Hi!
I've got 15 electropop singles (danceable radio edits).

Ozone 9 tells me that (in the loudest parts of the song)
some of the singles are -6dB RMS, some are -5dB RMS, some are -3dB RMS.
Each single is mastered to its optimal loudness (so that they sound good).

It might be hard to turn up the -6dB RMS songs and make them louder without loosing punch...

But I want my album to have loudness competitive with major artists albums.
When people listen to my track after The Chainsmokers' track I don't want my tracks to lose because of being quieter (when listeners took my track from my album and created their own custom playlists - and they have my tracks next to major artists' tracks)


So what do you recommend?

Should I back off the mastering limiters of my louder (-3db RMS) songs, and make them -6dB RMS? So that all songs on the album are -6dB RMS?

Or I better check the loudness of my 15 songs in LUFS and disregard completely the RMS levels?
And I make all songs have the same LUFS value?

Thanks!!

 

I'll add that most of the 15 songs have (more or less) similar frequency balance in their choruses
(checked in iZotope TonalBalance 2)

 

I would leave each song as it is, because each song is unique. You have set up each song to sound the best it can. Of course, if the songs on the radio or internet platforms are too quiet, they will automatically make them louder and if they are too loud they will automatically make them quieter.

Pro mastering engineer here: forget about the meters and set relative volumes by ear. Master what will be the loudest song on the album first and get it where you want it in terms of RMS and Peak levels. Then make all the other songs relate to it by ear, not by meter.

This may require making some level changes to song intros and outros in order to make transitions work, but again, this must be MUSICAL and sound good by ear. Mastering is NOT about making everything loud as hell. Mastering is about finesse of both the amount of dynamic range processing AND the color of the sound.

Peak level has fuck all to do with perceived loudness.

Turn the too loud stuff down (or compress/limit it a bit less) so that the album as a whole works if that is how you are going to be distributing.

While peaking higher then -1 or so is generally a bad idea, if a particular track does not need to be that hot, that is fine.

Now individual tracks on a streaming service do often get loudness normalised to about -14LuFS (integrated) or so, and those services generally only turn down, so if that matters to you then doing a per track loudness that is greater then -14LuFS is possibly worth considering (That still leaves you space to do the different gains for the tracks on the album thing).

Spoiler: How do streaming services analyze audio material and how is the output loudness determined?

How do streaming services analyze audio material and how is the output loudness determined?

When you upload your songs to Spotify, iTunes, etc., a distinction is first made as to whether it is an album or a single track. If it is a single track, your track is analyzed using a special algorithm (e.g. ITU 1770) and the integrated loudness is determined.

This value is then stored as metadata for the song. If your song does not meet the desired loudness requirements, the level is either lowered or normalized to the loudness required by the portal, or raised if your song is too quiet. Spotify, for example, then also switches on a limiter that catches all possible level peaks at -1 dBfs, so that no clipping can occur.

The ITU 1770 algorithm is also used by Replay Gain (Spotify uses this program, for example), Winamp and Foobar2000.
UPDATE: July 20, 2021 - ATTENTION: Spotify no longer uses a limiter in the standard setting!

MP3 encoding

- Assume the highest possible bit rate and the filters you need. MP3 mainly has difficulties with the high frequencies. Don't waste bandwidth with 32 kHz instead of 44.1 kHz. Your song will sound better.

- Don't overcompress everything with a compressor/limiter. Leave some of the dynamics of the song so that the encoding algorithm has something to work with.

- Set the encoder to “Best possible quality”. This will allow you to get the best possible results. It takes longer, but it's worth it.

- Remember: MP3 encoding makes the resulting material slightly hotter than the original mix in most cases Limit the output level of the material intended as MP3 to - 1 dB, instead of the usual -0.1 dB or - 0.2 dB. Avoid digital overloads (overs).

- A high-quality MP3 encoder is LAME.

- 160 kbs: The lowest bit rate that is acceptable for files
- 320 kbs: The highest quality, but shows the largest files, but is hardly distinguishable from a CD.

 

Last edited: Feb 12, 2025

Well for dance club CD i go LUFS -9 [some people go to extreme -5/6 squash it!] as for RMS i think -6 is loud. i usually go -8/9 with headroom -7 to -1.0... As for online upload they will add a limiter & even turn it down.

