Time alignment vs phase alignment for subwoofer and main

[lovetl2002]

Using REW I measured system delay of my subwoofer referenced to my main speaker and it’s 3ms. I used this delay to crossover at 90hz but unfortunately it gave me a dip. Then I tweak the delay with tries and errors and found that 7ms with phase inverted gave me the best SPL performance on crossover area. I’m tempted to use 7ms but will that extra 4ms delay cause audible difference?

 

[holbob]

Are you measuring through rew using the alignment tool? I used to use that when I had a monoprice monolith sub, and found it fixed phase and delays perfectly. I didn't have to choose.

 

[dedobot]

Interesting, I'll follow the thread. I think 7ms delay is within the expected range when adjusting the subwoofer in phase with the mains.

 

[Chris A]

I’m tempted to use 7ms but will that extra 4ms delay cause audible difference?

In my experience, yes, it will, if you're crossing at 80 Hz or higher (i.e., almost half wavelength added delay of the loudspeakers relative to the subwoofer using an added 4 ms).

Thus far, you haven't mentioned the crossover frequency you're using and whether or not you've got higher-frequency loudspeaker time alignment in place. I've also found that the method used to determine subwoofer-loudspeaker time alignment can also be an issue. Where exactly is the microphone (including above floor level), and did you move it around and take other measurements to see if you're designing on a "knife edge response" at your listening position?

For reference, I cross my loudspeakers nominally at 40 Hz (not 80 Hz or higher). That extra 4 ms, if applied at 40 Hz crossover, is less than 90 degrees of phase lead by the subwoofer(s)--which is generally inaudible. Polarity reversal says you're trying to eliminate an extra 180 degrees of phase from your subwoofer response relative to the mains.

Chris

 

[dedobot]

The other option is no time domain corrections and the sub goes 2.4 meters forward

 

[Keith_W]

Welcome to ASR!

In general, phase alignment is FAR MORE IMPORTANT than time alignment. This is why:

1. As a rule of thumb, the audible group delay threshold is about 1 period of the frequency in question. For 90Hz, that would be 11.1ms.
2. Failure to align the phase will cause aberrations in the amplitude response. This is far more audible than a subtle increase in group delay.

Read Bob McCarthy on time and phase alignment.

What I do: I time align the subs first. Then I align the phase. This will cause the time alignment to shift. To me that is an acceptable price to pay for the reasons above.

 

[lovetl2002]

Are you measuring through rew using the alignment tool? I used to use that when I had a monoprice monolith sub, and found it fixed phase and delays perfectly. I didn't have to choose.

Yes. I also tried phase alignment tool but its delay gave me a dip at another frequency (120 hz), though the crossover is fine. Only 7ms resolve all the dips.

 

[Dual]

Please see:

Sound Systems: Design and Optimization: Modern Techniques and Tools for Sound System Design and Alignment
by Bob McCarthy

The phase in the crossover region needs to be aligned. In small rooms, eg regular listening rooms, it is hard to measure phase at low frequencies because we are dealing with standing waves reflections and other artifacts in the region of interest (low frequency). Determining the time of arrival of a wave that is 3.4 m long aka 100Hz is not trivial either.

See also https://www.merlijnvanveen, e.g

Cheers, happy alignment ;-)

 

[gnarly]

What I've been taught at several Smaart classes, is that both time and phase need to be correct for optimal phase-traces-overlay, between sub and main.

At the lower left hand corner of Dual's snip from merlijn, an example of 100Hz half cycles is shown.
If time alignment is correct / relative to 0 cycles offset, then phase can dialed in to find overlay without further consideration of time. All-pass can be used to bring whichever side, sub, or main, has the lesser phase rotation, up to the rotation of the greater side.
If time alignment is in fact relative to any of the half-cycles, it will be off by the implied time offset. Phase can still be manipulated to overlay even with the fixed time being off, but it will not be possible to achieve overlay for as wide a frequency range or as precise, as when adjusting phase with 0 time offset.

Pragmatically, whenever I'm convinced I've dialed in a good phase alignment, I reduce the fixed delay increment another half-cycles worth of time, invert the polarity from wherever it was at, and see if phase alignment can be further improved. If phase gets worse, I was at t=0 before already , so best previous alignment gets verified.

But the big thing / big problem for me...is it's simply damn hard to get good sub / low freq measurements indoors.
Reflections swamp the measurements. Must admit, I'm super skeptical of most all the indoor measurements I see posted of subs or mains down low, made indoors.

Anyway, I like to measure a sub out on the driveway, and lock in a phase trace and an acoustic distance, relative to the sub's front grill.
And do the same thing for the main, or better yet, time and phase align the whole thing stacked on the driveway.
Then indoors, either stack them the same way, or use a laser distance finder to front grilles, to determine what delays need to be put in place to make them arrive together, as if stacked on driveway. Indoor time and phase alignment is a bitch!!! ..... Lol

 

[Dual]

I used SMAART to align indoors sound systems for a while, until I tried Dirac (bass control). Works like a charm, full round bottom - just the way we like it ;-) I have no idea how they do it. Dirac plays some sweeps and then calculates for a while... Not only bass comes out great but also overall impulse response is fixed. Stage, image, depth, clarity all on a new level. I still use SMAART and / or REW for general measurements, verification etc.
Happy measurements ;-)

 

[markstein]

I propose also trying MSO (multi subwoofer optimizer). It"s a free software. I guess you can use it to even align one sub with your two mains. You can manually play with delay and PEQ and see the effect at all listening places you measured (measuring groups). You can also show phase and frequency response in one chart.

