"Loudness" (amount of compression and saturation etc) should just be an artistic choice. I'm glad it moved in that direction instead of just being a marketing trick.
@@dutchdykefinger Oh I like compression/saturation just as much as the next guy. But I think it's clear that at a certain point in time things got really out of hand. Andrew "not only is the loudness war over, I won it" Scheps did some infamously over-compressed/saturated mixes. And I'm even a big fan of Andrew Scheps' work, but at the time he was just a little 'too good' at complying to such marketing demands. Many remasters of old albums also sadly fell prey to the loudness wars. Sure they sounded louder at lower volumes (and thus 'better' in most circumstances), but significantly worse when just really sitting down to listen to it.
@@dutchdykefinger every mp3 I've ever looked at does not have any headroom. Have you found one or why do you say 1.5db of headroom since literally noone does that, can't find a single mp3 that has it. So if you do that then your track is just gonna be 1.5db lower than literally anything else in your library
@@Tiban no , not really, it does eat up alot of it, but not 1.5db. and if you already used True peak limiting in the mastering chain, theres only ever about 0.5 - 0.9db extra created from the mp3 conversion. and if you look at your local mp3s; most of them seem to go way over 0dbfs; thats why my point is valid too - your master WILL be quieter if you use true peak limiting ASWELL as -1.5db of headroom.
NO!!! They are not over until record companies stop the madness. Most new music that I buy is still ridiculously loud. It will take time for the companies and THE ARTISTS to figure it out. Unfortunately, this may take YEARS....
@@TheMirolab This is because, very often, A&R people are not gain compensating when they are reviewing music, as well as many other people with a say in the big money music company chains. Many of them don't know and don't care about this quality concern and let themselves be fooled by loudness.
A friend of mine uploaded a quieter track in spotify: it was really quiet (made on purpose hoping that spotify turned up the volume). That did not happen, i think spotify and other stores do NOT turn up volumes. just down
Make sure the option for level matching is turned on, and make sure you're not using the web client, as it doesn't even offer that option. But yeah, you can't just expect to upload a quiet master and have Spotify turn it up an limit it till it sounds as loud as everything else, because there are so many reason why one track sounds louder than another that can't be caught with a LUFS meter. The loudness wars are not over.
Hi folks. Just to let you know the standards and algorithms for LUFS/LKFS measurement are available online. LUFS uses the k-weighted filter for it's measurements. It's a two-band filter with a low-cut on the bottom and high-shelf on the top that helps to mimic the perceived loudness curves of our ear (Fletcher-Munson curves) . The filter co-efficients are available in the Rec. ITU-R BS.1770 (page 4) paper, on MATLAB documentation, and also if you dig into the JS- on the actual Loudness Penalty website. It isn't a perfect system for measuring loudness but was made as a one-size-fits-all as it is easy and inexpensive for software and hardware engineers to implement into their tools. I've had clients which try to get me to 'trick' it and there isn't really a way to do it without destroying a perfectly good mix- either by boosting loads of bass or keeping high frequencies to a minimum.
You are 100% spot on my dude. I used to be OBSESSED with being as loud as other people in my field and TBH it made the finishing process for every song a NIGHTMARE. I would export 10,0000 times making tweaks and this and that. Now I just make it, master it, export it and boom. So much more simple.
For all newbies that get caught by numbers : try to understand the loudness perception before loudness algorithms! An orchestral piece playing smoothly won’t sound the same as heavy low end trap beat even when they both read -14 lufs (standart for many streaming services) so, master your music as loud as it allows you to do so without sacrificing dynamics and tonality! Happy mixing
I have started to make -14 LUFS so UA-cam won't do anything and Spotify will boost roughly 3dB. Some soft limiting and checking levels are good in Youlean, and if needed for own mp3's or so, just add compressor/limiter plugin in chain after Youlean. Also one Youlean after that to check how levels are, and bang! You can save song with -14LUFS muting last compressor, and then activate it and save louder version.
Also, when mixing tracks, use that compressor to check how it sounds after compressing so it will be good. If only mixing without compressor, it may cause trouble when doing final compression. Just check everything is good using compressor on/off.
