REW Alignment tool - Guides or Manual?

[lexicon]

Tricky to get your head around this concept I agree.
Delays in ms work conversely to distance in metres or feet. 1ms delay does not quite equate to 1ft closer in distance (speed of sound related) but it’s close enough - 88.5ms = 100ft.
By decreasing sub distance the receiver will see the sub as closer to the other speakers.
As it is seen as closer, the receiver will apply MORE DELAY in ms to slow it down to give the other speakers more time to catch it up.
At least that’s how I understand it.

 

[jtalden]

I get confused all the time so it is good to question. In this case I think No. 9 statement is correct.
A decrease in the SW distance setting (relative to mains distance settings) in a AVR is the same as increasing the SWs delay in the DSP. In your particular case however we needed a reduction of SW delay in the DSP instead of an increase, so we needed to increase the SWs distance in your stereo preamp.

Because we sometime get it wrong we then remeasure after making the change to confirm the timing. The phase tracking should then be as predicted and the impulse overlay chart of the SW and mains impulses should show the relative positioning that we targeted.

 

[jtalden]

Cross post above. We are in agreement now.

Note that the REW tool indicates the change needed in both ms and distance for us so we don't need to manually do the conversion. REW does not indicate the direction of the distance change only the magnitude. It is bit confusing to me, but it's easy enough to just confirm any change with measurements.

 

[Harrycr]

Thanks
So we just enter whatever REW suggests into the AVR.

For time alignment of 2 x subs:
Match level of sub A to sub B
Time align of the 2 x subs
Eq as one (combined mono) with same filters entered into each sub.

With time alignment of the 2 x subs what speaker is used as the reference?

Thank you

 

[lexicon]

What equipment are you using? Doesn’t matter which speaker you use as the timing ref as long as you use the same one for all. I use 1.8.

 

[jtalden]

Thanks
So we just enter whatever REW suggests into the AVR.

Sort of...
We use the REW alignment tool to select favorable setting changes for; DSP delay timing (AVR distance settings), gain, and polarity for the 2 SWs. The tool allows us to save these settings as a measurement and records the changes applied in the measurement notes.

For time alignment of 2 x subs:
Match level of sub A to sub B

We are looking to provide a relatively smooth SPL response that will respond well to EQ. Matching levels at the LP is usually fine. Matching SW gain settings is also usually fine. It can sometime be beneficial to use other gain settings. Relative SW levels is often not a big factor. Often all that is needed is to adjust the timing. The room setup and SW capacity are significant factors so there may sometimes be benefits to adjusting gain settings. The REW tool allows us to evaluate the impact of timing, gain and polarity setting changes to chose a favorable combination.

Eq as one (combined mono) with same filters entered into each sub.

This is the conventional approach.

 

[Harrycr]

Thank you
To time align subs to work together as mono, one sub placed up front of the room and other behind the couch.
Do I use a speaker (a main) as a reference to align to each sub or use the other sub?
To time align doesn't the reference speaker have to have a tweeter?

Thanks again

 

[jtalden]

Use the same main speaker for both SW measurements.

 

[Harrycr]

Thanks again.
Is it best to time align before or after eq?

 

[jtalden]

Adjust and verify...

For timing of 2 SWs I would not expect any timing change would be needed as a result of mono EQ. We select the timing that results in the best phase relationship pre EQ. The EQ impacts the phase of both SWs in similar manner thus retaining the same relationship after EQ.

For timing of XOs, the 2 independent EQs can change the relative phase tracking across the XO range. I haven't seen a case where the timing was significantly changed in my limited DIY setups, but there are a wide variety of setups so it may be significant in some cases.

 

[Harrycr]

Thank you
For time alignment with subs to mains there is a "Time Align" in the control, can you please describe what is this for? Do I do this before the Alignment Tool.
There are others as well do we need to use any of these other than Alignment Tool?

Thank you

 

[jtalden]

The "Time align" function is not used when evaluating the timing of multiple SWs or XOs. It aligns the timing between all the measurements selected. The original relative timing is then lost. Time align is helpful to create a beamforming average measurement, that is one where multiple mic locations are first time aligned and then vector averaged. This has no usage for common system setup activities. REW help may have more info on these controls.

All the measurements needed for timing SWs or XOs need to use REW acoustic timing (or loopback timing) feature. This retains the current relative timing between all the measurements. The alignment tool is then used to determine if any timing changes are helpful.

 

[Harrycr]

Thanks again.
For the Offset LR-48 80hz XO, I got 8.79ms.
With FDW settings how do you get to these figures?

 

[jtalden]

I don't understand the question.
Using a short FDW is helpful to determine the best delay timing for the direct sound as the room influence is reduced. The FDW is then removed to confirm there is still favorable SPL support. If a more favorable SPL support is found with a slightly different delay setting we would normally use that setting instead. The idea is not to deviate too far from the setting found using the FDW. A favorable setting can usually be found within ±1/4 wavelength of the XO frequency, but I would not be concerned with any setting within -1 to +1/4 wavelength.
I suggest you review the exercises of Post-27 and the information in Post-36. You can also use the OP's measurements to see if you can find the same favorable SPL results. This method takes some practice.

