WiiM Sound and Sound Lite - my impressions and measurements

[dominikz]

This will be another of my huge posts - sorry about that in advance!

NOTE: I've received the WiiM Sound (WS) and Sound Lite (WSL) free of charge from WiiM for testing purposes. The opinions shared are my own, however, and have not been moderated by WiiM in any way.

WiiM Sound - how does it sound?
This is the first speaker I received. When I got it, initially I just streamed some music without much analysis.
While I was pleasantly surprised by the bass extension of such a small speaker, I immediately noticed something was wrong with the overall tonality. Main issue for me is the brightness of the stock tuning, but midrange tonality also seemed a bit off to me as well. It didn't seem like something a simple treble control (high-shelf filter) could solve. There can also be some boominess in the bass, but that can be easily fixed with RoomFit.
To check my perception I did a few rounds of direct, level-matched comparisons with my Revel M16 speakers and with that confirmed the brightness and uneven tonality issue is definitely there. Creating a treble tone control with PEQ (link to instruction) can help to an extent, but not entirely because the issue is not linear (as we will see later).

I also added the WiiM Sound as a center channel with multichannel Dolby Digital content, with LR being assigned to Revel M16 and Sub to the SVS SB-1000 Classic (all driven from the WiiM Amp Ultra). For the Surround channels I temporarily assigned my Mini (so that I can set the Sound as Center), but then removed the Mini from the group. This gave me a 3.1 setup.
In this configuration the tonality difference between the Sound and my M16s was quite obvious, with the WiiM Sound being much brighter - and IMHO distractingly so. The fact that EQ doesn't work in this mode of operation makes this even more of a problem, IMO.
Another issue, but not specific to WiiM Sound, is that there's no automatic calibration of sound level and delay between channels in multichannel mode. I.e. you need to manually adjust the relative levels of WiiM devices until you are happy with the overall balance at your listening position. I'd personally much prefer a method that aligns the relative levels and delay in multichannel mode using the microphone connected to the phone (i.e. similar to RoomFit), while allowing manual fine-tuning. Perhaps something like this will come in the future (the announced "SurroundFit", perhaps?)

Lastly, the bass response from the speaker can very easily get distorted. This may not be as easily audible with music, but is very easily audible with pure tones. This is expected given the small woofer size, however - WiiM can't defy physics (at least not yet ).

In summary, WiiM Sound is OK sounding (especially for background music listening), but honestly I expected much better from WiiM given how great their streamers and streaming amps are.

WiiM Sound Lite - how does it sound (lite)?
Out-of-the-box tonality seems the same as the original WS to me - I perceive it as too bright, with a midrange scoop. Basically you can assume everything from the previous section is valid here as well.
However, now having two WiiM smart speakers I could test a few additional use-cases.

• Pairing the WS with WSL in a stereo pair helps hide some of their tonality sins. The added width and depth of stereo takes the focus away from tonality issues, which definitely improves the experience.
  • Still, adding PEQ on top to tame the excessive treble can improve it a bit.
  • However it is not without issues: in direct level-matched comparison (even with corrective EQ configured) the WS(L) pair is still brighter and not as smooth in the midrange as my Revel M16 pair.
  • Stereo soundstage seemed a bit undefined/unstable to me - it has an almost "phasey" sound to it (like some kind of phase cancellation thing is happening). This is very apparent when comparing side-by-side with my Revel speakers.
  • Grouping and channel allocation seems to persist after standby, which is nice.
  • There seems to be an issue with the follower device not following volume of the main speaker every time, especially if using presets to adjust volume. I've opened a support ticket on this.
• I've also tried the WS+WSL as surround L/R in a Dolby 5.1 configuration with the WAU driving a pair or Revel M16 as front L/R and a SVS SB-1000 Classic subwoofer. This worked fine and was actually quite fun to try.
  • But the current limitation where RoomFit/EQ are both inoperable with 5.1 content is a serious issue. Without bass correction (RoomFit) it easily gets too boomy, and without EQ the surrounds don't match the fronts on tonality. Until this functionality is added (SurroundFit?) I don't believe I'll actually use 5.1 much.
• Lastly, I tried "double stereo" playback by grouping WS [as L], WSL [as R] and WAU [as LR] in the same room, but opposite sides. This was far from being accurate (there's no sound stage to speak of), but it was fun to try.

What's up with the dual-tweeters?
Now isn't that an interesting question?
As we can see from the marketing pictures, the two tweeters are pointing approx. 45° to the sides:

The intention was I guess to:

• Make the sound dispersion wider
• Allow for 'true' stereo playback with a single speaker, i.e.:
  • Above the crossover frequency (2.4kHz):
    • The left tweeter plays the left channel only
    • The right tweeter plays the right channel only
  • Below the crossover frequency (2.4kHz):
    • The single woofer plays a sum of both left and right channels

This sounds nice in theory, but in practice it IMHO brings several challenges:

• Two tweeters mounted on the same speaker will cause a complex pattern of acoustic interference in the dispersion. We'll see that in the following sections. My assumption is that this is what causes the "phasey" sound when pairing two speakers in a stereo group.
• We can't really perceive a stereo soundstage if both tweeters are so close together - unless perhaps if our room dimensions are really lucky, and speaker is placed optimally to reflect the sound from each tweeter's adjoining side-wall exactly to the listening position.

