Broken VU output amplitude after update

[OurBodilyFluids]

Is the VU-Output Amplitude completely broken for anyone else after the latest update?

I understand that it doesn’t serve a functional purpose beyond being visually appealing, but in previous versions, it loosely corresponded with volume levels. Now, the dial barely moves, even at high volumes!

Video of meter at 60% volume.


View attachment IMG_0569.mov

 

[hgo58]

Is the VU-Output Amplitude completely broken for anyone else after the latest update?

I understand that it doesn’t serve a functional purpose beyond being visually appealing, but in previous versions, it loosely corresponded with volume levels. Now, the dial barely moves, even at high volumes!

Video of meter at 60% volume.

View attachment 26693View attachment 26692View attachment 26691
View attachment 26694

Yes. The so called "fix" of the VU meter has degraded the functionality even more than it already was.

I really hope that @WiiM Team will find another group of people to take over and make a real functioning meter.

 

[OurBodilyFluids]

Yes. The so called "fix" of the VU meter has degraded the functionality even more than it already was.

I really hope that @WiiM Team will find another group of people to take over and make a real functioning meter.

Correct me if I’m wrong, but wouldn’t it be relatively straightforward for developers to patch the firmware with a proper VU algorithm that computes a true RMS value?

It’s not as though the Ultra’s hardware is limiting it to a cosmetic feature rather than a functional VU meter… @WiiM Team

 

[WiiM Support]

Hi Team,

We’re actively working on a fix for this issue. In the meantime, please select the Original Signal option for reliable performance. For Output Amplitude, we’ll adjust the minimum threshold from –25 dB to –60 dB in the next update, as the current value is too low. The VU algorithm is already calculating the true RMS value. Thanks for your patience—we’re committed to resolving this quickly and ensuring the best experience.

 

[hgo58]

Hi Team,

We’re actively working on a fix for this issue. In the meantime, please select the Original Signal option for reliable performance. For Output Amplitude, we’ll adjust the minimum threshold from –25 dB to –60 dB in the next update, as the current value is too low. The VU algorithm is already calculating the true RMS value. Thanks for your patience—we’re committed to resolving this quickly and ensuring the best experience.

True RMS is for analog measurement, so unless the VU meter is measuring the analog output of the DAC, that is the wrong value.

The RMS dB is a relatively value, while the digital dBFS is absolute and just a logarithmic representation of the sample volume value.

 

[Machiavelli]

firmly believe that an estimated 97% of the (silent) wiim users are happy with the existing gimmick of flouncing needles. gives them at least a bit of a yamaha, mcintosh or accuphase feeling. for serious use cases it is (censored by myself).

 

[hgo58]

firmly believe that an estimated 97% of the (silent) wiim users are happy with the existing gimmick of flouncing needles. gives them at least a bit of a yamaha, mcintosh or accuphase feeling.

Guess you are right. Although I don't think it's even looking nice. And hard to see from a distance.

for serious use cases it is (censored by myself).

So why not make it serious? It's not hard to do. And then the last 3% would be happy too.

 

[Machiavelli]

long live the 3% and of course you are absolutely right.

 

[hgo58]

If people want to keep the analog scale it's also not a problem. Both scales can be included and there are good conversion tables.



This is not exact mapping as the dB is relative but can be used nicely as long as the max is 0 dBFS (all bits are 1).

 

[OurBodilyFluids]

Hi Team,

We’re actively working on a fix for this issue. In the meantime, please select the Original Signal option for reliable performance. For Output Amplitude, we’ll adjust the minimum threshold from –25 dB to –60 dB in the next update, as the current value is too low. The VU algorithm is already calculating the true RMS value. Thanks for your patience—we’re committed to resolving this quickly and ensuring the best experience.

Thanks for the update!

 

[pjorgenunes]

My two cents:

Levels have been more than discussed here (I got nothing positive to add)

Agree it should be seen from 3-4 meters away. Maybe a wider needle or some colours that contrast more.

Why don’t make it nice?! I’m stucked with the black and white because it’s the only I can almost see from 3,5 meters away.

If possíble, why not a two horizontal bars of square LEDs (if colours could be cosumized the better). With the levels right and this kind ok “tape recorder” vumeter I would be happy and surely see it from distance.

 

[cc_rider]

@WiiM Team here’s a HAD project that attempts to get the meter ballistics correct.

Digital VU Meter with Analog Physics

To start off, thanks to JLCPCB for sponsoring this project. JLCPCB Prototype for $2 (Any Color): https://jlcpcb.com/DYE I've seen tons of digital VU meters that aspire to look like analog movements of days past but most of them just directly set the needle position to the measured value, making...

hackaday.io

 

[Angel]

The VU algorithm is already calculating the true RMS value.

Dear WiiM team, thank you very much for this information.
To be honest, I'm a little puzzled by the "true RMS" value.

Maybe you could comment on this more theoretical questions for our understanding:

When I send a digitally generated sine wave with 100% FS signal amplitude from my PC to the WiiM optical digital in, the Original Signal "VU meter" reads 0dB (needle pointing to 0). All gains EQ RoomFit off, and Output Volume set to 100%, so does the output VU meter.
This could and should be the reference for 0dBFS, either as peak amplitude meter (relative to FS) or true RMS (relative to FS) .

However, when I send a digitally generated SQUARE wave with 100% FS signal amplitude from my PC to the WiiM optical digital in, the "VU meter" reads 0dB (needle pointing to 0) again!

However, a 100% FS square wave has +3dB more RMS power than the 100% FS sine wave.

See also this Wikipedia article, as pointed out by hgo58:

dBFS - Wikipedia

en.wikipedia.org

"dBFS is defined in AES Standard AES17-1998,[13] IEC 61606,[14] and ITU-T Recs. P.381[15] and P.382,[16] such that the RMS value of a full-scale sine wave is designated 0 dB FS. This means a full-scale square wave would have an RMS value of +3 dB FS."

