Why Mastering to -14dB LUFS is Completely WRONG!!

Measureing LUFs for a couple of seconds is not integrated, integrated is measuring the whole song.

The other flaw is that you don’t subtract the peak level from your LUFS average. If your meter says -11 and it peaks at -3, your LUFS is -11. Chris Muth, former chief engineer at Sterling Sound. I lived the loudness wars from the vinyl era through CD and streaming.

The Advice I've heard, and that somehow makes sense to me is "As loud as possible without sacrificing quality"

The information in this video is completely wrong. This is not how you measure loudness and will give you a false result. The -14LuFS figure is NOT a peak measurement. It is the average value (integrated loudness) across your whole song. In this way some parts can be louder than -14Lufs short term loudness (which is what he is measuring) as long as you have quieter parts to balance it out.

Yeah, I went down the -14 Lufs road once, everything was quieter than what was out there, across the board. It's the usual gatekeeping from the music industry and it's people giving you BS info to keep themselves ahead. If you're making any pop oriented music, -7 Lufs. I don't care who says it too loud, I don't care who says it's clipping, these crabs are trying to keep me in a bucket and I refuse.

Thanks for your diligence on this. Were the songs you analyzed subject to the streaming -14Lufs standard or were they direct from whatever distribution you can them them from? Would ones mastering level be a moot point if the streamers simply raise or lower volume to suit? Also, do you think the gain reduction (Spotify) has an effect on sound quality of an ex: -6lufs track? Thanks again!

in the article they never stated if the loseless files they bought were the same files that are uploaded to streaming services

This is largely on point. As others have mentioned you do need to play all the source audio to get the Integrated rating. Point stands though, the -14LUFS is to normalise levels across a range of material for a good listening experience - it's not a target to aim for when mastering. Also worth noting different genres call for different LUFS rating. Most pop, rock and dance music suits a level of -8 or above imo (even when 'turned down' to -14), but if you push a nu-jazz or lofi hip-hop song to that level you will likely ruin it. I think the biggest problem here is that Spotify has (or had) a guide saying that files above -14LUFS will be turned down, and that this can cause distortions or audio degradation. In my experience, having mastered 100's of songs at -8 or higher that have gone on to streaming services, none of them have suffered any noticeable loss in quality, and I think what the crux of this video is about... is that if I had mastered these songs to -14 in the first place they would not sound as clear, powerful, punchy etc.. in fact, the client would be saying, bro, why is this so quiet compared to all the other tracks I have in my collection i.e. downloaded from Bandcamp, Beatport etc..

One major contributing factor to the current loudness war was moving from tape based recording to digital. With magnetic tape, there was a “universal” reference level of 0 VU (volume units) equaling a known voltage level… in this case .7 ish volts into a specified impedance. With digital, since there was only a maximum level ceiling of 0 dBFS, reference levels now became whatever someone chose as their in-house standard, which is born out in all of the streaming services movie and programming soundtracks being all over the map. Originally, ‘zero’ reference levels were established because of the physical limitations of the medium used to carry the material (tape and vinyl can only be pushed so far before they compromise fidelity). With digital, fidelity as a goal has been compromised for data compaction and loudness. I miss ‘excellent fidelity’.

It's good to hear someone actually say it out loud .... thank you !

Thank-you for pointing this out. I was first introduced to the concept of a Crest Factor by Bob Katz book Mastering Audio (1st edition circa '03 - there are newer versions) and he advocated for a K-14 standard. I've found that as I and music I love has aged, the stuff that holds closer to K-14 and K-12 holds more of my interest over time than more compressed works. Yes, that's no way to win the chart in the current market place, but I'd put the thought out there for artists to have a mix of your work available that can be mastered to -14, -12, -10, -8, etc. dB So that when you and your fans are older, you have another way to re-vitalize your work.

Looking at the dB change in music from the 80s to the 90s is crazy. No wonder they had their own sound back then

Just a quick question, to obtain the desired lufs on your track, do I just turn up the volume on the master until It's reached or should I use other things, like a limiter for example?

Your video is much quieter than the last 5 videos I watched here on Youtube, broski!

Yup. I've been mastering at -8 for many years, because from all my testing, its the loudest I can get songs without introducing distortion or artifacts and still keep that punch that hits you in the chest.

Cool vid! Are you sure the streaming services bring level up to -14dBFS? Initially they only brought levels down. Have you found something that says that has changed and they raise it up?

So, if you take the LUFS of the loudest section of the song, doesn't this artificially increase your LUFS value? If you measure the whole track then your peak section of -8 LUFS drops down to -10/-11 or lower depending on the track. Surely what is important is dynamic range? If you master to -14 LUFS and your peak is -5dB then that's no different to -10dB with a -1dB peak. Also, just from a musical point of view, the streaming platforms' arbitrary choice of -14dB LUFS and -1dB peak (more or less) gives us 13dB of dynamic range to use with no penalty, so maybe this is an opportunity to re-educate ourselves and the public to having music with some actual dynamics? The -8dB 'standard' was for CD production and maybe we've moved on from that? We used to high-pass everything at 40Hz to stop the needle from popping out of the record. Mastering to -6dB just squashes the life out of everything and makes it sound crappy regardless of how well done.

For CD mastering it was always shoot -9 dB average level with a max of -1 dB to avoid clipping DAC. Note that it hasn't changed from RMS. It sounds like most of the references you cited were within a dB or so from that standard. From what I've seen is that the streaming services will adjust the level down to -14 dB LUFS but won't adjust up to that level.

Fake information that will lead your songs to worse quality when uploading on platforms with normalization applied. Yes, mastering to CD or club levels is important when you go for live performance. But these days everyone is sharing their products via yt or spotify and those are really normalized to -14. So when you upload on one of these platforms song that was mastered to higher values, it for sure has impact on dynamics, volume and overral quality. Try for yourself and youll see. Learnig by yourself is better than listening to someones opinion, who obviously isnt saying all of the information.

Very interesting ! What about the LUFS for mixing ? -16db -18db ?? Thank you for your advices