Using REW to Determine Time Delays Between Drivers

[marcopollo]

The question is about how to set the time delays on multi--amped (bi-amp, tri-amp, quadruple-amp, etc.) driver channels in a single loudspeaker using a DSP crossover.
I have a quadruple-amped (4 ways) loudspeaker wih DSP crossover, and I would like to accurately adjust the delays of each channel relative to the listening point so that the emission of each transducer is temporally aligned. Up until now, I have been doing this by measuring the relative position of each speaker with a ruler, but I know that this is not the correct method, as it does not take into account numerous factors.

 

[staticV3]

Using an audio interface, any microphone, and Room EQ Wizard, you can precisely measure time delays.

For this, connect the left channel Line out directly to the left channel Line in and the right Line out->Amp->Speaker->Microphone->Microphone input.

Then in the Measure dialog, turn on Loopback timing reference.

Once the sweep is finished, go to the Impulse window it'll show the precise delay.

 

[RayDunzl]

I would like to accurately adjust the delays of each channel relative to the listening point so that the emission of each transducer is temporally aligned.

In REW, if the speakers are not time/distance aligned, if both play the test swept tone at the same time, there will be two impulses on the graph.

This is a gross example, where the mic is about a foot off the center of the speaker centerline:

When the mic is in the exact center and equidsitant, the two impulses perfectly combine:

:

The little wiggle at 1ms is a bounce off the back of the sofa or the maike stand.

 

[marcopollo]

In REW, if the speakers are not time/distance aligned, if both play the test swept tone at the same time, there will be two impulses on the graph.

This is a gross example, where the mic is about a foot off the center of the speaker centerline:

View attachment 402246

When the mic is in the exact center and equidsitant, the two impulses perfectly combine:

View attachment 402248:

The little wiggle at 1ms is a bounce off the back of the sofa or the maike stand.

Thanks, but I would like to measure the time delay of each way in a speaker, in order to time-align them properly in the crossover

 

[alanhuth]

I'm wondering if you performed the above suggestion, but compared left tweeter to left mid, you could time-align them, and then repeat for the rest of the drivers.

 

[marcopollo]

I'm wondering if you performed the above suggestion, but compared left tweeter to left mid, you could time-align them, and then repeat for the rest of the drivers.

Sorry, I didn't understand! I have to turn off three ways and keep only one on at a time!
Then with the measurements in hand I can temporally align the four ways.
One last question: does this method also work with woofers and subwoofers?

 

[RayDunzl]

One last question: does this method also work with woofers and subwoofers?

Impulse Response of a 20-100Hz sweep:

The trace is more spread out, see the time scale, but I don't see why it couldn't be useful.

 

[Keith_W]

Firstly, are you using minimum phase DSP or linear phase DSP?

The standard way to time align individual drivers in a speaker is with reference to the tweeter.

 

[marcopollo]

Firstly, are you using minimum phase DSP or linear phase DSP?

The standard way to time align individual drivers in a speaker is with reference to the tweeter.

I really don't Know. Here attached is the datasheet and the user manual

 

[Keith_W]

You are using mixed phase. The reason why this is relevant is because the time alignment method is slightly different depending on whether you use linear phase FIR or minimum phase IIR.

With linear phase FIR: align tweeter peak to the peak of the second driver.

With minimum phase IIR: align tweeter start to the start of the second driver. In your case, this is what you should be doing.

I don't use REW for time alignment, so I don't know what the number it automatically spits out means. I would want to look at the curves to see where it obtained that number from. I use Acourate, and the method we use is slightly different. It still involves examining the curves though.

 

[Ruhled]

Minidsp has an article in their howtos on how to do this accurately with rew for years now. Sort i dont have a link but Google should turn it up.

 

[marcopollo]

Minidsp has an article in their howtos on how to do this accurately with rew for years now. Sort i dont have a link but Google should turn it up.

Thanks! found: https://www.minidsp.com/applications/rew/measuring-time-delay
and in my language too ! https://audiolai.it/blogs/progettar...emporale-driver-per-diffusori-a-due-o-piu-vie