Beta Test: Automatic timing alignment for Main Speakers and Subwoofer

[slartibartfast]

It was never intended to change the "phase" setting, but the phase relation between mains and sub. This is ontrolled through the latency.

The "phase" switch changes the polarity. It's a misnomer, even if a common one.

The latency gives finer control than the phase switch though and it is common to use the latency instead.

 

[harkpabst]

The latency gives finer control than the phase switch though and it is common to use the latency instead.

The phase switch doesn't actually delay anything. It inverts the polarity. Inverted polarity might help getting along with a lower latency setting in certain cases, but that's all.

Latency setting is really all that is needed. It obsoletes the polarity switch.

 

[slartibartfast]

The phase switch doesn't actually delay anything. It inverts the polarity. Inverted polarity might help getting along with a lower latency setting in certain cases, but that's all.

Latency setting is really all that is needed. It obsoletes the polarity switch.

I agree that latency is all that is required but the method used for adjusting it automatically by WiiM doesn't maximise amplitude at the crossover point. Isn't that the goal?

 

[Smartplug]

it uses wiim mic instead the mobile mic. Isn’t the mic in wiim located at the rear and what if this device was placed very close to the wall would it still be accurate? I’m not sure why it wasn’t used phone mic where your listening position.

 

[slartibartfast]

The phase switch doesn't actually delay anything. It inverts the polarity. Inverted polarity might help getting along with a lower latency setting in certain cases, but that's all.

Latency setting is really all that is needed. It obsoletes the polarity switch.

Last night I played pink noise to the Real Time Analyser in REW and watched the response near the crossover at various latency settings. The optimum setting with the maximum level was 3ms compared to 8ms from the automatic process. So this automatic timing seems to be purely to eliminate any audible delay between the sub and main speakers and not really helpful for integrating a subwoofer. WiiM themselves say further adjustments will be necessary from the listening position but neglect to say how to go about making these adjustments.

 

[harkpabst]

Last night I played pink noise to the Real Time Analyser in REW and watched the response near the crossover at various latency settings. The optimum setting with the maximum level was 3ms compared to 8ms from the automatic process. So this automatic timing seems to be purely to eliminate any audible delay between the sub and main speakers and not really helpful for integrating a subwoofer. WiiM themselves say further adjustments will be necessary from the listening position but neglect to say how to go about making these adjustments.

Not exactly satisfying.Did you also try flipping the polarity?

 

[slartibartfast]

Not exactly satisfying.Did you also try flipping the polarity?

Flipping the polarity gave a very similar response to my 3ms latency setting.

 

[Smartplug]

Last night I played pink noise to the Real Time Analyser in REW and watched the response near the crossover at various latency settings. The optimum setting with the maximum level was 3ms compared to 8ms from the automatic process. So this automatic timing seems to be purely to eliminate any audible delay between the sub and main speakers and not really helpful for integrating a subwoofer. WiiM themselves say further adjustments will be necessary from the listening position but neglect to say how to go about making these adjustments.

It’s still on beta so wiim is tweaking to get it more in line what the ideal should be. As time goes by, this feature will evolve and will be enhanced like what they did on rc. My prediction at the end of next year rc would be more sophisticated than ever.

 

[slartibartfast]

It’s still on beta so wiim is tweaking to get it more in line what the ideal should be. As time goes by, this feature will evolve and will be enhanced like what they did on rc. My prediction at the end of next year rc would be more sophisticated than ever.

Didn't you see WiiM's response? They aren't even trying to match phase at the crossover frequency.

"Apologies for the confusion - I was referring to the mic on the WiiM device.

Also, please note that the current setup doesn’t account for the distance between your listening position and your subwoofer or main speakers. For the best sound adjustment, it helps to know the exact distance from your listening position to each speaker. Otherwise, further calibration may be needed, ideally using a mic placed at your listening position."

 

[harkpabst]

Flipping the polarity gave a very similar response to my 3ms latency setting.

So, at least this remains consistent ... did you mention the crossover frequency yet? I don't remember.

