Time / Phase Aligning My Speakers / Subs with REW

Up front: I'm not anything like an expert on this and this is just me showing how I used the alignment tool feature in REW along with an interesting way to set the timing reference for an all digital system like I have. If you can see problems with what I did (other than the one where I forgot to move the cursor back to 300Hz to do the phase alignment) then please feel free to say so in the comments.

Time alignment has never been a priority for me, since the speakers I've made in the past have always been "close enough" physically for it not to be a concern. And phase alignment with a subwoofer always shows in the frequency response, so I've taken care of that there.

With that said I thought I'd try to "nail" the timing and compare that to the setup that was just tuned by the frequency response without any delays added. The beauty of the DSP is that I can set different configurations and quickly switch back and forth to listen and decide if there is a difference.
Another disclaimer: my hearing is well past it's best before date and while I do wear hearing aids that help fill in what's been lost, I don't have golden ears.
In all cases I heard a difference, but can't say with certainty that the difference I heard was due to timing or small differences in the frequency response. And I can't say that one sounded best, or even better than the others. They sounded different - that's all. And within a few hours of listening to a new configuration, my ears (or more precisely my brain) got used to the sound.
I'll add that everything I tried sounded great. The differences were subtle and in all instances I aimed for the most even frequency response I could get.

As I pointed out above, I made a mistake while filming this. I failed to move the cursor back to the 300Hz crossover point before hitting the "align phase at cursor" button. But even then the final results STILL sounded great!
I later redid the whole thing again just for peace of mind sake. And it's my opinion (you may disagree) that doing these little tweaks are valuable for just that: peace of mind. I don't think time alignment on a single baffle is make or break thing. I think phase alignment is important, but like I said above, any major phase problems will show up in the measured frequency response.

I'll say it again: I'm not an expert or pretending to be one. I know quite a lot about designing and building loudspeakers, but I'm also aware of how much I DON'T know. Admitting I don't know it all shouldn't be taken as proof that you know more than me. You may, and the best way to show that is to be politely helpful. If you have tips and advice, think about how YOU would like to be told them.
Also realize that I'm first and foremost a practical man. I recognize factors that are genuinely important and pay a lot less attention to ones that are of marginal value. Time alignment is one of those marginal factors. We do it because we can and for peace of mind, but shouldn't expect miracles as a result.
Perfection only exists in the minds of those who have never tried to produce it.

You can help support the work I do in making these videos:
Support this channel on Patreon:

#diyspeakers
#johnheisz
#audio

My "Scrap bin" channel:

My main channel:

[IBuildIt]

Up front: I'm not anything like an expert on this and this is just me showing how I used the alignment tool feature in REW along with an interesting way to set the timing reference for an all digital system like I have. If you can see problems with what I did (other than the one where I forgot to move the cursor back to 300Hz to do the phase alignment) then please feel free to say so in the comments.

Time alignment has never been a priority for me, since the speakers I've made in the past have always been "close enough" physically for it not to be a concern. And phase alignment with a subwoofer always shows in the frequency response, so I've taken care of that there.

With that said I thought I'd try to "nail" the timing and compare that to the setup that was just tuned by the frequency response without any delays added. The beauty of the DSP is that I can set different configurations and quickly switch back and forth to listen and decide if there is a difference.
Another disclaimer: my hearing is well past it's best before date and while I do wear hearing aids that help fill in what's been lost, I don't have golden ears.
In all cases I heard a difference, but can't say with certainty that the difference I heard was due to timing or small differences in the frequency response. And I can't say that one sounded best, or even better than the others. They sounded different - that's all. And within a few hours of listening to a new configuration, my ears (or more precisely my brain) got used to the sound.
I'll add that everything I tried sounded great. The differences were subtle and in all instances I aimed for the most even frequency response I could get.

As I pointed out above, I made a mistake while filming this. I failed to move the cursor back to the 300Hz crossover point before hitting the "align phase at cursor" button. But even then the final results STILL sounded great!
I later redid the whole thing again just for peace of mind sake. And it's my opinion (you may disagree) that doing these little tweaks are valuable for just that: peace of mind. I don't think time alignment on a single baffle is make or break thing. I think phase alignment is important, but like I said above, any major phase problems will show up in the measured frequency response.

I'll say it again: I'm not an expert or pretending to be one. I know quite a lot about designing and building loudspeakers, but I'm also aware of how much I DON'T know. Admitting I don't know it all shouldn't be taken as proof that you know more than me. You may, and the best way to show that is to be politely helpful. If you have tips and advice, think about how YOU would like to be told them.
Also realize that I'm first and foremost a practical man. I recognize factors that are genuinely important and pay a lot less attention to ones that are of marginal value. Time alignment is one of those marginal factors. We do it because we can and for peace of mind, but shouldn't expect miracles as a result.
Perfection only exists in the minds of those who have never tried to produce it.

[JDG]

John What you've done, and then the It's even better than good! Great brother. Splendid to see. How and what and why? Good information and your work is exceptionally beautiful. You have passion. I really enjoyed it. Greetings from the Netherlands Limburg Maastricht and of course continue with what you are doing. Enjoy it. Peace and love and of course coffee beans.
✌🏾🌍😘☕️💯%

[leslubiesdetienne]

Very cool! I've been reading up on all this methodology for years and never got around to implementing it so it's nice to see a demo of the whole process. thanks!