Something people also need to understand: DON'T master to specific numbers either. Not only you're not focusing on delivering the best sounding master you can, but also the platforms algorythms are constantly changing, so an "optimized master" today, won't be optimized tomorrow. Btw dude, remember the -1dB mentioned by the online platforms is not the dB we commonly know, but the True Peak/Inter Sample Peak, which is often higher than the common dB. The rule is "never go above -1dB True Peak" written as -1dBTP.
Another important fact that a lot of people seem to forget is that a lot of genres were born and/or thrived during the "loudness war". For example, EDM. Sure, dynamic is better in a lot of cases, but for genres like this the consumer is already used to a really limited sound, because that amount of limiting is an element of that genre I just want anyone to name 1 EDM song below -8 lufs.
I liked what you said at 11:32 about not making different volume level versions...I had thought about it but then...only one master is sent to a distro like CD baby. Thank you for the Videos!
Glad you put this video out. I've never understood why so many people master to -8 Lufs, when it's just going to get turned down anyway. As far as mastering for different streaming sites, I don't that either. I usually just make one file that meets the requirements of two different sites. That is, I master for the highest lufs (-12 lufs required by (I can't remember who)) and for a true peak of -1.5 (from Spotify, I believe). Mastering a file that meets the requirements from the different sites ensures you won't have any trouble no matter where you send it. The only caveat in this is that I believe some sites require a smaller dynamic range.
Very sensible! I'm still amazed at how some masters feel louder than others even when the Lufs read the ame. I guess THAT'S where the real skill of engineering is...
Please use a Spotify ripper program and compare your actual master to the version of your master ripped from Spotify.... come on !! you're the guy to do this.
@@RodrigoroRex there’s ripping methods , someone ripped my tracks off Spotify for me to compare to my own masters. The Spotify versions held up well sonically but were all 7-9db lower in volume than what I had uploaded. Loudness wars over for sure.
Great tutorial! I love the absence of bullsht around the tutorial itself! You're very straight forward and everything you say holds value to the tutorial! Thanks for that.
On TV there is a lot of times when the volume of the TV show is very low. Because of that you bump up the volume. When the commercial comes on it has normal volume. Because you increased the volume, because of the TV show, the commercial is now super loud.
I talked once with an old Mastering engineer , he said that literally the streaming service are going to trun the fader down or up to much their levels, but in extreme loudness its gonna get compressed.. great video my friend
This right here. I know the answer is that the perceived loudness has to do with the mixing, but i would love this being discussed more as its something i feel a lot of people don't talk about.
Create headroom by taking out frequencies that aren't adding valuable information to the track. That's generally gonna be in the low-mids, in my experience - the infamous "smiley face curve". Any couple of dBs you can scoop, means a perceived increase in every other frequency you didn't scoop.
I recommend you watch this video: ua-cam.com/video/EiRMYoqU3ys/v-deo.html In a streaming setting where loudness is "normalized", all songs are adjusted to have the same integrated LUFS value so they have the same perceived loudness. Outside of these platforms the songs with the higher RMS or LUFS value will be perceived as louder and you can achieve this through heavy compression and careful level mixing. In the video I recommended you can see how being slightly above the target value of a streaming platform can make a song sound more interesting than the other songs with higher LUFS value because it is more dynamic. I don't know why but to my experience, it seems songs are still being mastered to reach the -8 LUFS target (or louder) regardless of the streaming platforms loudness control. I'm no proper mixing or mastering engineer so don't take what I wrote as gospel.
The Sonnox Inflator is a great plugin for that, with a limiter can give you that feeling of louder without gaining true peak i think theres other plugin that does that (Reviver) but basically is a “saturation” technique so when you go to the mastering process you can gain even more volume.
@@tired.mp3 u cant have 10 mixes for each streaming service. Just one.. and no they don't normalize unless its turning it down, never up. So make sure your LUFS are hitting in the ball park of commercial music.
Technically, back then, the consumer is the mastering engineer if they're the one in charge of turning volumes up or down so it would be audible/wouldn't blow their eardrums off.
