Completed Automatic timing alignment for Main Speakers and Subwoofer

[Chris Bolkan]

it's just a """DIY""" project using what they already had ( multiroom delay)... doesn't really make sense... unless you really want to "bother yourself pulling cables"... and it would be interesting to see if it's ultimately effective...the wiim moved to the listening point.. (who did it and for example controlled with rew etc

proposing such an approach and thus considering the subject "closed" is, let's say...funny...

Well, if it works by just extending a couple of cables and does a good job of time aligning the subs to mains, is that any more of a PITA than using REW and a separate mic and then inputting the data? If this works, I think it is viable and makes sense. And if it works it would have been nice for someone from WiM to say that and explain on this forum maybe. I'll know after I try it out.

 

[canard]

Well, if it works by just extending a couple of cables and does a good job of time aligning the subs to mains, is that any more of a PITA than using REW and a separate mic and then inputting the data? If this works, I think it is viable and makes sense. And if it works it would have been nice for someone from WiM to say that and explain on this forum maybe. I'll know after I try it out.

indeed we can also see things this way... ;-)
but well it would be good to first validate and control the thing via other serious approaches... ;-)

but good luck...

 

[rccarguy]

Do you need to move wiim to the location of the LP when just using stereo speakers? (so each speaker is set different distances)

If it does have that ability how come no manual distance adjustment? (for each speaker profile)

 

[slartibartfast]

Well, if it works by just extending a couple of cables and does a good job of time aligning the subs to mains, is that any more of a PITA than using REW and a separate mic and then inputting the data? If this works, I think it is viable and makes sense. And if it works it would have been nice for someone from WiM to say that and explain on this forum maybe. I'll know after I try it out.

Let us know how well it works.

 

[slartibartfast]

Do you need to move wiim to the location of the LP when just using stereo speakers? (so each speaker is set different distances)

If it does have that ability how come no manual distance adjustment? (for each speaker profile)

The alignment is only for sub/main. We survived for decades with just a balance control for stereo

 

[s3igell]

Can you tell me why Bluetooth or Wi-Fi delay would be an issue?

In my layman's opinion, I would think that if the same amount of Bluetooth (Wi-Fi) delay is added to both the test signal for the main speakers and for the sub, it would not affect the timing adjustment. Am I wrong?

The "Timing Alignment" function works by measuring the absolute delay between when the DAC says "Make this Sound" and when that sound is actually "Recorded". If there are any indetermined delays in the communication of that "Make this Sound" signal, what is the process to use as that timing mark and how will it know what was "Network Delay" versus "Audio Device Chain Delay"??

 

[s3igell]

I still believe that the last few pages have been discussion of a Feature different than WiiM proposed and subsequently implemented. WiiM's admittedly few postings seem to indicate their goal was to quantify and address the Gross Delays in the Subwoofer Signal/Amplification Chain.

 

[slartibartfast]

The "Timing Alignment" function works by measuring the absolute delay between when the DAC says "Make this Sound" and when that sound is actually "Recorded". If there are any indetermined delays in the communication of that "Make this Sound" signal, what is the process to use as that timing mark and how will it know what was "Network Delay" versus "Audio Device Chain Delay"??

I have read that USB microphones in general don't work well for timing measurements due to delays. You can still use pink noise and an RTA to set timing alignment.

 

[slartibartfast]

I still believe that the last few pages have been discussion of a Feature different than WiiM proposed and subsequently implemented. WiiM's admittedly few postings seem to indicate their goal was to quantify and address the Gross Delays in the Subwoofer Signal/Amplification Chain.

Exactly. WiiM have said as much but still people don't believe it

 

[MrA]

Accurate time delay measurements are comparing initial impulse peaks from the speakers and sub(s). The highest (ie 'fastest') frequencies from both will provide this data but a tiny recessed mic within the casing of a WiiM device cannot be expected to achieve anywhere near the accuracy of a dedicated mic (as shown by the impulse readings in my earlier post #48).

This will be especially true with the main speaker tweeters where such a small internal mic is likely to roll-off significantly below the frequency range required for accurate comparison.

I'm afraid WiiMs approach is compromised not only by room location but also by the mic being used.

 

[harkpabst]

Thank you for clarifying. Except my speakers are not dipoles. Nothing that exotic. They are just pro audio speakers that have manufacturer's recommendations to EQ them for critical application.

Sorry, my aging eyes had for some reason misread "infinite baffle subs" as "open baffle subs", so I falsely assumed you had an all OB setup. My fault.