As far as I have tested, both a full wave 100 FS sine and square will give "0dB" reading at the current VU meter, supporting more a 0dBFS definition based on peak amplitude. For avoiding digital clippling, where both, FS sine and square wave, would be just at the FS limit, a 0dB peak reading for both would be favourable.

But at least, any clear definition is welcome.

 

[Angel]

we’ll adjust the minimum threshold from –25 dB to –60 dB in the next update

That might be the most important thing for the meter right now ...

 

[hgo58]

This means a full-scale square wave would have an RMS value of +3 dB FS."

What do you mean? The max dBFS value is defined to be 0. There are no positive dBFS values.

The mapping from digital signal level is very simple.

dBFS=20*log10(sample level / max level)

Where max level is 0xffffff for 24 bit sampling (if unsigned Integer is used).

If the max level is reached we have clipping at 0 dBFS.

The calculated meter value in a given periode (e.g. 100ms) is based on the max sample value in that period.

We could just as well use percentage but the logarithmic scale is better for display on a meter.

As you can see from the mapping in #9, the 0 dB is mapped to -18 dBFS (in Europe) but this is just a definition used when performing digital recording, to give a good headroom.

The current WiiM VU meter should show the RMS power value and that would be ok if the full signal range was included. Adding a little gain to the output and you will see the values go well above the +3 dB and out of the scale. Still without any digital clipping. So it's not working correctly.

So I will prefer to have the absolute value of the signal as dBFS to be used instead. It would be a much better indicator when adjusting pre-gain and EQ.

Edit: When the max level is encountered it would be nice to have this indicated on the VU meter with a clipping "flag".

 

[slartibartfast]

What do you mean? The max dBFS value is defined to be 0. There are no positive dBFS values.

The mapping from digital signal level is very simple.

dBFS=20*log10(sample level / max level)

Where max level is 0xffffff for 24 bit sampling (if unsigned Integer is used).

If the max level is reached we have clipping at 0 dBFS.

The calculated meter value in a given periode (e.g. 100ms) is based on the max sample value in that period.

We could just as well use percentage but the logarithmic scale is better for display on a meter.

As you can see from the mapping in #9, the 0 dB is mapped to -18 dBFS (in Europe) but this is just a definition used when performing digital recording, to give a good headroom.

The current WiiM VU meter should show the RMS power value and that would be ok if the full signal range was included. Adding a little gain to the output and you will see the values go well above the +3 dB and out of the scale. Still without any digital clipping. So it's not working correctly.

So I will prefer to have the absolute value of the signal as dBFS to be used instead. It would be a much better indicator when adjusting pre-gain and EQ.

Edit: When the max level is encountered it would be nice to have this indicated on the VU meter with a clipping "flag".

It was a reply to WiiM's statement that the VU meters were measuring RMS values rather than peak. His tests showed that both a sine wave and a square wave with the same 100% peak value resulted in 0dB on the VU meter. If the meter was measuring RMS the square wave would have been +3dB.
That is my interpretation anyway

 

[canard]

What do you mean? The max dBFS value is defined to be 0. There are no positive dBFS values.

The mapping from digital signal level is very simple.

dBFS=20*log10(sample level / max level)

Where max level is 0xffffff for 24 bit sampling (if unsigned Integer is used).

If the max level is reached we have clipping at 0 dBFS.

The calculated meter value in a given periode (e.g. 100ms) is based on the max sample value in that period.

We could just as well use percentage but the logarithmic scale is better for display on a meter.

As you can see from the mapping in #9, the 0 dB is mapped to -18 dBFS (in Europe) but this is just a definition used when performing digital recording, to give a good headroom.

The current WiiM VU meter should show the RMS power value and that would be ok if the full signal range was included. Adding a little gain to the output and you will see the values go well above the +3 dB and out of the scale. Still without any digital clipping. So it's not working correctly.

So I will prefer to have the absolute value of the signal as dBFS to be used instead. It would be a much better indicator when adjusting pre-gain and EQ.

Edit: When the max level is encountered it would be nice to have this indicated on the VU meter with a clipping "flag".

if still the case...there is a drc from, """""discreet""", can be explained, to get around the non-adaptations of level to contourn digital saturations etc...
I'll let the reflection take place to imagine its different impacts, see also in connection with this story of decoration VU meter etc....
and much more generally here...(
because it distorts the way we understand stories about levels, pregains, levels at corrections, etc. ;-) )

No comment, especially since the idea is to enjoy a dancing needle...to music that is 95% dynamically compressed
;-)

 

[hgo58]

It was a reply to WiiM's statement that the VU meters were measuring RMS values rather than peak. His tests showed that both a sine wave and a square wave with the same 100% peak value resulted in 0dB on the VU meter. If the meter was measuring RMS the square wave would have been +3dB.
That is my interpretation anyway

Yes. +3 dB. Not +3 dBFS.

 

[hgo58]

if still the case...there is a drc from, """""discreet""", can be explained, to get around the non-adaptations of level to contourn digital saturations etc...
I'll let the reflection take place to imagine its different impacts, see also in connection with this story of decoration VU meter etc....
and much more generally here...


No comment, especially since the idea is to enjoy a dancing needle...to music that is 95% dynamically compressed
;-)

It's about having a meter that is useful and correct or one that looks nice.

Maybe we can have both?

 

[canard]

It's about having a meter that is useful and correct or one that looks nice.

Maybe we can have both?

Do you understand the profound impact of the DRC "anti-saturation digital" here...?
everything is distorted...
;-)




Even though I understand your justified request for many months...
;-)