For the sake of clarity: “normalized” here means “loudness normalized”.
You’re kind of answering your own question here. Have you measured your reference tracks to see how they compare to your own?
I can’t thank you enough for this much detail. Everything you said is great, tysm.
1 Like
I have. The reference tracks’ mp3 clips at +0.3db and is way noticeably louder than my mix.
OK so there’s half the answer; have you measured the reference track’s LUFS?
@xenon, very curious about something. Two things possibly …
One is, have you confirmed that the actual rendered mp3 (or wav or whatever) is actually also at -2.0 TP and -9 LUFS? I ask 'cuz it’s sounding as though you are satisfied with your mix in REAPER but it’s the rendered product that is quiet? Am I correct with that?
Easiest way to tell is to load it onto Youlean’s online site to get a readout of your metering stats: click HERE. Click on the “Analyze from File” button and find your render and hit okay.
If those #s are not the same as what you see in the Youlean plugin in your DAW, the problem is likely to be the volume on your mix bus. If your mix bus volume is lower than zero it’ll mess up your metering, cuz REAPER’s fx on the mix (on all buses) are pre-fader. That means any volume adjustments will be AFTER Youlean. Easy, easy mistake to make (I’ve done it many times).
The other sneaky thing that’s easy to miss along those lines is if you have any volume automation on the mix bus. Again, that’s all post-fx (well, unless you chose a pre-fx volume envelope).
My apologies if that’s already been addressed in this thread. I only had a few moments and, well, this is a long thread and I didn’t have time to read it all. SORRY!
Hope that helps!
(P.S., I did read enough of the thread to learn that @Lophophora’s real name is Jean Marc! Never knew that!)
1 Like
Hi @Danny_Danzi,
I have carefully read your post and I appreciate the wisdom and hindsight towards what is often a controversial topic. I’m not sure anyone intentionally tried to turn this into a “dark art” though. In my opinion the confusion (or controversy) comes from 2 main factors which are
1/ the fact that some people who mix and master have made technical recommendations their own when they were initially addressed to streaming platforms
2/ the fact that some people talk about LUFS when they don’t really know how it works and why it was invented in the first place
Something else: I have one question about this
This recommendation is based on what exactly? There are so many albums - past and present - considered among the greatest in the history of pop & rock music that are not at this level. So why should LUFS be restricted to such a specific, limited range?
You seem to consider Bob Ludwig as someone who knows what he is doing, to which I totally agree. So I did a quick and simple research and measured three songs of the last three albums he has mastered (or at least is credited for) and the way he sets the loudness seems to contradict what you wrote:
Passenger - Sword from the Stone is mastered at -11,6 LUFS
Mike Fleetwood & Friends - Rattlesnake Shake is mastered at -12,4 LUFS
Madonna - Like a Prayer (dance mix) is mastered at -17,7 LUFS
Maybe I misunderstood what you wrote? Otherwise I’d like to know why your recommendation is so different from what Bob Ludwig (and other mastering legends too, Bob Katz is almost always mastering at low LUFS levels, for instance) usually does.
This is absolutely true if you’re looking at a DAW meter because they are usually showing RMS (over 300ms) + peak levels. But it isn’t the same with LUFS… LUFS has been specifically designed to replicate the way humans perceive loudness, closely matching the way we hear. In fact, it has been engineered to avoid exactly what you are describing.
LUFS calculation goes through 4 stages: response filtering, average-power calculation, channel weighting and summation. The response filtering stage is called ‘K-weighting’ and it combines a shelf equalizer (4dB boost above about 2kHz), intended to replicate the acoustic effects of the head, with a high-pass filter that simulates our reduced sensitivity to low frequencies when it comes to assessing loudness (12dB/octave filter turning over at roughly 100Hz). You can read the full description here, specifically the Annex 1: Specification of the objective multichannel loudness measurement algorithm.
@Lophophora I didn’t accuse anyone here of making it a dark art. Just so you know. I said “no one here”.
Next, the Lufs numbers I shared were what I go for on my music.
That’s a great idea, thanks for sharing. I can’t believe I never thought to try that. Brilliant! 
Wait 'till you hear it - you’ll never go back to sending it through the master bus!
@AJ113 Sorry my phone won’t let me quote.
OK so wait…to do this, are you mastering inside your actual audio project or are you exporting the snare track to fly in to the master?
Sorry for my ignorance. I can see where you’d do it on a project. But if I exported that project and then mastered it to two track, wouldn’t I have the same problem with the snare losing crack with a hot limiter?
I was reading your original post and then it occurred to me…I never master inside of a project unless someone has provided me with stems. So I figured you’re either doing it in your project or bringing in a snare track to send to another out while mastering that blends with the two track? Let me know how you’re doing this if you get a minute? Thanks!
I’ve done it both ways. As you surmise, currently I master on the fly so it’s simply a case of sending the snare to its own output in order to bypass the two-bus processing.
If you’re mastering as a separate process then the snare needs to be rendered separately when you render the mix. When you come to mastering, import both files, but route the snare so that it bypasses the mastering tools.
Either way works just as well in terms of the snare not mushing.
Yes, if you sent it to the limiter, but you’re not going to, right? 
I call it ‘karaoke snare’, because it’s like the snare doing it’s own thing printed on top of a fully mastered track.
The only downside is that if you ever move the master fader, you have to move the snare’s output track fader by the same amount.
If it sounds a little bit too much like a karaoke snare, and unnatural compared to the rest of the mix, don’t worry. It takes a few attempts to get used to the extra clarity and fidelity, and to adjust your ears accordingly. Remember, you’re used to hearing a snare crushed by a limiter. Even at -12 LUFS the snare takes a bit of a beating (see what I did there? LOL). Now you will be hearing the snare in all it glorious cracking, unadulterated majesty. 'Tis a wonderful sound to behold.
1 Like
@AJ113 Lol! I literally laughed out loud reading some of this. Thanks so much for explaining it like that. You know, I have plenty of horsepower on my PC’s to master in-project but…lol…I just have to say it because…I’ve always fought against it and always wanted to say it…here it comes…
I have all these “dark art” secret tools (LMAO!) to master that would probably drive me nuts to see in a project with everything else. Why did it feel so good and mysterious to say that? Hahaha!
The older I get, the more I have to stay on point and remembering things sometimes annoys me. I still can’t believe I have NEVER (watch it will happen now because I said it) forgotten to turn the ARC monitor correction plug off before an export. I’d definitely screw something up with my mastering rig loaded into a mix project. Lol!
I’m gonna try it. Although, I feel the import of the snare stem would be the better option for me. Thanks again @AJ113 for sharing this. It’s going to be super valuable in those times when someone begs me to master something crazy loud. I’ve walked away from doing loud masters for many years because of how it has always ruined certain elements of a good song. Now that I know to do this it can be done with any instrument that may have a mushy type artifact from a BWL.
See that @xenon, now we both learned something and no longer have to worry about the snare losing impact if we decide we want a really loud master. I still say keep your dynamics and master the song volume to match the song first and foremost. But now you got some extra things to try.
I don’t consider myself to be old but I know I would probably forget to turn it off if I used it this way. Fortunately, Reaper has a monitoring fx section which allows me to leave the ARC plugin on all the time without having to worry about turning it off when rendering.