NOTE: So with both WS and WSL speakers the dual tweeter operate as follows:

• If we play a file where left and right channels contain a different signal, each tweeter will play the respective channel signal.
  • This is what happens with regular stereo audio content on a single speaker, or when WS or WSL are configured with "LR" when added to a multi-device group:
    
• If we play a file where left and right channels contain the same signal, both tweeters will play the same signal.
  • This is what happens with mono audio content on a single speaker, or when WS/WSL are configured to each play a single channel when added to a multi-device group (e.g. "L", "R", "C", "SL" or "SR"):
    
• If we play a file where only the left channel contains a signal, while the right one is empty then only the left tweeter will play, while the right tweeter will remain silent.
  • This can currently only be achieved with test or manipulated audio files, by leaving one channel empty.

Now on to see some measurements...

Methodology and equipment

The measurements were made through the WS/WSL Aux In input (EQ, RoomFit and Dynamic Bass were all disabled).

Measurement equipment:

• Cross-Spectrum Labs calibrated Dayton Audio EMM-6 measurement microphone
• RME BabyFace Silver Edition audio interface
• miniDSP UMIK-1 calibrated measurement microphone
• DIY loudspeaker measurement turntable stand
• Room EQ Wizard (REW) and VituixCAD software

High-frequency anechoic response was measured at 50cm distance and about 1,2m height, resulting in ~5,5ms reflection-free window.
This means that after temporal gating resolution and accuracy above 1kHz is high, but drops to about 1/3 octave at ~550Hz and lower below that.
The response was also measured at 1m distance, and produced the same result - which means that the drivers integrate well at 50cm and that responses at this distance are representative of device performance.

Low-frequency anechoic response was measured with three separate methods:

• Single microphone nearfield woofer measurement - done for both WS and WSL
• Two microphone mode-compensation method (explained here) - done for WS only
• (Almost) free-field response measurement - done for WSL only

All three methods produced equivalent woofer responses, increasing the confidence in the results. There appears to be no bass reflex port in either WS or WSL.

WiiM Sound (WS) - on-axis response (left and right tweeters play the same signal):

WiiM Sound (WS) - on-axis response (only the left tweeter plays, right tweeter is silent):

WiiM Sound Lite (WSL) - on-axis response (left and right tweeters play the same signal):

WiiM Sound Lite (WSL) - on-axis response (only the left tweeter plays, right tweeter is silent):


We can see that both WS and WSL give a very similar response, which is unsurprising given the identical architecture.
There is a peak in bass response around 70Hz, followed by a dip around 130Hz. The response above that is kept within +/- 2dB until about 12kHz, which seems pretty OK at first glance. Above 12kHz the response is chewed-up, likely due to some kind of acoustical interference caused by the physical design of the device.

But the really interesting part is the horizontal dispersion/directivity - here only shown for the WS, but the same applies to WSL.

WiiM Sound (WS) - horizontal directivity (left and right tweeters play the same signal):

WiiM Sound (WS) - horizontal directivity (only the left tweeter plays, right tweeter is silent):


While directivity is indeed pretty wide with both tweeters playing, it is riddled with cancellations when we move off-axis - unfortunately a natural product of this kind of acoustic design.
Here's how that looks if we compare responses between 0°-30°:

With only one tweeter playing, there is no cancellation, but there is directivity discontinuity around the crossover, at least in part due to physical misalignment of the tweeter and woofer. However the responses from 0°-30° are much better aligned to each other, because now there's no interference with the other tweeter:


So where does the brightness come from?
By now I hope we could see that the brightness issue is caused by the dual-tweeter design, and the resulting directivity mismatch between the single woofer and the tweeters. Basically to get a flat on-axis response (which is actually 45° off-axis to either woofer), both tweeter level need to be pretty high, which generates a lot of off-axis high frequency energy. Combine that with the cancellations caused by dual-tweeters, and I'd say that explains a lot of the issues I'm hearing.

Conclusion
As I mentioned at the beginning, these are not bad speakers - it is just that I expected more. If you pair them and apply some EQ, they can sound pretty good. I wouldn't replace my main speakers with them, however!

That's it for now! Hope some will find it useful!

 

[Roedelheim]

This will be another of my huge posts - sorry about that in advance!

NOTE: I've received the WiiM Sound (WS) and Sound Lite (WSL) free of charge from WiiM for testing purposes. The opinions shared are my own, however, and have not been moderated by WiiM in any way.

WiiM Sound - how does it sound?
This is the first speaker I received. When I got it, initially I just streamed some music without much analysis.
While I was pleasantly surprised by the bass extension of such a small speaker, I immediately noticed something was wrong with the overall tonality. Main issue for me is the brightness of the stock tuning, but midrange tonality also seemed a bit off to me as well. It didn't seem like something a simple treble control (high-shelf filter) could solve. There can also be some boominess in the bass, but that can be easily fixed with RoomFit.
To check my perception I did a few rounds of direct, level-matched comparisons with my Revel M16 speakers and with that confirmed the brightness and uneven tonality issue is definitely there. Creating a treble tone control with PEQ (link to instruction) can help to an extent, but not entirely because the issue is not linear (as we will see later).