[KipdoesStuff]

Way above my pay grade, lol. Nice chat though, I could follow along. I'm one of those guys who listens to them straight out of the box and the only adjustment is volume.

[sudd]

when i did Phase Aligning it was so much to learn and i did not found much information on it. but also my drivers were so physically on the same plane(output of the cones that is) that my hypex dsp did not adjust for so low increment, you can calculate the ms offset if you measure the cone location difference.

but for my subwoofer and main speaker its much larger, i ended up doing that by ear, the measurement looked helpful but dont know if it helped.
in the overlap frequency you get either a dip or bump, bump preferable, then i could adjust the delay to i moved to the other bump in time(phase), andi could tell that my sub with either by head or behind, some drum music helped me out here.

[fredygump]

well, looks like I know what I'm doing for the rest of the evening! Now I gotta try this technique out on the speakers I'm working on.

[MrFrobbo]

Great vid for education, but as you say even a small move of your listening position would change all the delays/invertions, so I just cant see the DSP and all this setup practical. Am I missing something?

Might be useful for setting a sub against the other cabs?

[jmi]

Really good video John. Time aligning the mid and tweeters is very important for good temporal summation at the listening position. If summed correctly you should notice an improvement in imaging. With woofers and subwoofers, I'd argue that it is less important (as long as the frequency response seems relatively good) due to the larger wavelengths, however if done correctly you can achieve more "even" bass throughout a given space - which can be great if multiple listening positions are required. It can be easier (or not) to approach bass summation using both delays and all-pass filters. Correcting for group delay in you speakers is perhaps another interesting avenue to go down!

[InsectMindNumb]

this is much easier actually than many videos ive seen...can you comment on how the IR window showed the delay/time alignment changing to come into alignment?

[impuls]

I'n the next vid I would like to see the off axis respons and how it changes by adjusting the XO filters up and down for the mid and tweeter. It would also be fun to hear your impression of how the sound height/with changes in real life. Having good flat off axis respons is a must for pinpoint imaging and to get the speakers to disappear. I'd say your 60percent done in this vid. Changing the polarity should have a massive effect in practice, I think the smoothing applied just hides the sharp dips from showing up. When I change my tweeter polarity on my physical filter it has a huge impact on SPL in the XO area, (but the changing the timing would help, I'm just not sure if its the best way to do it).
Thanks for showing us how the miniDsp works in practice! I think a lot more people will try it having seen it demonstrated :)

[hdmoviesource]

When I click align phase slopes at cursor it doesn't do anything. Are there any requirements? I'm using acoustical timing reference. I measured my left sub and right sub and there's no movement when I click it, am I missing something?

[randomgoose]

Hey, John. I have no knowledge and relatively zero audiophile knowledge.... but I'm watching because of your charisma and that pancake eating video (although your video editing surpasses both). I had questions, but the more I watched the more I appreciated your dedication. I can't hear it, but the set up looks magnificent.

I know this channel isn't for me, but I'm allowed aren't I!

[wio]

Huge fan of your work! I’m no expert at all when it comes to audio. Especially the physics involved. I’m curious if you’ve ever tried doing this using a tape measure and the speed of sound?

How do you pick which component you want the timing to be based off of?

[UberAlphaSirus]

Good stuff John. This is really helpful. What mic are you using?
Im using the built in DSP of my sony db1080 6.1. Not as fancy as your dsp, but I'm not bi, tri or quad amping like you. It phases by speaker distance. Can eq full range speakers and the .1 needs an external amp but you can set the phase, distance and xover point. I'm having trouble getting my 15" subs to blend into these Kef towers. Didn't relise REW had so many gadgets. Thanks fella.

[henzyk]

Hi Sir ! want to ask you,

There are Left & Right side speakers,
After we phase align one side,
We are still using same reference speaker driver as reference right ?

Then as per Left & Right, how we do Time Alignment ?

Eg. For Mid Left with Mid Right, Because those those driver produce identical frequency band. They can be time aligned.
Eg: in the car where Listening position can be off axis

Or we do Full Left & Full Right to do Time alignment ??



Thank you for you to share knowledge. Thank u very much..

[Websurferz]

What dsp software did you use? Didnt get thr name

[mabehall]

What a great, informative video. Tks. Wish I had wired my CSS audio speakers for possible active filter so I could compare with the passive system.

[stevefoster]

Very interesting! I'm curious if changing the delay times lie that have any noticeable impact on imagining? It seems like it might if different frequency bands are arriving at your ears with a few milliseconds difference in timing.

[davidketley]

Surely you need to use a laser measure to make sure the speakers are physically aligned first?

[kguenot]

J'utilise a 95% cette méthode, mais tu devrais faire du fenetrage (ir windows) afin de rendre la phase plus visible et d'ajuster au mieux les delais. Des délais proches de 9ms sont bien trop importants ce n'est pas cohérent en centimètres et ca influ sur la justesse général de l'enceinte. ;)