However, that doesn't change a thing regarding what I said. Keeping speaker EQ and room EQ separated is clearly the more powerful concept. You can easily do that using e.g. REW, but not relying on WiiM's room correction so far. Also, restricting room correction to a range below the Schrödinger frequency is generally a good idea, especially if correction is based on just one measurement. More advanced room correction systems (using multiple measurements and separate speaker EQ) can absolutely provide improvements in higher frequency ranges as well.

There's bee so much talking about latency and delay in one or the other context that I don't feel like stepping in. The important bits have all been mentioned, even if buried in lots of other stuff.

OK, don't listen to what I just promised, I have to add a couple of facts:
The delay caused by signal processing and the delay caused by different positioning of speakers relative to the listening position can all be measured in one go. There's no principal difference and they usually in a comparable order of magnitude. However, this only works if the mic can be placed in the listening position. Automatic timing alignment cannot currently do that, unless you move your Wiim to the MLP. However ...

... the delay caused by sub and speaker placement can actually be calculated easily using a tape ruler. 34 cm longer distances causes 1 ms of delay. Add that to whatever automatic timing alignment told you. Done. What's the problem?

Still not satisfied with the blending of mains and sub? In that case you'll need something like REW or ARTA (no longer maintained, I think) or CLIO or whatever, anyway. And not just for trying to capture a latency value. If you take out the real measuring equipment anyway, go ahead and optimise based on the resulting frequency response.

Latenc of USB measuring mics is not much of an issue. Programs like REW let you choose between different timing references and usually we're not really interested in absolute phase measurements (which needs compensation for the distance between mic and DuT anyway).

 

[Chris Bolkan]

Sorry, my aging eyes had for some reason misread "infinite baffle subs" as "open baffle subs", so I falsely assumed you had an all OB setup. My fault.

However, that doesn't change a thing regarding what I said. Keeping speaker EQ and room EQ separated is clearly the more powerful concept. You can easily do that using e.g. REW, but not relying on WiiM's room correction so far. Also, restricting room correction to a range below the Schrödinger frequency is generally a good idea, especially if correction is based on just one measurement. More advanced room correction systems (using multiple measurements and separate speaker EQ) can absolutely provide improvements in higher frequency ranges as well.

There's bee so much talking about latency and delay in one or the other context that I don't feel like stepping in. The important bits have all been mentioned, even if buried in lots of other stuff.

OK, don't listen to what I just promised, I have to add a couple of facts:
The delay caused by signal processing and the delay caused by different positioning of speakers relative to the listening position can all be measured in one go. There's no principal difference and they usually in a comparable order of magnitude. However, this only works if the mic can be placed in the listening position. Automatic timing alignment cannot currently do that, unless you move your Wiim to the MLP. However ...

... the delay caused by sub and speaker placement can actually be calculated easily using a tape ruler. 34 cm longer distances causes 1 ms of delay. Add that to whatever automatic timing alignment told you. Done. What's the problem?

Still not satisfied with the blending of mains and sub? In that case you'll need something like REW or ARTA (no longer maintained, I think) or CLIO or whatever, anyway. And not just for trying to capture a latency value. If you take out the real measuring equipment anyway, go ahead and optimise based on the resulting frequency response.

Latenc of USB measuring mics is not much of an issue. Programs like REW let you choose between different timing references and usually we're not really interested in absolute phase measurements (which needs compensation for the distance between mic and DuT anyway).

I 100% agree . I am so impressed by this little box that I am rooting for ALL of the features to work as well as everything I have played with so far. For sure I can just measure the physical distance and enter a delay. That's what I do now. They call it a sub alignment tool. It is very easy to lengthen a couple of cables and see if it does the calculation reasonably well. If it does I will take back everything bad I said at first. I did not get to the little test because we have company. Will do it as soon as I can! (of course I'm still hoping for capability to biamplify stereo with an additional DAC as a NEW feature )

 

[slartibartfast]

I 100% agree . I am so impressed by this little box that I am rooting for ALL of the features to work as well as everything I have played with so far. For sure I can just measure the physical distance and enter a delay. That's what I do now. They call it a sub alignment tool. It is very easy to lengthen a couple of cables and see if it does the calculation reasonably well. If it does I will take back everything bad I said at first. I did not get to the little test because we have company. Will do it as soon as I can! (of course I'm still hoping for capability to biamplify stereo with an additional DAC as a NEW feature )

I have 3ms latency setting in the WiiM Home app which seems to be best according to REW and my sub and speakers are the same distance from my listening position. It will be interesting to see the outcome of your test.

 

[Chris Bolkan]

OK! I did the experiment and from everything I can tell it measured perfectly and as accurately as it should ever need to considering a 100Hz crossover. I found some long RCA cables and a extra power cord I had laying around so it was super easy to set up. I attached a couple pictures of my system and test setup.

