Pre-Equalization gain for parametric equalization (PEQ)

jed1

Senior Member
Joined
May 7, 2024
Messages
45
@WiiM Support Does WiiM apply pre-equalization gain before PEQ to avoid clipping? For example, with fixed output, volume is set at 100% automatically. If any of eq setting has gain value greater than 1, it can potentially cause a clipping in digital domain. Does WiiM somehow calculate maximum impulse amplitude based on PEQ setting and apply pre-equalization gain? Or, do you plan to have a manual pre-equalization gain setting in the app?
 
Reduce the volume limit setting to accommodate it as per this link

 
Reduce the volume limit setting to accommodate it as per this link

Does this also apply to the case of toslink output at fixed volume? Thanks
 
Yeah. I have my WiiM Mini’s optical output volume limit fixed with a 94% limit based on the 4 band PEQ settings I got from the HouseCurve app and the settings REW suggested as per that link.
 
@WiiM Support If I understand correctly, "Volume Limit" function in the WiiM app is a limiter not attenuator. That means it is not a bulk volume attenuation prior to PEQ. That does not help PEQ application that requires pre-EQ gain lower than 0dB. That makes current PEQ function useless for such case. I will wait until it is correctly implemented.
 
@WiiM Support If I understand correctly, "Volume Limit" function in the WiiM app is a limiter not attenuator. That means it is not a bulk volume attenuation prior to PEQ. That does not help PEQ application that requires pre-EQ gain lower than 0dB. That makes current PEQ function useless for such case. I will wait until it is correctly implemented.
It does work before the PEQ stage in the DSP path, so can be used to provide the desired headroom.
 
It does work before the PEQ stage in the DSP path, so can be used to provide the desired headroom.
I understand the volume limiter is applied before the PEQ. However, it is "limiter" not "gain control". They are not interchangeable. The limiter clips signal above the set % while the gain control reduces the signal to the set %. For the PEQ to work correctly, it needs "gain control" not "limiter".
 
I understand the volume limiter is applied before the PEQ. However, it is "limiter" not "gain control". They are not interchangeable. The limiter clips signal above the set % while the gain control reduces the signal to the set %. For the PEQ to work correctly, it needs "gain control" not "limiter".
I don't think so. The limiter reduces the maximum volume. Try it, set the limiter to 20% and the max volume will be quiet.
 
No, your assumption is wrong. It is named limiter, but that's not how it's working.


And not just max volume. Volume is reduced at whatever level.

@Brantome bet me to it by a few seconds.
Well yes that was what I meant really 😀. It caught me out when I had forgotten I'd set the limit to 20% and thought the amp was screwed 🤣
 
Now, I understand the volume limiter in the app is not "limiter".
You got it. (y)

Even the percentage to dB conversion is not intuitive.
No, it's certainly not. If you restrict yourself to negative gain correction then there's nothing to worry about (or at least nothing that pre-EQ gain could fix). So far I got along doing so in my setup. Also, PEQ is mostly enthusiasts' playground, so they might know about and take care of the issue. But there's definitely room for improvement.
 
Hello! I think it could be solved in software in the eq and peq, in the Eiim Home app, to monitor highlights from 0dB up and have a relative volume slider that "shifts" the whole curve so that it stays below digital 0dB the voice. Or, if it is not possible automatically, you should manually add a slider called Volume or Gain, which pushes the peq and eq curves in various ways so that we can stay at 0 dB. In that case, the final output signal will be quieter due to the highlights, but it will not be distorted. (I think the same is implemented in the Hibymusic software and in Volumio, it intelligently pushes the level down if we use emphasis). Please, Wiim engineers, make it happen here too.

Dear members! If someone could post this idea in the right place, I would appreciate it. The ticket doesn't work for me for some reason.
@wiim
@Brantome
 
Hello! I think it could be solved in software in the eq and peq, in the Eiim Home app, to monitor highlights from 0dB up and have a relative volume slider that "shifts" the whole curve so that it stays below digital 0dB the voice.
This is surely doable in software (PEQ, rather not GEQ). Other software - like REW - can do it. It's not enough to account for positive gain, though. The combined effective result of all EQ filters must be considered to loose as little headroom as possible.

Or, if it is not possible automatically, you should manually add a slider called Volume or Gain, which pushes the peq and eq curves in various ways so that we can stay at 0 dB. In that case, the final output signal will be quieter due to the highlights, but it will not be distorted. (I think the same is implemented in the Hibymusic software and in Volumio, it intelligently pushes the level down if we use emphasis). Please, Wiim engineers, make it happen here too.
That's exactly what we already have. :)

The only downside is that you currently need an external tool (like REW) to calculate the required attenuation and then you need to know that 1% volume reduction equals 0.6 dB of headroom.

You can file a feature request yourself, there are several product related "feature request" sub-forums.
 
I posted to the feature request;

https://forum.wiimhome.com/threads/pre-gain-per-eq-profile.3765/

Maximum amplitude is a simple summation of spectral response (=20*log10(sum(frequency response^2)*delta_frequency). Even without the summation, the upper bound is simply max of PEQ gains.

As long as the app allows manual pre-gain per PEQ profile, that will give users capability to adjust gains.

Here is my use case; I use 2 PEQ profiles per 1 set of speakers. One with subwoofer and another without subwoofer. From time to time, I bypass the subwoofer (which has a foot switch input to bypass). So, I need different pre-eq gains.
 
Hello! I'd love to see it happen, and maybe I didn't say anything stupid about it. It has already been solved in other types of much more lame software. This would be the best: you can even calculate the PEQ itself, what leveling should be done to eliminate the distortion, but we get the best signal to noise ratio. I can't send ideas properly, I don't speak English well. Thanks for helping the Wii listen to this request.
 
Even without the summation, the upper bound is simply max of PEQ gains.
How big are the differences in required pre-gain in your use case? Just applying the inverse of max. boost (to be on the safe side) sounds a bit too coarse for me to make sense. We'll want to give up as little volume and resolution as possible.

In the end, not applying appropriate negative pre-gain doesn't mean that distortion will occur, but that distortion can occur, depending on programme material.

Anyway, I see no downside of (possibly) automatically calculated pre-gain stored per EQ profile, so I'm a supporter.
 
Last edited:
How big are the difference in required pre-gain in your use case? Just applying the inverse of max. boost (to be on the safe side) sounds a bit too coarse for me to make sense. We'll want to give up as little volume and resolution as possible.

In the end, not applying appropriate negative pre-gain doesn't mean that distortion will occur, but that distortion can occur, depending on programme material.

Anyway, I see no downside of (possibly) automatically calculated pre-gain stored per EQ profile, so I'm a supporter.
I use replaygain for all my local tracks. I suppose that lowers the gain enough in most cases to keep the waveform well away from clipping. Some albums such as Dark Side Of The Moon have very low replaygain values though.
 
Back
Top