I also added the WiiM Sound as a center channel with multichannel Dolby Digital content, with LR being assigned to Revel M16 and Sub to the SVS SB-1000 Classic (all driven from the WiiM Amp Ultra). For the Surround channels I temporarily assigned my Mini (so that I can set the Sound as Center), but then removed the Mini from the group. This gave me a 3.1 setup.
In this configuration the tonality difference between the Sound and my M16s was quite obvious, with the WiiM Sound being much brighter - and IMHO distractingly so. The fact that EQ doesn't work in this mode of operation makes this even more of a problem, IMO.
Another issue, but not specific to WiiM Sound, is that there's no automatic calibration of sound level and delay between channels in multichannel mode. I.e. you need to manually adjust the relative levels of WiiM devices until you are happy with the overall balance at your listening position. I'd personally much prefer a method that aligns the relative levels and delay in multichannel mode using the microphone connected to the phone (i.e. similar to RoomFit), while allowing manual fine-tuning. Perhaps something like this will come in the future (the announced "SurroundFit", perhaps?)

Lastly, the bass response from the speaker can very easily get distorted. This may not be as easily audible with music, but is very easily audible with pure tones. This is expected given the small woofer size, however - WiiM can't defy physics (at least not yet ).

In summary, WiiM Sound is OK sounding (especially for background music listening), but honestly I expected much better from WiiM given how great their streamers and streaming amps are.

WiiM Sound Lite - how does it sound (lite)?
Out-of-the-box tonality seems the same as the original WS to me - I perceive it as too bright, with a midrange scoop. Basically you can assume everything from the previous section is valid here as well.
However, now having two WiiM smart speakers I could test a few additional use-cases.

• Pairing the WS with WSL in a stereo pair helps hide some of their tonality sins. The added width and depth of stereo takes the focus away from tonality issues, which definitely improves the experience.
  • Still, adding PEQ on top to tame the excessive treble can improve it a bit.
  • However it is not without issues: in direct level-matched comparison (even with corrective EQ configured) the WS(L) pair is still brighter and not as smooth in the midrange as my Revel M16 pair.
  • Stereo soundstage seemed a bit undefined/unstable to me - it has an almost "phasey" sound to it (like some kind of phase cancellation thing is happening). This is very apparent when comparing side-by-side with my Revel speakers.
  • Grouping and channel allocation seems to persist after standby, which is nice.
  • There seems to be an issue with the follower device not following volume of the main speaker every time, especially if using presets to adjust volume. I've opened a support ticket on this.
• I've also tried the WS+WSL as surround L/R in a Dolby 5.1 configuration with the WAU driving a pair or Revel M16 as front L/R and a SVS SB-1000 Classic subwoofer. This worked fine and was actually quite fun to try.
  • But the current limitation where RoomFit/EQ are both inoperable with 5.1 content is a serious issue. Without bass correction (RoomFit) it easily gets too boomy, and without EQ the surrounds don't match the fronts on tonality. Until this functionality is added (SurroundFit?) I don't believe I'll actually use 5.1 much.
• Lastly, I tried "double stereo" playback by grouping WS [as L], WSL [as R] and WAU [as LR] in the same room, but opposite sides. This was far from being accurate (there's no sound stage to speak of), but it was fun to try.

What's up with the dual-tweeters?
Now isn't that an interesting question?
As we can see from the marketing pictures, the two tweeters are pointing approx. 45° to the sides:

The intention was I guess to:

• Make the sound dispersion wider
• Allow for 'true' stereo playback with a single speaker, i.e.:
  • Above the crossover frequency (2.4kHz):
    • The left tweeter plays the left channel only
    • The right tweeter plays the right channel only
  • Below the crossover frequency (2.4kHz):
    • The single woofer plays a sum of both left and right channels

This sounds nice in theory, but in practice it IMHO brings several challenges:

• Two tweeters mounted on the same speaker will cause a complex pattern of acoustic interference in the dispersion. We'll see that in the following sections. My assumption is that this is what causes the "phasey" sound when pairing two speakers in a stereo group.
• We can't really perceive a stereo soundstage if both tweeters are so close together - unless perhaps if our room dimensions are really lucky, and speaker is placed optimally to reflect the sound from each tweeter's adjoining side-wall exactly to the listening position.

NOTE: So with both WS and WSL speakers the dual tweeter operate as follows:

• If we play a file where left and right channels contain a different signal, each tweeter will play the respective channel signal.
  • This is what happens with regular stereo audio content on a single speaker, or when WS or WSL are configured with "LR" when added to a multi-device group:
    View attachment 35251
• If we play a file where left and right channels contain the same signal, both tweeters will play the same signal.
  • This is what happens with mono audio content on a single speaker, or when WS/WSL are configured to each play a single channel when added to a multi-device group (e.g. "L", "R", "C", "SL" or "SR"):
    View attachment 35253
• If we play a file where only the left channel contains a signal, while the right one is empty then only the left tweeter will play, while the right tweeter will remain silent.
  • This can currently only be achieved with test or manipulated audio files, by leaving one channel empty.

Now on to see some measurements...

Methodology and equipment

The measurements were made through the WS/WSL Aux In input (EQ, RoomFit and Dynamic Bass were all disabled).