I had previously physically measured the difference in signal path between my mains and subs (which are IB type located underneath the main speakers in the basement). My setup is very symmetrical and easy to measure. I came up with about a 3 foot difference in signal path which ends up about 3ms at standard temperature and pressure. So I already had the mains delayed by 3ms in the box manually.

When I let the Ultra do the measurement automatically (after I positioned it close to the sweet spot) it came up with the following output latencies:
Main speakers: 13ms
Subwoofer:16ms
And the comment: "The main speakers lead the subwoofer by 3ms

As far as I am concerned it did a perfect job of what I needed it to do and it was actually super easy to set up.

So all WiiM needed to do was lay this out and explain it to users.

I continue to be impressed with this box and I officially retract my original comment on this thread except to say a little guidance would have been helpful.

Chris

 

[slartibartfast]

OK! I did the experiment and from everything I can tell it measured perfectly and as accurately as it should ever need to considering a 100Hz crossover. I found some long RCA cables and a extra power cord I had laying around so it was super easy to set up. I attached a couple pictures of my system and test setup.

I had previously physically measured the difference in signal path between my mains and subs (which are IB type located underneath the main speakers in the basement). My setup is very symmetrical and easy to measure. I came up with about a 3 foot difference in signal path which ends up about 3ms at standard temperature and pressure. So I already had the mains delayed by 3ms in the box manually.

When I let the Ultra do the measurement automatically (after I positioned it close to the sweet spot) it came up with the following output latencies:
Main speakers: 13ms
Subwoofer:16ms
And the comment: "The main speakers lead the subwoofer by 3ms

As far as I am concerned it did a perfect job of what I needed it to do and it was actually super easy to set up.

So all WiiM needed to do was lay this out and explain it to users.

I continue to be impressed with this box and I officially retract my original comment on this thread except to say a little guidance would have been helpful.

Chris

Interesting. I have the same latency setting as you and my speakers and sub are also equidistant. Where do you see
Main speakers 13 ms
Subwoofer. 16 ms
I don't remember seeing that in my tests for the Amp.

 

[Chris Bolkan]

Hi slartibartfast.
The WiiM app on my phone displayed that information after the test. It's gone now. I don't know how to get it back. I ran the test twice just to look for consistency and it displayed the info both times.

 

[slartibartfast]

Hi slartibartfast.
The WiiM app on my phone displayed that information after the test. It's gone now. I don't know how to get it back. I ran the test twice just to look for consistency and it displayed the info both times.

OK, I wonder what it compares each speaker latency to. The test signal I suppose. Maybe the Ultra displays different info, the Amp only stated the difference between mains and sub.

 

[Chris Bolkan]

It appears to me that the app is measuring the time of flight of each signal from the speaker to the microphone and calculating the difference. It figures out which is the sub and which are the mains by the tones it puts out. Clever.

I think its cool they provide time of flight, because now I know my mains are approximately 13 feet from the listening position (based on approximately 1ft/ms)

What I do not know (but it doesn't matter to me personally) is if once the auto measurement is performed the app automatically sets the Ultra or if I am supposed to go in and set it manually based on the info it provides. Since it measured the same value I had already set nothing changed, so I'll never know unless I set it up with different initial manual entries and run the test again. I didn't even think about that when I ran the test.

Again this is a case where it seems to me the app runs perfectly and does a wonderful job, but some instruction is missing.

 

[Wiimer]

I am curious if the auto-sync feature can do the same thing. I asked the team previously and did not get a reply.

 

[DG77]

OK! I did the experiment and from everything I can tell it measured perfectly and as accurately as it should ever need to considering a 100Hz crossover. I found some long RCA cables and a extra power cord I had laying around so it was super easy to set up. I attached a couple pictures of my system and test setup.

I had previously physically measured the difference in signal path between my mains and subs (which are IB type located underneath the main speakers in the basement). My setup is very symmetrical and easy to measure. I came up with about a 3 foot difference in signal path which ends up about 3ms at standard temperature and pressure. So I already had the mains delayed by 3ms in the box manually.

When I let the Ultra do the measurement automatically (after I positioned it close to the sweet spot) it came up with the following output latencies:
Main speakers: 13ms
Subwoofer:16ms
And the comment: "The main speakers lead the subwoofer by 3ms

As far as I am concerned it did a perfect job of what I needed it to do and it was actually super easy to set up.

So all WiiM needed to do was lay this out and explain it to users.

I continue to be impressed with this box and I officially retract my original comment on this thread except to say a little guidance would have been helpful.

Chris

My man! Nicely done and a great setup!