Skip to main content
Guitar Hero
April 24, 2019
Answered

How does the "Normalise Volume" option work?

  • April 24, 2019
  • 23 replies
  • 3387 views
I see there's an option to normalize volume in deezer, but I'm skeptical to use it as I don't know exactly what it does.

  • Will it normalize each track to a given volume target by simply adjusting the general volume of that track, thus keeping all the dynamics of the song intact?
  • Or will it dinamycally adjust the volume in real time, leading to annoying moments, when you expect the music to get really loud or really quiet but it doesn't, as the normilizer algorithm is trying to compensate that volume difference by compressing the dynamic output of the track?
It is annoying when you're listening to a playlist or to flow but and when the next track gets going your speakers suddenlly start distorting because of a big difference in levels but I'd really rather deal with that than to loose the dynamics of the songs I'm listening.
Best answer by Rudi
@meuvoy

That functionality is only available for mobile - if the phone is compatible.
So I guess for PC/Mac is best to find a 3rd party equaliser app, in order to control the output.
If you think this is a feature which should be present in the desktop app, perhaps start your convo in our ideas forum 😉

23 replies

Roadie
December 2, 2021

I have the same problem with volume normalization turned on. Some songs are very quiet, others are very loud. For example: Life Cried - Burnt To Ashes is crazy loud in comparison to other tracks (with volume normalization turned on). Very strange. The song is not very dynamic, so with integrated LUFS over the entire song, it should tend to be more quiet than more dynamic songs. In a dynamic song, the louder parts must be even louder to balance the quieter parts, such that the desired LUFS value will be satisfied. What could be the reason for this?

Roadie
June 16, 2022

For those who visit this thread and are still looking for some answers.

 

I have been trying out the normalization and have some observations:

the 3 songs I used all differ from overall loudness.

Salt - The devil Wears Prada peaks at 0dB and is squashed like a tomato run over by a dump truck. no dynamics. Turning on normalisation will adjust the volume of the whole track by -9dB

Brother Moon (Live) - Gungor also peaks at 0dB but has significantly more dynamics over all turning on normalisation will adjust the volume by approx.. -4 or -5 dB thus peaking higher after normalisation than song one.

Let There Be (Live) - Gungor again peaks at 0dB this time the song stays quiet for a long time before at the end building in energy. this song only got lowered by -4dB keeping the quiet part intact.

song two and three are from the same album and follow after one another. although only an opinion I feel comfortable saying I could not hear a difference between the two songs. not one of them stood out as louder or quieter. (which could be the work of the mastering engineer from the album or Deezer loudness matching songs on albums)

 

I wish I could provide more data by running the audio through an audio plugin which analyses dynamics. but I would have to change my setup to do that.

Roadie
January 1, 2026

I created an account just to say this.

Deezer users care about their sound quality, else they wouldn’t be on Deezer. Some go to further extents than others but any form audio processing should ideally be transparent to the users.

I personally don’t hear a difference if I increase the volume levels after turning on equalize audio - as it tends to bring the overall level down by default, but there’s a always a possibility of psychoacoustics in play and to answer the question from Grapplerzorn: If Deezer published a comprehensive numbers analysis, would it affect your decision - 100% yes. I like the convenience factor, and in the tracks I’ve heard so far I don’t hear much of a difference, but depending on the methodology it might potentially negatively affect only certain tracks with a very wide / small dynamic range, and I would never know.