Measurement equipment:

• Cross-Spectrum Labs calibrated Dayton Audio EMM-6 measurement microphone
• RME BabyFace Silver Edition audio interface
• miniDSP UMIK-1 calibrated measurement microphone
• DIY loudspeaker measurement turntable stand
• Room EQ Wizard (REW) and VituixCAD software

High-frequency anechoic response was measured at 50cm distance and about 1,2m height, resulting in ~5,5ms reflection-free window.
This means that after temporal gating resolution and accuracy above 1kHz is high, but drops to about 1/3 octave at ~550Hz and lower below that.
The response was also measured at 1m distance, and produced the same result - which means that the drivers integrate well at 50cm and that responses at this distance are representative of device performance.

Low-frequency anechoic response was measured with three separate methods:

• Single microphone nearfield woofer measurement - done for both WS and WSL
• Two microphone mode-compensation method (explained here) - done for WS only
• (Almost) free-field response measurement - done for WSL only

All three methods produced equivalent woofer responses, increasing the confidence in the results. There appears to be no bass reflex port in either WS or WSL.

WiiM Sound (WS) - on-axis response (left and right tweeters play the same signal):
View attachment 35254
WiiM Sound (WS) - on-axis response (only the left tweeter plays, right tweeter is silent):
View attachment 35256
WiiM Sound Lite (WSL) - on-axis response (left and right tweeters play the same signal):
View attachment 35255
WiiM Sound Lite (WSL) - on-axis response (only the left tweeter plays, right tweeter is silent):
View attachment 35257

We can see that both WS and WSL give a very similar response, which is unsurprising given the identical architecture.
There is a peak in bass response around 70Hz, followed by a dip around 130Hz. The response above that is kept within +/- 2dB until about 12kHz, which seems pretty OK at first glance. Above 12kHz the response is chewed-up, likely due to some kind of acoustical interference caused by the physical design of the device.

But the really interesting part is the horizontal dispersion/directivity - here only shown for the WS, but the same applies to WSL.

WiiM Sound (WS) - horizontal directivity (left and right tweeters play the same signal):
View attachment 35259
WiiM Sound (WS) - horizontal directivity (only the left tweeter plays, right tweeter is silent):
View attachment 35260

While directivity is indeed pretty wide with both tweeters playing, it is riddled with cancellations when we move off-axis - unfortunately a natural product of this kind of acoustic design.
Here's how that looks if we compare responses between 0°-30°:
View attachment 35261
With only one tweeter playing, there is no cancellation, but there is directivity discontinuity around the crossover, at least in part due to physical misalignment of the tweeter and woofer. However the responses from 0°-30° are much better aligned to each other, because now there's no interference with the other tweeter:
View attachment 35262

So where does the brightness come from?
By now I hope we could see that the brightness issue is caused by the dual-tweeter design, and the resulting directivity mismatch between the single woofer and the tweeters. Basically to get a flat on-axis response (which is actually 45° off-axis to either woofer), both tweeter level need to be pretty high, which generates a lot of off-axis high frequency energy. Combine that with the cancellations caused by dual-tweeters, and I'd say that explains a lot of the issues I'm hearing.

Conclusion
As I mentioned at the beginning, these are not bad speakers - it is just that I expected more. If you pair them and apply some EQ, they can sound pretty good. I wouldn't replace my main speakers with them, however!

That's it for now! Hope some will find it useful!

Three big for the work you did. I do not understand the most, but I can imagine the hours you needed.

 

[GeeTee]

unbias, comprehensive and analytical review .. 'WHAT HiFi' is calling you ..

 

[GeeTee]

@dominikz .. your overall impression is probably the main reason i have not added wiim sound or the lite to my system. I felt some of your findings were reflected in various forum comments i have seen.
Excellent review
** and well done to WIIM for letting you give an unbiased review

 

[Roedelheim]

unbias, comprehensive and analytical review .. 'WHAT HiFi' is calling you ..

View attachment 35269

But at the end of the day the result is similar. I do not have the technical skills like Dominik, but when we looked for a new kitchen radio, both my wife and I came to the by-ear conclusion the Lite is mediocre at best. The clear winner came from a competitor.

 

[Wiimer]

This will be another of my huge posts - sorry about that in advance!

NOTE: I've received the WiiM Sound (WS) and Sound Lite (WSL) free of charge from WiiM for testing purposes. The opinions shared are my own, however, and have not been moderated by WiiM in any way.

WiiM Sound - how does it sound?
This is the first speaker I received. When I got it, initially I just streamed some music without much analysis.
While I was pleasantly surprised by the bass extension of such a small speaker, I immediately noticed something was wrong with the overall tonality. Main issue for me is the brightness of the stock tuning, but midrange tonality also seemed a bit off to me as well. It didn't seem like something a simple treble control (high-shelf filter) could solve. There can also be some boominess in the bass, but that can be easily fixed with RoomFit.
To check my perception I did a few rounds of direct, level-matched comparisons with my Revel M16 speakers and with that confirmed the brightness and uneven tonality issue is definitely there. Creating a treble tone control with PEQ (link to instruction) can help to an extent, but not entirely because the issue is not linear (as we will see later).