 

-3, -4 and -6 all on the same album?... there are at least two things which come to play.

first of all, a good mastering engineer uses his ears (in addition to measuring.)

secondly, electronic music often has sub bass content which can not usefully be LUFS measured. (if it would be possible, the measuring algorithm had dedicated frequency filters for sub bass)

 

Yes, ditch RMS!. RMS does not measure loudness but is a technical term used to calculate power / voltages etc.... root mean square this is the effective voltage of the waveform

also your tracks should not have all the same lufs value.... rather they should have short term the same lufs value at their loudest parts like the 5 loudest seconds of chorus or sth. like that

 

NO they don't. They also turn up, and to avoid clipping they use their own limiters if turning up. therefore you should always aim for AT LEAST -14 LUFS to avoid being pushed into their 0815 limiter that is not suited for your track. everything louder is fine, some genres need the -6 lufs, others don't.

 

then simply make your tracks sound at subjectively same volume, so listeners won't complain so,
you don't really have to bother with RMS/LUFS (because most engineers aren't either), just make sure to stay under -1dB TruePeak

 

Yes @nctechno, you're right!

Spotify: -14 LUFS
Apple Music: -16 LUFS
Amazon Music: -9 to -13 LUFS
YouTube: -13 to -15 LUFS
Deezer: -14 to -16 LUFS
CD: -9 LUFS
Soundcloud: -8 to -13 LUFS

However, when streaming platforms introduced normalization procedures so that tracks played at the same level, “loud” became much less important. For example, if your track is louder than Spotify's threshold, the platform simply lowers the average level of your track.

In doing so, you sacrifice a large part of a song's dynamic range, taking away a lot of character and depth. It's much easier to preserve dynamic range in a less compressed track, which is why some mastering engineers don't compress near the LUFS threshold for different audio outputs.

Revealing sound deficits:
- Overly compressed and loud productions can reveal their weaknesses through normalization.
- Dynamic songs tend to benefit from normalization as their quality is better preserved

 

Wait, what?
MP3 @ 320k is dam well easily distinguishable from a CD.. Who said that? Whoever it was obviously has not listened to a lot of MP3 encoded material and the artifacts associated with ANY lossy encoding platform. Which is why I tossed MP3 out with the bath water some time ago and switched to FLAC.

 

This comes from the manual Bobby Owsinsky - Mastering like the pros! The emphasis was on hardly!

It is understandable that there are differing opinions on the audio quality of MP3 files. While 320 kbps is the highest bit rate for MP3s and offers acceptable sound quality in many cases, experienced listeners may actually notice differences compared to uncompressed formats such as FLAC or WAV. These differences may be due to the lossy compression of MP3, which can introduce artifacts and a certain amount of sound distortion.

FLAC, on the other hand, is a lossless format that preserves the audio quality of a CD without compression. Many audiophiles and music lovers prefer FLAC to enjoy the best possible sound quality. So it is quite normal that you have switched to FLAC if sound quality is important to you.

 

No, each song should retain its own optimal loudness. Here's how I would approach it:

1. Render each song at its optimal loudness.
2. Put all the songs into a new project.
3. Take the quietest song as a reference and normalize it to -2dBTP (to give yourself some headroom for step 5).
4. Match all the other tracks to the reference, so that all tracks sound equally loud (using an LUFS meter if you want).
5. If you like, you can then introduce some macro dynamic and vary the levels of the individual tracks slightly, so that in the album context, one track has more punch than another (depending on your music and your concept).
6. After that, insert a gain plugin or limiter on your master bus and increase the level until the maximum peak of the album is at -1dBTP.
7. Done.
   

 

If you are correct, this is the most useful advice I've read in a long time. I tend to have OCD with metering.

 

listening to mixes with eyes is a terrible modern habit (and I confess to having such problem too),
human sensing has limited capacity, so if you can't close your eyes when doing mixing on a computer (which frankly you can't) you always compromise on what you could have achieved years back without computers, it's that simple,

that said, well there are few things to keep in mind:

if you have an overly hot mix, then you might have squashed dynamic range too much (but again, if others do, and more importantly if your track sounds fine, who cares), balance lies in taste, not numbers,
on the other hand if you work for clients with specific demands (like Ads or any content for TVs, VODs etc..) you still gotta conform to some norms (like that imperfect -23 LUFS spec, EBU, ITU and such...) but that doesn't mean your mix shouldn't already sound amazing to begin with, and if you have your mind set and some kind of skillset, then it shouldn't be that big of a problem to re-adjust nuances


audio industry is obsessed with all kinds of norms only because companies can sell you gear that supposedly achieves so, but it's not a magic and listeners don't give a damn anyway