Mastering for maximum loudness

How to do this, that and the other. Share, learn, teach. How did X do that? How can I sound like Y?
RELATED
PRODUCTS

Post

I've been working on mastering a track of mine for maximum loudness and am running into an issue where when I turn up the gain on my master limiter, I get a lot of issues like reverbs becoming way too loud, distortion happening, and the mix pretty much turning to crap. I have no idea how to address this or what the cause is, but I would guess it has to do with too much of the signal hitting the limiter ceiling? Does this mean I've basically made the track as loud as it can possibly go? It's at about -10 LUFS during the loudest parts; ideally I want to get it up to -5 or even higher, without the audio content becoming messed up. I'm fairly inexperienced with mastering and would appreciate any help in further understanding the topic.

Post

You have to go back to the mix and address the loudness there, if you just push the stereotrack you'll get all these problems you've described.
Soft Knees - Live 12, Diva, Omnisphere, Slate Digital VSX, TDR, Kush Audio, U-He, PA, Valhalla, Fuse, Pulsar, NI, OekSound etc. on Win11Pro R7950X & RME AiO Pro
https://www.youtube.com/@softknees/videos Music & Demoscene

Post

legendCNCD wrote: Sun Nov 28, 2021 12:00 pm You have to go back to the mix and address the loudness there, if you just push the stereotrack you'll get all these problems you've described.
What sort of adjustments should I make in the mix to get a potentially louder master?

How I've been doing mastering is basically, after I've finished the mix, I'd push the gain on the master limiter until the track hits the distortion ceiling. The problem is that the track is only at about -11 integrated LUFS before it starts falling apart, and I'd really like to push it to at least -8. I've also noticed that it's really hard to get the track above -8, even if I cram the limiter all the way up and let the track distort to hell.

Post

It is very hard for me to say without hearing the mix, but I'd go back to mixing in mono first to get volume gains and rough eq done properly. Then listen on low volume for bassdrum vs. bass to be correct - also here you can hear easily which element is the most highest in gain - if you can hear just that then its too high in gain. Basically the loud mix comes from the balancing of the elements. Maybe you have bassdrum in way too high volume and it eats possibility to have high loudness mix for example.
You may also want to clip some peaks off from individual instruments/channels if there are massive peaks, with softclip/peaklimiter or their combination.
Basically I would start the mix from the beginning again by resetting everything (and saving to new project name) (and having somekind of references available also from the same type of music maybe).
Soft Knees - Live 12, Diva, Omnisphere, Slate Digital VSX, TDR, Kush Audio, U-He, PA, Valhalla, Fuse, Pulsar, NI, OekSound etc. on Win11Pro R7950X & RME AiO Pro
https://www.youtube.com/@softknees/videos Music & Demoscene

Post

exist01 wrote: Sun Nov 28, 2021 10:57 pm
legendCNCD wrote: Sun Nov 28, 2021 12:00 pm You have to go back to the mix and address the loudness there, if you just push the stereotrack you'll get all these problems you've described.
What sort of adjustments should I make in the mix to get a potentially louder master?
Flatten the waveform. Saturate the peaks, fix elements that play at the same moment and add up in amplitude. Tame one-shots and risers which can get suprisingly loud. Reduce the volume of some background (or completely irrelevant) parts. Watch for phasing sounds that change amplitude over time and compress them.

Gate all the noises and cut sounds short.
Blog ------------- YouTube channel
Tricky-Loops wrote: (...)someone like Armin van Buuren who claims to make a track in half an hour and all his songs sound somewhat boring(...)

Post

Your mix should be constantly as close to 0 dB as possible. Without clipping, without any plugin on master.

It’s all in the mix. Have been there learned it the hard way.

It does not matter if the mix sounds good if the gainstaging and frequency seperation is faulty. A good mix sounds good and is loud, but a mix that sounds good and your master gain is volatile / extensively dynamic IS not a good mix.

Best mixing tool is proper selection of patterns, envelopes and elements. Not EQ and compression.

What you describe is exactly what you get when your master bus dynamic is out of control. Mostly driven by dynamics of drums, as the poster above said you need to get rid of transient peaks.

You should completely delete the word mastering out of your head. It’s just a proxy, an excuse. A false hope. Mastering makes only sense after the mix is perfect. And then it’s like 3% sugar on top.

There is very good vid from mixbustv about loudness in mix and crest factor. Watch it, it’s sets everything straight clear.

Post

Probably you are already aware of this, but it is important to highlight:

It's supposed to be hard to get to -5 LUFS without sounding like trash, because -5 LUFS is far too loud. A more reasonable goal for a decent mix in almost any genre to around -14 LUFS with minimal limiting, and -12 with transparent limiting.

So how do some artists get their music to much higher levels than that? Well,

a) they compromise the arrangement. Decisions that should be musical/artistic are made in order to facilitate a technical goal of loudness. An obvious example of this would be excessive pumping (often via sidechain compression). Is pumping always bad? No. Is it bad that you have to duck that bass to match the loudness of another track? Yes.

b) a lot of it does sound distorted. Maybe some of that distortion was done on individual tracks, maybe you like the way it sounds... but it is distorted.

So people are right when they say that you need to get more loudness at the mix stage... but be warned that you won't have "the same song but louder" at the end of the process. Probably you need a sparser, more filtered and more distorted track. It might sound worse when level-matched.