I also added the WiiM Sound as a center channel with multichannel Dolby Digital content, with LR being assigned to Revel M16 and Sub to the SVS SB-1000 Classic (all driven from the WiiM Amp Ultra). For the Surround channels I temporarily assigned my Mini (so that I can set the Sound as Center), but then removed the Mini from the group. This gave me a 3.1 setup.
In this configuration the tonality difference between the Sound and my M16s was quite obvious, with the WiiM Sound being much brighter - and IMHO distractingly so. The fact that EQ doesn't work in this mode of operation makes this even more of a problem, IMO.
Another issue, but not specific to WiiM Sound, is that there's no automatic calibration of sound level and delay between channels in multichannel mode. I.e. you need to manually adjust the relative levels of WiiM devices until you are happy with the overall balance at your listening position. I'd personally much prefer a method that aligns the relative levels and delay in multichannel mode using the microphone connected to the phone (i.e. similar to RoomFit), while allowing manual fine-tuning. Perhaps something like this will come in the future (the announced "SurroundFit", perhaps?)

Lastly, the bass response from the speaker can very easily get distorted. This may not be as easily audible with music, but is very easily audible with pure tones. This is expected given the small woofer size, however - WiiM can't defy physics (at least not yet ).

In summary, WiiM Sound is OK sounding (especially for background music listening), but honestly I expected much better from WiiM given how great their streamers and streaming amps are.

WiiM Sound Lite - how does it sound (lite)?
Out-of-the-box tonality seems the same as the original WS to me - I perceive it as too bright, with a midrange scoop. Basically you can assume everything from the previous section is valid here as well.
However, now having two WiiM smart speakers I could test a few additional use-cases.

• Pairing the WS with WSL in a stereo pair helps hide some of their tonality sins. The added width and depth of stereo takes the focus away from tonality issues, which definitely improves the experience.
  • Still, adding PEQ on top to tame the excessive treble can improve it a bit.
  • However it is not without issues: in direct level-matched comparison (even with corrective EQ configured) the WS(L) pair is still brighter and not as smooth in the midrange as my Revel M16 pair.
  • Stereo soundstage seemed a bit undefined/unstable to me - it has an almost "phasey" sound to it (like some kind of phase cancellation thing is happening). This is very apparent when comparing side-by-side with my Revel speakers.
  • Grouping and channel allocation seems to persist after standby, which is nice.
  • There seems to be an issue with the follower device not following volume of the main speaker every time, especially if using presets to adjust volume. I've opened a support ticket on this.
• I've also tried the WS+WSL as surround L/R in a Dolby 5.1 configuration with the WAU driving a pair or Revel M16 as front L/R and a SVS SB-1000 Classic subwoofer. This worked fine and was actually quite fun to try.
  • But the current limitation where RoomFit/EQ are both inoperable with 5.1 content is a serious issue. Without bass correction (RoomFit) it easily gets too boomy, and without EQ the surrounds don't match the fronts on tonality. Until this functionality is added (SurroundFit?) I don't believe I'll actually use 5.1 much.
• Lastly, I tried "double stereo" playback by grouping WS [as L], WSL [as R] and WAU [as LR] in the same room, but opposite sides. This was far from being accurate (there's no sound stage to speak of), but it was fun to try.

What's up with the dual-tweeters?
Now isn't that an interesting question?
As we can see from the marketing pictures, the two tweeters are pointing approx. 45° to the sides:

The intention was I guess to:

• Make the sound dispersion wider
• Allow for 'true' stereo playback with a single speaker, i.e.:
  • Above the crossover frequency (2.4kHz):
    • The left tweeter plays the left channel only
    • The right tweeter plays the right channel only
  • Below the crossover frequency (2.4kHz):
    • The single woofer plays a sum of both left and right channels

This sounds nice in theory, but in practice it IMHO brings several challenges:

• Two tweeters mounted on the same speaker will cause a complex pattern of acoustic interference in the dispersion. We'll see that in the following sections. My assumption is that this is what causes the "phasey" sound when pairing two speakers in a stereo group.
• We can't really perceive a stereo soundstage if both tweeters are so close together - unless perhaps if our room dimensions are really lucky, and speaker is placed optimally to reflect the sound from each tweeter's adjoining side-wall exactly to the listening position.

NOTE: So with both WS and WSL speakers the dual tweeter operate as follows:

• If we play a file where left and right channels contain a different signal, each tweeter will play the respective channel signal.
  • This is what happens with regular stereo audio content on a single speaker, or when WS or WSL are configured with "LR" when added to a multi-device group:
    View attachment 35251
• If we play a file where left and right channels contain the same signal, both tweeters will play the same signal.
  • This is what happens with mono audio content on a single speaker, or when WS/WSL are configured to each play a single channel when added to a multi-device group (e.g. "L", "R", "C", "SL" or "SR"):
    View attachment 35253
• If we play a file where only the left channel contains a signal, while the right one is empty then only the left tweeter will play, while the right tweeter will remain silent.
  • This can currently only be achieved with test or manipulated audio files, by leaving one channel empty.

Now on to see some measurements...

Methodology and equipment

The measurements were made through the WS/WSL Aux In input (EQ, RoomFit and Dynamic Bass were all disabled).