 

[slartibartfast]

So, at least this remains consistent ... did you mention the crossover frequency yet? I don't remember.

The crossover is set to 70Hz but the sub is set to +15 dB in the settings to let room correction use only cuts. Does this effectively raise the crossover frequency or produce some overlap between the sub and speakers due to the filter slopes and the sub output without room correction being much higher than the speaker output?

 

[harkpabst]

Yes, it probably does. RC doesn't really care if it's the mains or the sub contributing to the SPL around the crossover frequency. In that setup the sub is probably a lot louder than the mains at fX (before correction). RC will be taming both, so effectively the sub plays higher up in the end.

You should be able to measure the sub and the mains separately. The latter most easily by simply switching off the sub. The former would require temporarily breaking the speaker connection. This won't do no harm for just one quick sweep.

 

[slartibartfast]

Yes, it probably does. RC doesn't really care if it's the mains or the sub contributing to the SPL around the crossover frequency. In that setup the sub is probably a lot louder than the mains at fX (before correction). RC will be taming both, so effectively the sub plays higher up in the end.

You should be able to measure the sub and the mains separately. The latter most easily by simply switching off the sub. The former would require temporarily breaking the speaker connection. This won't do no harm for just one quick sweep.

When I was doing the latency tests I noticed the frequency response was a lot more uneven at the bottom end than I expected it to be. At first I suspected it was because I was using REW with my DIY calibration file but it was actually because for that test I placed the microphone on the arm of the sofa and the room correction was done with the microphone on the back of the sofa, so a bit higher and further from the speakers. The magenta lines show the difference. The first two vertical lines are 31Hz and 63Hz and the second image shows a 20dB drop between them.

 

[harkpabst]

When I was doing the latency tests I noticed the frequency response was a lot more uneven at the bottom end than I expected it to be. At first I suspected it was because I was using REW with my DIY calibration file but it was actually because for that test I placed the microphone on the arm of the sofa and the room correction was done with the microphone on the back of the sofa, so a bit higher and further from the speakers. The magenta lines show the difference. The first two vertical lines are 31Hz and 63Hz and the second image shows a 20dB drop between them.

View attachment 13645
View attachment 13646

The resolution of the vertical axis is pretty low here. Stretching it to 5 dB (instead of 15 dB) divisions would be helpful. But it still makes the difference stand out.

What you see is probably just what really happens in those two different mic positions. Did you use any stand to elevate the mic from its boundary surface?

This might also be seen as a reality check what makes sense to correct for and what doesn't. Move the mic (or your head) around just a couple of inches and measurements may look vastly different. Just cross your arms behind your head while listening and notice how much different the sound is. Pull your auricles to any direction or press their edges flat to your scull ... everything changes dramatically ... but it's pretty impossible to tell which is righ, it's just different.

It's still common sense that the mic should have some distance from the surrounding surfaces, so some kind of mic stand made from thin tubing (no like of books) is highly recommended.

 

[MrA]

If WiiM implement this correctly, time alignment should be as simple as their room correction PEQ process....

For a single 'sweet spot'
Place the phone/mic representing a point 'between the listeners ears' and run the appropriate sweeps - ideally one for the main speakers and a second for the sub(s). This will then provide correct time delays and permit an appropriate adjustment - in almost all circumstances the time delay will apply to the main speakers.

For a wider range area
As with the above but several sweeps for each of the main speakers and sub(s) in different positions. These will then provide an 'average' adjustment where all would be 'good enough' but none quite as accurate as the first arrangement.

I have always preferred (and used) the first of these where only the listening position gives a totally believable soundstage and where frequency response is without any nulls or peaks. However, as I have three listening chairs adjacent to each other, the other two also sound very good but there are also other areas (especially at the door behind the chairs) where bass is overpowering.... irrelevant for listening of course.

(I use an Earthworks E23 on a mic stand with the mic just ahead of the chair headrest - chair slightly reclined so the measuring point is in the 'correct' place. Neither WiiM PEQ nor time correction are necessary as I already have all the correct perimeters and adjustments).