Lots of Mastering Guides and gurus online said to master to -12 LUFS. So I pushed my limiter with a -3 dB headroom. However, all other tracks were louder than mine, even with normalization. So I pulled some reference tracks into my DAW and measured them. They were -7 and -5 LUFS. 😐😨😳 Had a talk with a friend of mine who's a Mix Engineer for Sony, Universal & Warner. He's like: "Yeah, nobody master at -12 LUFS, you want to do it minimum at -10 LUFS, then let streaming services normalize it." Turns out to lower your loudness overall makes your track louder than making it dynamically hit -12 LUFS. Blew my mind, why don't people teach you that? I've probably spent 10 years thinking my production was bad due to wrong mix volume 😷
By watching this video, I´m really happy to see that there´s still a "hope" for this new generation, and the great values and achievements we did in the past, our experience and knowledge will go on. Congratulations, just one more subscriber. Even at my age, I want to learn more and more.
It is good practice to be at minus 3db peak limit to allow for aliasing avoidance issues and levelling stuff the sites do across the board, much better for them to boost the volume if needed, on your track, rather than squash it down in volume for these reasons.
2 Toughts on This: 1) I know many people who have disabled loudness norm. in spotify, due to the reason that tracks from EDM Artists like Hardwell, KSHMR and so on transfere their energy best if their is no processing going on. And there is also no loudness normalisation on beatport releases - so if you want to sound as loud as the "stars" you have to be somwhere around -6 to -5 LUFS. 2) I think its way more important to check the artefacts which compressing (wav. to acc. or mp3.) creates. Those artefacts can create somer louder signals and then you get even more "penalty" then you should get. Be Aware of this. You can check it with ozone for example. Just my thoughts on this topic.
Haven't tried this myself, but you can have a louder chorus/drop if you turn down every other section by a few dB as the overall average loudness is less
The guidelines are great, thank you. At the end of the day one has to use one 's ears according to what is best for the song and depending on what material it hosts inside of it. If one is not willing to feel out how loud a collection of files should be mixed and thereafter mastered at as one file then one won't find it's sweetspot. Don't think the sweetspot is not loud enough, it is, then when mastering it's all about where you can take that sweetspot to a louder one if needed according to what processes the material can handle ect. Don't mix at mastering levels either, leave that headroom for your dynamics that will later trigger your mixbus chain and or mastering chain. The mix's headroom takes hands with what mastering chain can achieve without destroying the music. Personally I find it easy to hear good or bad pushback from any device or plugin used when making music or pushing that music upwards. I hope my attempt at commenting in a non technical way helps.
Here are some tips for anyone who can't decide how should they master their song: - If you think most people will be listening on headphones in a bus, train or a public place, or at lower volumes, lower dynamics will benefit your song because they will hear details better even if the noise floor is really high. - Higher dynamics basically will be more exciting but it won't be as much fun when listening in bad conditions. - For CD masters, loudness war still continues, so you should be aiming for maximum loudness (Unless it's slow, jazz, classical etc. because their song structures benefit from dynamics too much)
Damn! So relevant and so to the point I really hope it comes across for everyone else. For me, you really helped to see it all a bit more clearly. I'm mostly self taught as an engineer thru necessity (which sounds stupid when I say it, but all engineering is equations mostly combined with a little trial and error) ... So as a musician / producer / composer (/engineer) this is more helpful than even you know lol ... You make me feel a little more sane in a crazy world of consumer driven "MORE" culture ... Things that are quiet actually make us listen more, because we have to. Creative use of dynamics are the "secret sauce" a lot of us producers and artists seem to miss for the first few years. If the only good thing to come out of the streaming revolution is the end of the loudness wars then maybe it was worth it ..... Maybe? Thanks again for another razor sharp upload my brother in loudness lmao
ps i wish we could all JUST master "commercial" music to around -14 LUFS regardless of CD's and P.A. systems with too much headroom (ironically because of inconsistent mastering levels and buzzing DJ's in their twilight years with blown out eardrums lol)..
