The right way to align speakers?

[pratul]

I am trying to get all my speakers aligned and have read about two ways to align them:
1. Align using the delays shown in the measurements
2. align using the impulse graph

I aligned them using the impulse graph but then the delays in the measurements don't match.

Which one is the right way?

Thanks!

 

[John Mulcahy]

Better plotting the impulse responses normalised. They look pretty well aligned, within +/- 0.1 ms. A little odd that the ref channel, C, shows a fairly large delay on its own measurement but I'd leave them as they are.

 

[pratul]

Thanks for the reply @John Mulcahy ! I was wondering about the delay shown on the measurement of the center channel. Maybe that is why the delays for the channels don't match while IR shows them as aligned. Guess, I will try measuring them again. Thanks for the reply and developing this wonderful application. I am hooked

 

[sm52]

1. Align using the delays shown in the measurements
2. align using the impulse graph

Can you please tell me where you read about these methods?

 

[pratul]

@sm52 Page 117

 

[sm52]

Thank you. Maybe you know the answers to questions when a series of measurements is taken using acoustic synchronization:
1. Can 'Estimate IR delay' be applied?
2. Can 'Offset t = 0' be applied?
3. Can 'Clock ajustment' be applied?
Or, after a series of measurements of several speakers, to find the delays relative to the reference speaker, it's needed to leave all measurements as they are?

 

[Jack1949]

@sm52 Page 117

FYI: That is a down level version of AustinJerry's REW document (Rev 5.43, June 3, 2017). The current version (Rev. 6.0, March 22, 2020) can be found here: https://www.dropbox.com/s/51jpnxet3bvew2k/REW 101 HTS Current Version.pdf?dl=0, page 113.

 

[pratul]

Thanks @Jack1949 !

 

[pratul]

Thank you. Maybe you know the answers to questions when a series of measurements is taken using acoustic synchronization:
1. Can 'Estimate IR delay' be applied?
2. Can 'Offset t = 0' be applied?
3. Can 'Clock ajustment' be applied?
Or, after a series of measurements of several speakers, to find the delays relative to the reference speaker, it's needed to leave all measurements as they are?

I'm sorry but I am just getting started myself. I'd just follow the doc above. Maybe others more knowledgeable can answer you question.

 

[sm52]

The current version (Rev. 6.0, March 22, 2020)

Thanks.

pratul, I looked at your measurements. It looks like your microphone is inverting the received signal. Condenser microphone? To fix, find the 'invert' button in the sound card control panel or in REW.

 

[pratul]

Thanks.

pratul, I looked at your measurements. It looks like your microphone is inverting the received signal. Condenser microphone? To fix, find the 'invert' button in the sound card control panel or in REW.

OK.. can you tell me more? What does it mean and what are the implications? I am using the UMIK-1 and have not seen anything that says I should configure it with the 'invert' button on. Maybe I missed it.
Thanks!

 

[sm52]

The phase characteristic changes. The impuls becomes positive. Maybe for your purposes this is irrelevant.

 

[pratul]

Thanks! Good to know.

 

[jtalden]

The posted measurements have the conventional polarity - positive polarity for the tweeter.

 

[sm52]

jtalden,
Please tell me your opinion on the issues in post #6.

 

[jtalden]

Thank you. Maybe you know the answers to questions when a series of measurements is taken using acoustic synchronization:
1. Can 'Estimate IR delay' be applied?

No, the relative timing of the group is lost if this is done on individual measurements.
Yes, if done on the first measurement and then that value same value is used as an offset for the rest of the group measurements, i.e., 'shift and update timing offset'.

2. Can 'Offset t=0' be applied?

No, When applied only to an individual measurement in a group of timed measurements because the relative timing is lost.
Yes, if all measurements in the group are moved by the same amount of time.

3. Can 'Clock ajustment' be applied?

Yes, The choice of the clock adjustment settings in REW 'preferences/analysis' does not impact speaker or driver timing analysis.

Or, after a series of measurements of several speakers, to find the delays relative to the reference speaker, it's needed to leave all measurements as they are?

The relative timing of the group of measurements being used for timing must not be changed, but we can move all the members in a group the same distance if we want. It is easiest to do that in the impulse overlay window. Select the group we want to move and place the cursor at the time that we want to become 0 ms and then apply 'Set t=0 at cursor'. All selected measurements are shifted together as a group and thus the same timing relationship will be preserved.

 

[sm52]

jtalden
Is it the same fr "Use loopback as timing reference"?
The questions dealt with the 'Impuls' window, the 'Controls' options.

 

[jtalden]

I don't know the when and how that control is properly used.

 

[sm52]

jtalden
Oк. Thank you for your help.