 

[Holmz]

Using REW I measured system delay of my subwoofer referenced to my main speaker and it’s 3ms. I used this delay to crossover at 90hz but unfortunately it gave me a dip. Then I tweak the delay with tries and errors and found that 7ms with phase inverted gave me the best SPL performance on crossover area. I’m tempted to use 7ms but will that extra 4ms delay cause audible difference?

But what sub izzit?

The sealed subs, ported, bandpass etc are all different.
If one wants the easiest one with the least phase or time EQ then the order should be low.
I guess that infinite baffle being 1st order is the winner, and sealed being 2nd order is in second place.

Since you already have the sub then you make it work.
But if you did not, then it might be possible and even likely that a couple of sealed subs maybe be easier to integrate and cheaper without the need for extras, than a ported sub.

Or as Confucius say, “Man who peel onion, he get watery eye.”

 

[ZolaIII]

But what sub izzit?

The sealed subs, ported, bandpass etc are all different.
If one wants the easiest one with the least phase or time EQ then the order should be low.
I guess that infinite baffle being 1st order is the winner, and sealed being 2nd order is in second place.

Since you already have the sub then you make it work.
But if you did not, then it might be possible and even likely that a couple of sealed subs maybe be easier to integrate and cheaper without the need for extras, than a ported sub.

Or as Confucius say, “Man who peel onion, he get watery eye.”

Don't create the time domain problem so you can skip worrying about it. Also improving decay times and delay to a point of not being very relevant and that only lefts you phase (and impuls) alignment to deal with. As open baffle will suffer from problems of its own using the room as the box that leaves you sealed 10" one's. General pick is mid 30's Hz SPL capabilities. You do crossovers to the in room decay order not the enclosure.

Rule of the tumb, make less problem so you later have to deal with less problems. Same thing goes for speakers with cardioid alike dispersion patern (minimising impact of room refractions and ensuring you get what's proper natural in room response). Of course that's possible up to standing waves but not really needed under uper bass.

 

[gnarly]

I used SMAART to align indoors sound systems for a while, until I tried Dirac (bass control). Works like a charm, full round bottom - just the way we like it ;-) I have no idea how they do it. Dirac plays some sweeps and then calculates for a while... Not only bass comes out great but also overall impulse response is fixed. Stage, image, depth, clarity all on a new level. I still use SMAART and / or REW for general measurements, verification etc.
Happy measurements ;-)

I should give Dirac a good second go. I tried its setup routine a few times, and it had very little effect sonically.
I also made transfer functions of what Dirac was doing electrically, of the filter it applies, for the left and right speakers. It appeared Dirac uses the left side speaker/sub as the one to match the right side to.
Phase alteration of the right side sub was the only marked change Dirac made to the tuning I had in place which was identical flat mag and phase tuning, linear-phase tuning made outdoors for both speakers/subs. It made no discenable fixed time changes. So I guess Dirac was accounting for the effect the room was making on the subs' phase.

I'm a bit of an obsessed student of measurements, using Smaart, REW, ARTA, and for the last year my new go-to Crosslite+.
The best technique I've found for time location of subs (or low frequency) indoors, is using wavelets / tone bursts.
With REW, its signal generator and its 2-ch scope make it pretty easy to find time-of-flight, comparing loopback trace on scope ch-1 vs acoustic mic grab trace on scope ch-2.
Crosslite+ can do that too via its scope, but also has the ability to extract wavelets at any frequency from either transfer functions or impulse responses. Let's you choose what frequency do you want to key time and phase alignment to. Get's you to the same phase overlay method, but much faster knowing where t=0 is.

 

[Dual]

The full version

Dirac Live - Room Correction Suite - Bass Control​

worked for me. Bass control is a bug step up in room correction. And it is much quicker and easier than manual alignment - uff - what a time safer!
But then, there is no free lunch. Yet, it was the best investment in getting better sound.
Cheers

 

[gnarly]

The full version

Dirac Live - Room Correction Suite - Bass Control​

worked for me. Bass control is a bug step up in room correction. And it is much quicker and easier than manual alignment - uff - what a time safer!
But then, there is no free lunch. Yet, it was the best investment in getting better sound.
Cheers

Thanks. My impression is that the Bass Control module is for multi-sub and/or regular stereo where separate subs may or may not be co-located with main speakers.
I wonder if the underlying measurement and tuning algorithms do anything differently than the Room Correction Suite (which I have w/o Bass Control.
If the Bass module does use a different process, I might snag it next time Dirac holds a sale.
Can you tell if the bass module would setup the subs on pair of stereo L & R stacks, any different than the regular room correction suite being applied only to the sub/low frequency sections?

I'm only interested in the tech as far as how it measures single subs, and then the mixed phase corrections it would use to implement their integration.
I'm not really interested in a highly optimized listening position/area, and am not a fan of multi-sub.
Must say, I'm not really a fan of room correction.
I like to get DIY speakers as quasi anechoically correct as possible, and then use acoustic treatments for room tuning.
So I use very little electronic filtering for room other than to cut a strong room mode.
All that said, I'll EQ a particular track to taste tonality wise, in a heartbeat.

 

[Dual]

Bass Control is another step! The woofers work in unison, thus augmenting each other, no matter of large speaker or subs. I was amazed by the upgrade.
I also use acoustic treatment, but in the bass department it needs real estate. So it is of limited success, but removing echoes and reducing reverb in the upper registers improved speech intelligibility and imaging. Bass Control takes care of the low end. Multi-subs is - by the way - to go in small rooms, see the work by Welti J. Audio Eng. Soc., Vol. 54, No. 5, 2006 347 and FLOYD E. TOOLE from Harman International (but perhaps you know that already)!

I completely agree: EQ, compress, saturate, add harmonics, sub-bass, space etc to your personal taste! Mix it up to surround! Just amazing!