Post

Thanks for all the tips and info! An update on my situation - I was able to get my track up from -11 to -8.5 integrated LUFS by getting the trial version of FabFilter Pro-L2 and using that to push the gain a few dB higher. Previously I was using the stock FL Studio limiter, which would mess up the whole mix when pushed, but Pro-L2 did not have that issue at all, and the track sounds pretty much the same, if not better/fuller during quieter parts. If I cram the gain higher still on the Pro-L2, it does start sounding bad, but in a different way, almost like an overpowering sidechain effect, but none of the reverb going crazy, clipping sounds I was getting before. So it seems the limiter you use to push loudness really has a significant effect here. Overall I think I'm okay with the track's loudness now so I will probably call it for this track. During the mixing stage of future tracks I will definitely keep in mind loudness maximization and see if I can push for a higher LUFS still. Oh and the reason I've been particularly concerned about getting a higher LUFS is because some labels seem to have all their tracks at a certain level which was a *lot* higher than -14, so being able to hit that with my tracks seemed important to getting the track accepted by that label.

Post

Yep its all in the mix, can do a -8dB LUFS mix where no limiter will touch almost nothing at all, but its not my target anymore :lol:
Soft Knees - Live 12, Diva, Omnisphere, Slate Digital VSX, TDR, Kush Audio, U-He, PA, Valhalla, Fuse, Pulsar, NI, OekSound etc. on Win11Pro R7950X & RME AiO Pro
https://www.youtube.com/@softknees/videos Music & Demoscene

Post

legendCNCD wrote: Tue Nov 30, 2021 6:44 pm Yep its all in the mix, can do a -8dB LUFS mix where no limiter will touch almost nothing at all, but its not my target anymore :lol:
That sounds awesome. I'm getting about -4dB of gain reduction on most of my peaks at -8.5 LUFS integrated. Quite a bit of red popping down into the graph. Perhaps I can try to reduce my peaks to improve this a bit.

From what I've gathered the really important thing is to tame the peaks so the bulk of the audio can be brought up louder before something hits the ceiling. I guess part of the difficulty is preserving the transients so they still sound punchy, but not take up too much headroom either. If everything's about the same level it would result in a thick, even-looking waveform (good crest factor?) which can then be boosted as a whole and reach very loud levels where everything in the mix is contributing somewhat equal amounts. Will have to experiment with these techniques some more! I really appreciate all the replies from everyone, I think I'm starting to understand things a lot better now.

By the way what's your target now if -8 isn't anymore?

Post

if you're interested in getting the most out of your heavy transients (drums), I've been using Kclip3 for pretty transparent clipping. I learned a lot by watching https://www.youtube.com/watch?v=uV-Ng8F ... aphometrix he goes super in depth on how to use a clipper and I can say it has helped me push my mixes louder and cleaner.

Post

exist01 wrote: Wed Dec 01, 2021 12:32 am
legendCNCD wrote: Tue Nov 30, 2021 6:44 pm Yep its all in the mix, can do a -8dB LUFS mix where no limiter will touch almost nothing at all, but its not my target anymore :lol:
That sounds awesome. I'm getting about -4dB of gain reduction on most of my peaks at -8.5 LUFS integrated. Quite a bit of red popping down into the graph. Perhaps I can try to reduce my peaks to improve this a bit.

From what I've gathered the really important thing is to tame the peaks so the bulk of the audio can be brought up louder before something hits the ceiling. I guess part of the difficulty is preserving the transients so they still sound punchy, but not take up too much headroom either. If everything's about the same level it would result in a thick, even-looking waveform (good crest factor?) which can then be boosted as a whole and reach very loud levels where everything in the mix is contributing somewhat equal amounts. Will have to experiment with these techniques some more! I really appreciate all the replies from everyone, I think I'm starting to understand things a lot better now.

By the way what's your target now if -8 isn't anymore?
Yeah in my experience its about sound selection, balancing and peak information reducing to a point.

"Target" is around -14 LUFS for streaming and I dont like to crush much dynamic chillout tunes :D thats why.. but depends, I think we did the Hazy Wings a bit higher in gain because its more of psytrance style.
Soft Knees - Live 12, Diva, Omnisphere, Slate Digital VSX, TDR, Kush Audio, U-He, PA, Valhalla, Fuse, Pulsar, NI, OekSound etc. on Win11Pro R7950X & RME AiO Pro
https://www.youtube.com/@softknees/videos Music & Demoscene

Post

If you are going for max loudness you need to take that into account from the very beginning, writing, arrangement and mix stages. If your mix has lots of dynamic range, large transients, or a high crest factor, you are gonna have trouble getting the loudness you want in the mastering stage to be transparent.

Post

Keep in mind that there is a difference between absolute loudness, as measured with a DB meter, and relative or "perceived" loudness. Squashing a track too much (e.g., less than -12 LUFS for streaming) may in fact lessen its general perceived loudness by removing too many dynamics, including transients. There is a balance to strike. Here's GRAMMY-nominated mastering engineer Alan Silverman making that case: "The Future of Mastering: Loudness in the Age of Music Streaming" https://www.youtube.com/watch?v=EiRMYoqU3ys

Post

So I have a question about peak reduction - how do you know specifically which things to compress? I am seeing that I am peaking at -2 dB or whatever, and I can mostly tell where it's happening, but is there a way to know what instruments specifically I should compress to result in the most effective amount of peak reduction (since usually there's a bunch of stuff playing at once during those peaks)? Should I solo the various elements and see what thing has the highest individual peak? And is peak build-up mostly caused by similar/overlapping frequencies, or frequencies that are further apart, like for example a cymbal + bass playing at once?

Post Reply

Return to “Production Techniques”