10:20 I think they want the tracks to be -1db at peak is because when the file gets compressed to a codec it gets a little louder and so it doesn’t go over 0db
Nice video and good tips those you share. With all the due respect, there are few things I don't agree with: 1) some streaming platform (for ex. Spotify) have the option to disable the normalization. I think that in a short time almost all platforms will make it as an option so that no genre of music will be penalized. 2) I would prefer to master my track without thinking about streaming platforms. I make dubstep and the modern songs have loudness like -1 LUFS and peaks up to +5dB. I prefere that my track is penalized by UA-cam and not by a live gig. The risk of "yawning audience" at a dubstep concert is really too scary. 3) if music is still art, it's the artist that decide, not a streaming platform. People can download the mp3 version and listen to utra loud music. But if we start to master according to Spotify, make them shorter and fast hooking for Tik Tok, and so on... I think we lost the meaning of music. This is just an exchange of opinions, I totally respect your point of view, so please, don't consider this as an attack!
Ok... Correct me please, If I get this wrong... According to my understanding the human ear hears loudness mostly via the average volume encountered. Above is the reason why a more compressed file always sounds louder in comparison to the less compressed one ie. a track by -10 LUFS with 8dB dynamic range sounds louder than a track by -10 LUFS with 15 dB dynamic range. Right? So to achieve loud results in comparison with the rest of the bunch means that the quieter parts should sound louder in comparison to the peaks ie. compression should be used... Naturally in a way that it doesn’t make the sound all too flat. Getting the very last part (compressed, but not flat) from the above is the true skill of a mixing (and in some cases the mastering) engineer. Right?
Defintly the prefer the dynamic version! The super compressed version gives me shivers just by looking at the waveform. It’s a solid black bar! Absolutely horrible that mastering engineers do this to the listeners. As a listener it makes me cry just seeing that hard clipped loud to the max version. I am glad the loudnesswar is no more and it also should never come back. As a hobbiest I would not want to have that sort of result and go back to the drawing board and mix it again from scratch. Sadly a lot of radio stations trash the sound that includes Veronica which I listen at work.
It's no surprise that the plugin shows +2.x dB at around 9 mins. into the video since your v2 master item had been turned down by 2.4 dB and it was already peaking a few 1/10 of a dB below 0 dB to begin with. This means that its peak level after turning it down by 2.4 dB was probably around -3 dB, thus, provided that the mix complies with the service's loudness regulations, they will normalize it to peak at 0 dB. No "secret" limiting or compression involved - just linear gain. If you don't overdo your mastering loudness-wise, you can upload fairly loud mixes peaking at 0 dB and it will not get touched by the algorithm.
WARNING: as much as I love this channel, this is VERY misleading. Master your track at the recommended Spotify volume and then be amazed and depressed when your song sounds significantly quieter than the next one. The default setting in Spotify is normalization but you can turn it off, and they’re not transparent about the process of “normalizing” at all. If you look in the Spotify for artists section on mastering, they give a vague explanation, but it’s clear that it’s not perfect. Bottom line, all the top mastering engineers still master to a -8 RMS for any mainstream style of music (hip-hop, electronic/edm, rock, pop). The art of mastering at this point is being able to make loud masters that are also dynamic. IT IS POSSIBLE. You just have to know what you’re doing. For dynamic music (classical, jazz, folk), yes, make it quieter. But what’s also never talked about is you don’t even use a limiter in these genres usually (altho a lot of contemporary jazz that’s not off blue note or verve definitely is trending towards a more contemporary limited sound, e.g. Brownswood records). My suggestion is use a maximizer and then 2-3 limiters in series as the first thing you do, and then work backwards. Start at your target volume and then make the master fit into that volume. Why? Because maximization and limiting IS compression. You need wayyyyy less compression than you think once you’ve done this, and then you just sculpt into the limiter. Try it. Game changer. Anyway, don’t believe the hype. Great channel, just a misleading video
Hey Justin, heard the same thing. I am wondering, how do you set up your mastering chain? I've been pushing my volumes up with just one limiter but seems like it's not ideal and I should use many limiters in a chain. I am just not sure how much gain to push in each of those limiters. Say I need +10dB increase in gain, do I slap on three limiters, each one giving approx. 3-4dB in gain?