Measurement equipment:

• Cross-Spectrum Labs calibrated Dayton Audio EMM-6 measurement microphone
• RME BabyFace Silver Edition audio interface
• miniDSP UMIK-1 calibrated measurement microphone
• DIY loudspeaker measurement turntable stand
• Room EQ Wizard (REW) and VituixCAD software

High-frequency anechoic response was measured at 50cm distance and about 1,2m height, resulting in ~5,5ms reflection-free window.
This means that after temporal gating resolution and accuracy above 1kHz is high, but drops to about 1/3 octave at ~550Hz and lower below that.
The response was also measured at 1m distance, and produced the same result - which means that the drivers integrate well at 50cm and that responses at this distance are representative of device performance.

Low-frequency anechoic response was measured with three separate methods:

• Single microphone nearfield woofer measurement - done for both WS and WSL
• Two microphone mode-compensation method (explained here) - done for WS only
• (Almost) free-field response measurement - done for WSL only

All three methods produced equivalent woofer responses, increasing the confidence in the results. There appears to be no bass reflex port in either WS or WSL.

WiiM Sound (WS) - on-axis response (left and right tweeters play the same signal):
View attachment 35254
WiiM Sound (WS) - on-axis response (only the left tweeter plays, right tweeter is silent):
View attachment 35256
WiiM Sound Lite (WSL) - on-axis response (left and right tweeters play the same signal):
View attachment 35255
WiiM Sound Lite (WSL) - on-axis response (only the left tweeter plays, right tweeter is silent):
View attachment 35257

We can see that both WS and WSL give a very similar response, which is unsurprising given the identical architecture.
There is a peak in bass response around 70Hz, followed by a dip around 130Hz. The response above that is kept within +/- 2dB until about 12kHz, which seems pretty OK at first glance. Above 12kHz the response is chewed-up, likely due to some kind of acoustical interference caused by the physical design of the device.

But the really interesting part is the horizontal dispersion/directivity - here only shown for the WS, but the same applies to WSL.

WiiM Sound (WS) - horizontal directivity (left and right tweeters play the same signal):
View attachment 35259
WiiM Sound (WS) - horizontal directivity (only the left tweeter plays, right tweeter is silent):
View attachment 35260

While directivity is indeed pretty wide with both tweeters playing, it is riddled with cancellations when we move off-axis - unfortunately a natural product of this kind of acoustic design.
Here's how that looks if we compare responses between 0°-30°:
View attachment 35261
With only one tweeter playing, there is no cancellation, but there is directivity discontinuity around the crossover, at least in part due to physical misalignment of the tweeter and woofer. However the responses from 0°-30° are much better aligned to each other, because now there's no interference with the other tweeter:
View attachment 35262

So where does the brightness come from?
By now I hope we could see that the brightness issue is caused by the dual-tweeter design, and the resulting directivity mismatch between the single woofer and the tweeters. Basically to get a flat on-axis response (which is actually 45° off-axis to either woofer), both tweeter level need to be pretty high, which generates a lot of off-axis high frequency energy. Combine that with the cancellations caused by dual-tweeters, and I'd say that explains a lot of the issues I'm hearing.

Conclusion
As I mentioned at the beginning, these are not bad speakers - it is just that I expected more. If you pair them and apply some EQ, they can sound pretty good. I wouldn't replace my main speakers with them, however!

That's it for now! Hope some will find it useful!

As always, thank you for the excellent measurements and review.

When comparing the on-axis response, it appears that the volume of the WS is about 1 dB higher than that of the WSL in your case. In my case, it’s the opposite. Does this mean it’s just a slight variation between units?

 

[dominikz]

When comparing the on-axis response, it appears that the volume of the WS is about 1 dB higher than that of the WSL in your case. In my case, it’s the opposite. Does this mean it’s just a slight variation between units?

I measured the WS and the WSL on completely separate occasions, so I can't guarantee the levels were set identically both times. The level mismatch could also be because of that.

But I can definitely test for this specifically and report back!

 

[Bryan]

Conclusion
As I mentioned at the beginning, these are not bad speakers - it is just that I expected more. If you pair them and apply some EQ, they can sound pretty good. I wouldn't replace my main speakers with them, however!

Your conclusion reflects the same felling I have about my WiiM Sound. I will say though over time, as a stand alone speaker I've been tweaking my WiiM Sound settings and it's become more enjoyable for listening purposes. However, I feel the journey has taken longer than I expected for a speaker like this.

I sort of feel the WiiM Sound is still trying to define it's best use-case. For example, the other day I also put it on my deck outside...and it probably is one of the better speakers I've had outdoors despite it not being intended to be a portable speaker. The speaker "filled the deck" with sound in a much more positive way than I had expected. If this speaker wasn't such an expensive item, I probably would keep it permanently outdoors during the warmer months.

 

[dominikz]

Conclusion
As I mentioned at the beginning, these are not bad speakers - it is just that I expected more. If you pair them and apply some EQ, they can sound pretty good. I wouldn't replace my main speakers with them, however!

Perhaps it makes sense to give some more context here.

First, I should say that for many use cases WS and WSL would probably work great for a lot of people.

The feature set, ease of use, form-factor and design are all very nice, and both seamlessly integrate into the existing WiiM ecosystem.
While they are not the best-sounding smart speakers available out of the box, sound quality is not bad either - especially if we compare them to the usual small BT and smart speakers many people use. We can also tune them as we please with EQ and RoomFit.

