phase at cursor' did not work. What frequency was the cursor placed at? The manual slider is used to determine if there are other more favorable timing settings than the one that 'align phase at cursor' found. When you understand the process well it is helpful, but not necessary, to use 'align phase at cursor' to find one of the favorable timings.

 

[jtalden]

I should have mentioned that 'align phase slopes at cursor' may be helpful for Step 1 timing adjustment, but I haven't needed to use it for that in my setup.

 

[AustinJerry]

Measurements:

1. FL (alone; SW off)
2. FR (alone; SW off)
3. SW (alone; Measure the redirected bass by using FR channel with the FR speaker disconnected)

Given these 3 measurements, I can outline the REW alignment tool steps needed in this case to evaluate the current SW distance settings vs other options for this current setup.

I am arriving at this discussion quite late, but I have an interest in understanding how to properly use the Alignment Tool. I want to start reading this thread having already run the necessary REW measurements, and would like clarification regarding how to properly configure the measurements.

- When measuring the FL and FR with SW off, should the speakers be configured to run as Large Full Range, or Small with the usual crossover in place?
- Should room correction (in my case Dirac Live) be off or on when measuring?

Thanks in advance for your feedback.

 

[jtalden]

All measurements should be as the intended usage. In most cases that would be with the XO chosen and active and with mains set to small.
It is usually very helpful to have EQ active. Any EQ in the XO range may not still be appropriate after timing changes however. This depends on how they were determined.

 

[AustinJerry]

All measurements should be as the intended usage. I most cases that would be with the XO chosen amd active and mains set to small.
It is usually very helpful to have EQ active. Any EQ in the XO range may not still be appropriate after timing changes however. This depends on how the were determined.

Thanks, that makes it clear. Dirac measures and calculates its filters for each speaker Independent of other speakers, as full range speakers without crossovers in place. I suspect timing changes will not affect the filters.

 

[AustinJerry]

@jtalden

OK, I ran the measurements:
- LCR set to small, crossover @100Hz, subs off, Dirac Live on
- Subs only from center channel output with speaker disconnected

Here is the starting point:

Placing the cursor on 100Hz and clicking Align Phase at Cursor produces this:

Note that the suggested delay is -1.09ms. My equipment (MiniDSP) does not allow a negative delay. Why is the delay negative? When I adjust the sub channel delay manually (the sub distance tweak), a delay of 8ms produced the best results So, I entered 8ms manually, and the result looks quite close to the -1.09ms result:

Am I doing something wrong? A negative delay won't work for me. Thanks for any advice you might have.

 

[jtalden]

Nothing wrong - excellent results. These are 3 favorable alignments. As usual, a 1.09 ms timing change from ideal is negligible for a 100 Hz XO. The SPL difference in these first 2 charts look almost identical. There is probably a small fraction of a dB difference if we looked close enough.
The 2nd chart is very close to the 'conventional' timing. The phase tracking through the XO range is pretty close from 85-180 Hz. The phase separation below 85 Hz can probably not be improved with just a timing change. It would likely require changes to the XO filters and/or the room setup. You could try inverting the SWs There is an outside chance that there may be better phase tracking results, but the SPL would not be any better. Note that the SPL of the 3rd chart is also very strong through the middle of the XO range. There is a modest crossing of phase a 100 Hz as the SWs are arriving 1 cycle later than the center speaker. Note that there is only a trivial SPL difference at the very extreme ends of the XO range where the phase starts to diverge.

I suggest the timing for the SWs be first chosen based on the FL+FR speakers. Then if the CC is a different design than the FL and FR, I normally suggest the the delay/distance of the CC be adjusted such that there is close phase tracking to the FL and FR through the midrange. I don't know if others agree with this. My thought is that there is more likely to be shared content across the 3 front speakers in the midrange and we tend to be sensitive to differences in the midrange. Low bass content in the CC is likely less common?

 

[AustinJerry]

Thank you for your very useful feedback. While I showed the results for the center channel, the results for the left and right channel were almost identical. I still don’t understand the meaning of a negative delay. Why would the tool give that result?

 

[jtalden]

I'm not sure what you are missing. The tool helps find the favorable timings. Those can either be negative or positive delays for SWs.

If the SWs are placed further away from the LP than the mains then with no delays to either the mains or the SWs then sound will arrive from the SWs later than sound from the mains. The tool will then suggest the SW delay be reduced to provide same arrival time. A negative delay will be shown in the tool.

DSP in the MiniDSP and in some SWs amps will add maybe 1 ms of signal delay in addition to any distance difference so that is another reason the SWs signal is can be delayed too much even if the distances were carefully chosen to equal in the setup.

If an AVR (or JRiver) is being used, the SWs distance setting in the AVR can be increase to reduce the delay of sound from the SWs. Alternatively, the distance to the mains can reduced to reduce the delay on the SWs. The AVR adjusts the relative delays based on the distance differences entered between the speakers; not on the absolute distances entered.

Possibly my suggestion of always adjusting the lower frequency speaker or driver when using the tool is the point of confusion. If so, that is just a suggested practice as a practical matter, It is not a necessity. No matter want we adjust in the tool the actual timing changes in the DSP or distance changes in an AVR are chosen as needed to implement the change.

 

[AustinJerry]

I don't know why I am having so much difficulty understanding the results. I have never implemented a negative delay for a SW, and TBH I don't really know how to. None of my devices allow a negative value in the delay setting.

