Can anyone help with Alignment?

[Robbie010]

Afternoon all,

I'm just wondering if there are any REW wizz's would be willing to take a look at my sweeps and help me work out time / phase alignment?

See the attached file.

I took individual sweeps of each driver in the left speaker, using the right channel as time ref but I'm just getting in a muddle trying to understand how to use the alignment tool in REW.

Much appreciated.

 

[Philbo King]

This might work better if you actually attached the file.

 

[DVDdoug]

No file is attached.

If the left & right speaker are an equal distance from the listening position the sound will arrive in-time and in-phase. It doesn't have to be "perfect"... Two different listeners can't be in the same place and your left & right ears can't be in the same place.

If there is a difference you can fix the time & phase by calculating from the speed of sound. The loudness difference is trickier and it needs to be measured because it includes reflected sound that can't be calculated.

 

[Robbie010]

Apologies..... I can't seem to upload it, it says the file does not have an allowed extension.......?

 

[ppataki]

Apologies..... I can't seem to upload it, it says the file does not have an allowed extension.......?

just zip it

 

[Robbie010]

What a guy. Thanks.

Attached.

 

[ppataki]

What a guy. Thanks.

Attached.

I think your measurements were too much filtered, since if I put them on top of each other they hardly have any common frequency range
As if they were all HPF-d and LPF-d....
This way it would be impossible to create a crossover

Anyway, to do the delay calculation I would take the tweeter as a reference and perform measurements as per the below:

1. Tweeter (ref) --> Tweeter - let's say this would be 0.001 ms (you can see that figure where it says Δt at each measurement)
2. Tweeter (ref) --> Mid - let's say this would be 0.1 ms
3. Tweeter (ref) --> Woofer - let's say this would be 0.5 ms

Now you calculate the delay (that you need to add to the tweeter and to the mid) like this - check column called Time align (ms)

Formulas look like this in Excel if your three delay values are in range I14:I16 (in my case they are but of course this can be changed)

With this table you can easily perform any delay calculations (you can add or remove more speakers, even a sub)

 

[Robbie010]

Thanks for taking a look.

Can the delays not be estimated by viewing the overlayed Impulse response?

I read elsewhere that the delays can be set by aligning the rise of the tweeter and mid with the woofer?

 

[ppataki]

Thanks for taking a look.

Can the delays not be estimated by viewing the overlayed Impulse response?

View attachment 479333

I read elsewhere that the delays can be set by aligning the rise of the tweeter and mid with the woofer?

There are multiple ways indeed
I will let the other fellow members comment on them, to be honest I always use the time alignment method only....

 

[Cbdb2]

Are those measurements right? Is it just me or does that midrange look almost useless? The woofer covers the same range and the mid impulse is 20db below the woofer and tweeter. What speaker is that?

 

[Robbie010]

I’m trying to setup a DIY 3 way speaker.

 

[Cbdb2]

I’m trying to setup a DIY 3 way speaker.

Ok I see. Im not a speaker guy so take this with a grain of salt. The measurements look like the mid should fill in the 3khz hole between the woofer and tweeter not overlap with the woofer high end. I dont know what drivers your using but does that woofer really go to 3khz with out break up?

 

[Sokel]

Poor woofer is working 1200Hz over what it should be working.
What speaker is that???

Edit: Tweeter is strange too, there can not be any meaning IR this way.

 

[Robbie010]

Ok I see. Im not a speaker guy so take this with a grain of salt. The measurements look like the mid should fill in the 3khz hole between the woofer and tweeter not overlap with the woofer high end. I dont know what drivers your using but does that woofer really go to 3khz with out break up?

Woofer is full range Fane 12-250TC. If anything, the midrange crossover is off.

 

[Salt]

You might , at first, every driver measure at the most narrow distance as possible to trace the response, ignoring enclosure (does not need a window essentially).
Then measure at 1 meter distance (that needs a window to exclude room interference) and look what you get.
....

 

[Robbie010]

I've already done multiple nearfield and far field measurements, all the drivers pretty much measure as per published info suggest they should, not withstanding room interactions. Crossovers I have chosen are 1,200hz & 5,000hz. I'm still dialling in the slopes and gain settings but wanted to get the time alignment sorted, hence this post:

Woofer - Fane 12-250TC:

Midrange - Faital Pro HF144:

Tweeter - Visaton TL16H:

 

[ppataki]

I've already done multiple nearfield and far field measurements, all the drivers pretty much measure as per published info suggest they should, not withstanding room interactions. Crossovers I have chosen are 1,200hz & 5,000hz. I'm still dialling in the slopes and gain settings but wanted to get the time alignment sorted, hence this post:

Woofer - Fane 12-250TC:

View attachment 479377

Midrange - Faital Pro HF144:

View attachment 479379

Tweeter - Visaton TL16H:

View attachment 479380

Your measurements that you shared do not really correspond to these graphs though
Your measurements show a very narrow frequency coverage for each driver - as if they were high-passed and/or low-passed

This is what I mentioned in my previous post

 

[Robbie010]

Your measurements that you shared do not really correspond to these graphs though
Your measurements show a very narrow frequency coverage for each driver - as if they were high-passed and/or low-passed

View attachment 479382

This is what I mentioned in my previous post

Understood.

Forgive my ignorance, my assumption was that FR / crossover settings didn’t matter.

So should the time alignment be done first, before any crossover settings are applied?

Should I take a full range sweep of each driver (within its published range) and time align the native responses then re-measure and move on to gain and crossover settings?

 

[Thomas_A]

Understood.

Forgive my ignorance, my assumption was that FR / crossover settings didn’t matter.

So should the time alignment be done first, before any crossover settings are applied?

Should I take a full range sweep of each driver (within its published range) and time align the native responses then re-measure and move on to gain and crossover settings?

The graphs do not make sense. Curves should meet at -6 dB at each crossover point if you e.g. use a 24 dB RL filter.

 

[Robbie010]

The graphs do not make sense. Curves should meet at -6 dB at each crossover point if you e.g. use a 24 dB RL filter.

They do now, or nearly do, I’m still playing around with the crossover settings. Those were done before I set the current crossovers.

As I mentioned, I assumed that crossovers weren't relevant with regards to time aligning the drivers.

I’ll post some new sweeps tomorrow.