Second, I compared WS and WSL against well-designed, full-size (6" woofer) bookshelf speakers paired with a 12" sub - which some might say is not really a fair comparison.
While that is partially true, there's also really no technical reason why a small speaker couldn't sound great - it wouldn't be able to play very deep or very loud, but at lower playback levels tonality could still be on par with a bigger speaker.

As always, the resulting design of any device is a compromise of different objectives, some of which have to be given precedence over others.

A priority in WS and WSL design appears to have been to make a single unit able to play stereo without downmixing to mono, and to have wide dispersion. Hence the (IMHO unfortunate) angled dual-tweeter design. A secondary objective seems to have been to adjust for a flat on-axis response.
When these two objectives are combined, we get a speaker which radiates too much treble energy off-axis.

Which brings us to...

So where does the brightness come from?
By now I hope we could see that the brightness issue is caused by the dual-tweeter design, and the resulting directivity mismatch between the single woofer and the tweeters. Basically to get a flat on-axis response (which is actually 45° off-axis to either woofer), both tweeter level need to be pretty high, which generates a lot of off-axis high frequency energy. Combine that with the cancellations caused by dual-tweeters, and I'd say that explains a lot of the issues I'm hearing.

Let me try to illustrate with another diagram; if we move from 0° (on-axis) to 90° (off-axis) on WS or WSL, we start to get less energy in the upper midrange, but more and more energy in the treble. This is because when we move away from the front of the device we also move towards the tweeters, which are aimed at ~45° angle to the side, and so naturally radiate more energy there (responses at 0°, 30°, 60° and 90° horizontal angles are compared below):


Compare that now to the behaviour of a really well-designed traditional loudspeaker (front-firing, single-tweeter), in an equivalent measurement:

As we can see, while the on-axis response target is also flat, the off-axis responses show a progressive reduction in treble energy. Actually, with the Neumann above 10kHz the response at 90° angle is almost 20dB down (subjectively roughly a quarter as loud) compared to on-axis
Now look at the previous diagram and notice that with WS/WSL the response above 10kHz at 90° angle is actually higher than on-axis!

The result is that WS/WSL sound bright, while KH120A sounds balanced - even though they were both designed with the same flat on-axis response target in mind! The difference is simply caused by the different sound dispersion characteristics of the two designs.

IMHO if WS and WSL were designed as 'mono' speakers (like the KH120A example above), with just a single woofer and single tweeter, they would have likely sounded better.
This would then be a traditional two-way front-firing design, where horizontal directivity could be better controlled (with some care in development), there would be no excess treble energy, and off-axis interference caused by the two tweeters would have been completely avoided.
I'd personally prefer such a design - it would be ideal for use in multi-speaker groups (stereo or multichannel)!

The dual-tweeter design WiiM opted for with WS/WSL (to have "stereo playback" with a single unit) makes it much more complicated to make a neutral-sounding speaker. IMO it is obvious that in this case a simple "flat on-axis" doesn't work, but at least to my knowledge there are no industry guidelines on what the target *should* be in such designs. This means significant research effort would be required to get it right.

So in that sense I can understand that WiiM had a pretty difficult task on their hands here.

Lastly, WiiM make streamers and streaming amps that can compete with devices at any price on both sound and features, not to mention the amazing support.
So my expectations for the WS and WSL were really very high. But loudspeaker design brings new aspects to the table, so perhaps it is not so surprising that their first attempt is not perfect in every way.

In any case I'm excited to see what they come up with in the future!

 

[slartibartfast]

Perhaps it makes sense to give some more context here.

First, I should say that for many use cases WS and WSL would probably work great for a lot of people.

The feature set, ease of use, form-factor and design are all very nice, and both seamlessly integrate into the existing WiiM ecosystem.
While they are not the best-sounding smart speakers available out of the box, sound quality is not bad either - especially if we compare them to the usual small BT and smart speakers many people use. We can also tune them as we please with EQ and RoomFit.

Second, I compared WS and WSL against well-designed, full-size (6" woofer) bookshelf speakers paired with a 12" sub - which some might say is not really a fair comparison.
While that is partially true, there's also really no technical reason why a small speaker couldn't sound great - it wouldn't be able to play very deep or very loud, but at lower playback levels tonality could still be on par with a bigger speaker.

As always, the resulting design of any device is a compromise of different objectives, some of which have to be given precedence over others.

A priority in WS and WSL design appears to have been to make a single unit able to play stereo without downmixing to mono, and to have wide dispersion. Hence the (IMHO unfortunate) angled dual-tweeter design. A secondary objective seems to have been to adjust for a flat on-axis response.
When these two objectives are combined, we get a speaker which radiates too much treble energy off-axis.

Which brings us to...

Let me try to illustrate with another diagram; if we move from 0° (on-axis) to 90° (off-axis) on WS or WSL, we start to get less energy in the upper midrange, but more and more energy in the treble. This is because when we move away from the front of the device we also move towards the tweeters, which are aimed at ~45° angle to the side, and so naturally radiate more energy there (responses at 0°, 30°, 60° and 90° horizontal angles are compared below):
View attachment 35281

Compare that now to the behaviour of a really well-designed traditional loudspeaker (front-firing, single-tweeter), in an equivalent measurement:
View attachment 35282
As we can see, while the on-axis response target is also flat, the off-axis responses show a progressive reduction in treble energy. Actually, with the Neumann above 10kHz the response at 90° angle is almost 20dB down (subjectively roughly a quarter as loud) compared to on-axis
Now look at the previous diagram and notice that with WS/WSL the response above 10kHz at 90° angle is actually higher than on-axis!