 

[Harrycr]

My bad. With FDW how do you work out the settings eg "window ref 12.7" "FDW 5" ? Or that just what values should be.

Ok with
Real Data 6.50ms
Offset LR-48 80hz xover 8.77ms

I notice that around the xover phase trace they cross is this normal?

Do we align sub 1 & sub 2 to make sure they are in phase.? Like post 73.

Thanks

 

[jtalden]

My bad. With FDW how do you work out the settings eg "window ref 12.7" "FDW 5" ? Or that just what values should be.

The 'Window ref time' normally does not matter for this purpose. The value will shift due to the original measurements and subsequent REW operations. [In rare cases I have seen the phase plot to be impacted so if the phase plot looks questionable we can reset this to 0.000 ms for all measurements.]

When setting the FDW value, I usually start around 5 or 6 cycles (applied to all) and decrease from there as needed if the direct sound rotation is not easily identifiable for both traces.

I notice that around the xover phase trace they cross is this normal?

A crossing condition near the XO suggests that the SPL support will be favorable. This can be wrong though if this is heavily impacted by room effects and the overall phase tracking through the XO range is poor. It is best to look for both a crossing near the XO frequency and close phase tracking/alignment of the 2 phase traces. This will result is favorable SPL support.

With FDW on, the direct sound phase is much easier to see and align. The resulting delay change that results indicates the delay change needed for the alignment of the direct sound. Strong SPL support with FDW then disabled that is reasonably near this delay value is then a favorable setting.

Do we align sub 1 & sub 2 to make sure they are in phase.? Like post 73.

Yes, start there and often finish there. In some case better SPL smoothness through the range will be found using some delay offset from this value. With multiple subs it may sometimes be helpful to adjust the relative gain or even to invert the polarity of one or more SWs.

A smooth SPL response that is amenable to EQ correction to our house curve is the overriding goal.

 

[Harrycr]

Thank you, very much appreciated.
Please keep in mind I'm not questioning you just getting a bit confused.
Align x 2 subs, but thought I had to have full range (with a tweeter) as a reference for time alignment?
I thought too time align x 2 subs with mains, the idea was to align each sub to a main then to both subs?

Process for align x 2 subs to front left and right mains :
1. Align Sub A with Sub B
2. Align both subs to a main
3. Align both subs to both mains.

Would you be able to add to this.

Thanks for your help.

 

[jtalden]

Align x 2 subs, but thought I had to have full range (with a tweeter) as a reference for time alignment?

Yes, one channel is the reference channel (L) and one is the measurement channel (R) for example. A tweeter is required in the L reference channel. It only provides the high frequency chirp needed to establish the start time for all the measurements on R channel.

I thought too time align x 2 subs with mains, the idea was to align each sub to a main then to both subs?

No, Step 2 can be done that way, but it is a lot more analysis work. We would likely eventually find a similar L vs R compromised answer for the mono SW system. I recommend using the sum of the mains with the sum of the SWs for the Step 2 analysis.

Process for align x 2 subs to front left and right mains :
1. Align Sub A with Sub B
2. Align both subs to a main
3. Align both subs to both mains.

Would you be able to add to this.

Analysis Steps
Step 1: Find favorable SPL for the SWs by evaluating delay, gain and polarity of SW1 vs SW2.
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 provide all the measurements needed to complete the analysis of both steps.

Setup:

• Mic at LP
• XO active (main speakers set to small)
• SWs set to work as mono
• REW Acoustic timing: active
• Sweep full-range for all 4 measurements (10-20k Hz for example)

Measurements:

1. L (alone)
2. R (alone)
3. SW1 (alone)
4. SW2 (alone)

For the SW1 measurement we can disconnect the wire to the R main speaker and turn off SW2. The receiver send the full range measurement sweep to the R channel, but the XO will filter out the high frequencies and only the low frequencies will arrive at SW1. We are thus measuring what is called the redirected bass.

 

[Harrycr]

Thank you
When doing the final step aligning Average L+R and Sub1+Sub2 do we use the aligned sum of the two subs?
Another words Average L+R and Aligned Sum (sub 1 & sub 2).

 

[jtalden]

Yes

 

[IWC Doppel]

Regarding time aligning when REW is saying measurements do not have timing reference will this be a problem for assessing and choosing delays from C or main to Subs ?

 

[jtalden]

Yes. Acoustic or loopback timing must be active in REW while taking the measurements for use in the alignment tool.

 

[IWC Doppel]

Thanks do you know how I do that it defaults to Loopback input R, the input device if the UMM-6 and the timing reference is L with the output device bing the HDMI to the Processor

 

[jtalden]

When using a HDMI mic like your we must use the acoustic timing option.
The REW reference channel chosen for all measurements must have a tweeter.
The REW measurement channel is used to select the speaker you want to measure.

 

[IWC Doppel]

Sorry I am being thick, I am looking at the sondcard options and dont know which options to select