Another anomaly: for Align Phase at Cursor, the tool recommends a a -1.09 for Center/sub, -1.18 for Left/sub, and then a +10.75 for Right/sub. That is a pretty large difference for the right channel.

In case you have time to validate my findings, here is a downloadable copy of the MDAT file: https://www.dropbox.com/s/af2r3mh1e8tnr2q/Sweeps%20Dirac%20On.mdat?dl=1

Thanks for your advice--it is appreciated.

 

[jtalden]

I will look at the file later today.

A recommended decrease in delay of the SW is the same as an increase in delay of all the mains. Most of us have a method to accomplish that even when the SW is already at 0 ms in a MiniDSP. We instead either; increase the distance setting of the SW in the AVR, or reduce the distance settings of all the mains in the AVR. If the AVR doesn't allow that for some reason or we are not using JRiver or have any other DSP method then the only option is to either physically move the SW closer to the LP or move all the mains further away.

In this case, the timing is very close to ideal with the original setting and no change is needed. The 8 ms delay also provides a very favorable SPL result so that can also be used.

 

[jtalden]

I reviewed the file. All is as discussed in Post 135 - 136.

The only new item is the question you posed:

Another anomaly: for Align Phase at Cursor, the tool recommends a a -1.09 for Center/sub, -1.18 for Left/sub, and then a +10.75 for Right/sub. That is a pretty large difference for the right channel.

I confirmed the same +10 to +11 ms SW delay recommendations (depending upon cursor position) when considering the FR main speaker. It is clear however when looking at the phase response the phase tracking is much better near the ~ -1.0 ms SW delay the tool recommended for the FL and CC. I am not sure why the tool found this value. My first thought is that the truncated sweep length of the SW (ending at only 300 Hz) may have contributed to this result. The tool may use a wider frequency range for proper calculation. JohnM would need to need advise.

Because the room effects are significant when measuring at the LP the phase response is often pretty chaotic and thus hard to interpret. This is why I normally apply a FDW filter to the data initially to better get an idea of the direct sound phase. When doing that on this data the tool finds the same ~ -1 ms SW delay for all 3 front speakers. I then removes the FDW to determine if a reasonable shift in the delay is helpful to account for the room influence on SPL/phase. I normally try to keep any additional shift in the delay as small as possible. I try to deviate within a ±1/4 wavelength, but accept larger values if the SPL is significantly better.

The room impacts the FL and FR differently in most cases so a compromise SW delay is often needed. It is thus easier to sum FL and FR and use FL+FR vs the SW to more quickly find a good compromise delay setting.

FL with no FDW: The tool finds -1.09 ms as the timing for good phase tracking.

FR with no FDW: The tool finds 10.42 ms timing, but more phase crossing is evident.

FR with FDW applied - The tool finds -0.67 ms as the recommended setting

FR no FDW applied - The tool charts shows the -0.67 ms setting as having better SPL and phase tracking

 

[Harrycr]

I have tried to align two subs as one.
Subs front and back (diagonally).
The back one polarity is inverted.
When phase aligning I get front sub 4.35 meters and back one 0 meters.

I would have thought back one would have the delay?

Is it best to measure with everything set to + polarity first then phase align?

Thanks

 

[jtalden]

Yes, using the same polarity is the more likely to provide the more favorable results. There is no problem analyzing with one polarity reversed as well. Just chose a polarity and timing that provides strong and smooth SPL support throughout the entire bass range of the SWs. If there is more than one setting that looks favorable just chose either of them or audition the sound both ways.

If the closer SW has significantly more delay than expected based on the differences in distance to the LP, that suggest there may be a better delay setting. Two SW don't have to be exactly time aligned based on the difference in distance, but there is a concern if there is a big discrepancy,

It's fine to use the 'align phase slope at the cursor' if we chose cursor position that avoids room effects in both SWs it is very likely to accurate. We should also manually confirm that there is no better setting by moving the delay slider to different settings.

 

[Harrycr]

Thanks again.
The furthest sub has all the delay while the closet sub has no delay at all. The closest one has the reversed polarity.
Furthest sub is inside the front left main speaker and the closest is diagonally inside the rear right speaker.

Hope that makes sense.

 

[jtalden]

You can post the mdat if you want me to review it.

 

[Harrycr]

Thanks currently the old laptop is not on line due to age. I'm waiting on a new one that hasn't made it over here yet, probably due to covid. All these posts are done on phone.
The receiver is a Yamaha Z9 awesome but has no HDMi.
I posted a thread no one answered but when I use the acoustic timing I setup it up in preferences and use one input into the Yamaha, I measure and it sends a signal to the main speaker and sub and I end up with a sub and main measurement crossover 80hz.

Thanks

 

[jtalden]

What is it about Post-118 that you question. There are 2 outputs of REW so 2 inputs to the Z9.

 

[Harrycr]

Ok think I get it.
I need to disconnect the right main speaker (measurement speaker) for the sub measurement.
This can be tricky in my situation.
I suppose there is no other way.

 

[jtalden]

Yes, we can disconnect the FR main speaker wire at either end.
The reference signal is sent to the connected FL speaker.

 

[Harrycr]

Thanks.
It's hard to disconnect the speakers.
I might be able to get my hands on a sound card, would this help.
For loopback on a soundcard do you know what the connection is e.g TS or RCA or XLR balanced or unbalanced?
Everything seems to be on USB mic setup.

Thanks again really appreciate your help.