The result is that WS/WSL sound bright, while KH120A sounds balanced - even though they were both designed with the same flat on-axis response target in mind! The difference is simply caused by the different sound dispersion characteristics of the two designs.

IMHO if WS and WSL were designed as 'mono' speakers (like the KH120A example above), with just a single woofer and single tweeter, they would have likely sounded better.
This would then be a traditional two-way front-firing design, where horizontal directivity could be better controlled (with some care in development), there would be no excess treble energy, and off-axis interference caused by the two tweeters would have been completely avoided.
I'd personally prefer such a design - it would be ideal for use in multi-speaker groups (stereo or multichannel)!

The dual-tweeter design WiiM opted for with WS/WSL (to have "stereo playback" with a single unit) makes it much more complicated to make a neutral-sounding speaker. IMO it is obvious that in this case a simple "flat on-axis" doesn't work, but at least to my knowledge there are no industry guidelines on what the target *should* be in such designs. This means significant research effort would be required to get it right.

So in that sense I can understand that WiiM had a pretty difficult task on their hands here.

Lastly, WiiM make streamers and streaming amps that can compete with devices at any price on both sound and features, not to mention the amazing support.
So my expectations for the WS and WSL were really very high. But loudspeaker design brings new aspects to the table, so perhaps it is not so surprising that their first attempt is not perfect in every way.

In any case I'm excited to see what they come up with in the future!

Do any other speakers use the same twin tweeter arrangement?

 

[dominikz]

Do any other speakers use the same twin tweeter arrangement?

Not sure if exactly the same, but some other smart speakers use a very similar arrangement - e.g. Sonos Era 100 looks very similar (you can find the 'exploded' view on the official page).
But reportedly Sonos has a huge audio research budget, several anechoic chambers, and years of experience designing such products... And as far as I know their research results are not publicly available.

EDIT: Seems Amazon Echo 4th gen and Harman Kardon Onyx Studio 7 use a similar architecture as well. There are probably others.

 

[Roedelheim]

Our decision at the end was the Era300. We do not need a lot of the features. Apple Music, Qobuz and some local radio stations. No idea how it looks inside and how it measures, but the overall and quite adult sound let us make this decision.

 

[dominikz]

Our decision at the end was the Era300. We do not need a lot of the features. Apple Music, Qobuz and some local radio stations. No idea how it looks inside and how it measures, but the overall and the quite adult sound let us make this decision.

Haven't heard the Era 300 myself, but I wouldn't be surprised at all if it sounds great!
I have the small Roam 2 and it is IMHO pretty good for such a small and portable speaker!

Sonos has strong audio engineering capability, despite the (relatively) recent problems with their app.

 

[slartibartfast]

Haven't heard the Era 300 myself, but I wouldn't be surprised at all if it sounds great!
I have the small Roam 2 and it is IMHO pretty good for such a small and portable speaker!

Sonos has strong audio engineering capability, despite the (relatively) recent problems with their app.

It makes you wonder why WiiM came up with their own speaker design when they already had a partnership with Audio Pro, another experienced speaker designer.

 

[dominikz]

As always, thank you for the excellent measurements and review.

When comparing the on-axis response, it appears that the volume of the WS is about 1 dB higher than that of the WSL in your case. In my case, it’s the opposite. Does this mean it’s just a slight variation between units?

I measured the WS and the WSL on completely separate occasions, so I can't guarantee the levels were set identically both times. The level mismatch could also be because of that.

But I can definitely test for this specifically and report back!

Here's a quick and dirty test.

For both of these tests WiiM Sound (WS) and Sound Lite (WSL) were paired in a stereo group; WS=left channel and WSL=right channel.

First test was with the microphone centered and fixed at ~2.75m from each speaker, measured with a separate sweep for each channel (speaker), with the other muted in WHA:

The second test was measured using the moving microphone method, moving randomly at around 2.5-2.3m distance from each speaker. Again each channel (speaker) was measured separately, with the other muted in WHA:


Neither test shows any obvious overall difference in relative level between my two units. Seems they are reasonably well matched.

Perhaps worth noting that these responses are with RoomFit (non-boost mode) applied below 400Hz, plus a high-shelf PEQ to reduce the highs a by a few dB.

 

[dominikz]

It makes you wonder why WiiM came up with their own speaker design when they already had a partnership with Audio Pro, another experienced speaker designer.

Maybe they wanted to have more control of the design? Who knows.
In any case I for one applaud them for the effort, and hope we see more of their smart speakers in the future!

 

[satheesht]

EDIT: Seems Amazon Echo 4th gen and Harman Kardon Onyx Studio 7 use a similar architecture as well. There are probably others.

I have a JBL authentics 300 which has front facing tweeter pair & single woofer and a downward facing passive radiator. It's a nice little package and for a long time i considered getting another one for stereo pairing. But finally went with a proper (WiiM) amp+speaker+sub instead. With a limited EQ and 4 tweeters and 2 passive radiators, it would have been a struggle if i had gone for stereo pairing I wonder.