Excellent vid, looking at most of the angles... Its funny when I went back to some of my historically favorite tracks over the years and started looking at them as reference tracks, I found that the ones that were more dynamic and well saturated with a nice central density always sounded better to me than the ultra loud ,peaks flattened out, solid block of sound like frequency dynamic. When i started mastering tracks more for my own ears rather than like trying to compete with some sort of quest for the loudness gold... mind you there is nothing wrong with that either... its about preference I guess... but no matter anymore if we host on these sites , they get the last hands on it...
The only problem I find on uploading quieter songs to Spotify is that you can disable normalization (I do that) so if you listen afterwards a "normal" mastered track it'll blast you away. In the genre I'm into (Metal) you can't risk sounding way quieter than classic tracks if people disables normalization, and I bet many Metal listeners disable this feature. Besides, at least in the genre I'm talking about, records are still being mastered loud in 2020, loud as 6 LUFS. I'm actually mixing and mastering my upcoming album and I've aimed for 7 LUFS, which is loud, but not as loud as some records I've listened this year. In the end, I decided to aim for a loudness level that doesn't squash the hell of my music just for loudness sake, yet that it can stand to the loudness of the records I've heard in the past 15 years, but more importantly, that it sounds great and when Spotify and UA-cam lower the volume, it keeps sounding great.
Hi there, I come with a question I asked myself today. We all know about loudness for streaming like youtube, spotify, etc etc… But what about UA-cam, Twitch "live streaming" services ? What volume are they streaming ? Can't find any answer even on supports. Was wondering if that also could be something to take in mind during mastering as there are more and more dj live streamings during this period. Cheers everyone, and thanks White Sea for your videos, always a good moment to watch ! Greetings, Gregory.
The thing about this is that when people have the "normalize" oprtion off, you get squashed by other songs if you made a -14 LUFS Master, so still even doe you loos some dynamic range mastering loud I thhink is the best option rn
Elim neden sürekli ses seviyesi düğmesine gitsin? Normallestir düğmesi sürekli açık. Ayrıca Spotify her güncelleme de otomatik olarak onu açar. Ve youtube normallestirme kapanmaz.
4:18 This version sounds so much better, more "3-D" than the squashed version. I would buy a copy of the dynamic one whereas I'd ignore the squashed one because it has "that sound" that is in so many pop tracks. The dynamic version has a lot more depth.
One thing that gets me annoyed is that, yes, we can master towards being close to the streaming standards, BUT, try turning of the loudness standardising on the Spotify desktop app and listen to something like God's Plan vs a track uploaded to streaming standards (my track Richard Carter - Blow Up, was mastered around -13luffs). His track is way louder without the control, so it seems like big studios don't actually care?
Something that really messes with my nerves is that there is no accurate tool measure that accurately, so they don’t even have to touch the volume at all... I would love it if streaming platforms made it more transparent.
There is more to it than this because there is still a great deal of loudness differences within any single streaming service. So they don't seem to normalize or limit in the way we may think. There is some unknown variable we don't know yet?
well they are even in UA-cam controlling Loudness but with a Normalization technology or algorythm (call it whatever you want) that is horrendous because i can switch to another video of other years or a videoclip or a show and i instantly hear the different loudness and peaks and distortions, the same occurs in Spotify ETC. So, it is everywhere digitally the control of loudness from 2015 or even 2014 but, it's badly processed we need years for this normalization to be great enough.
Men, you can sound louder with presence, boosting 3k - 6k with control. I heard song bad mixed in spotify boosting 10k-12k and sound quieter. Using normalizer of spotify.
High loudness at festivals? I wouldn't do that. Best advise is always dynamic pushed a bit into loudness territory for festivals and loud volume. Live sound engineers (hi!) correct your sound! Also due to sound restrictions/corrections. Which could lead into lower volume or comb filter effects.
I think that's right but many electronic music producers still mastering their tracks very loud maybe because it's for the style of the track, but I also know that spotify has a selectable volume normalizing so isn't better to be louder?
Master every track to around -10 LUFS and call it a day. In this way you won't be punished by streaming normalisation very much, while ur limiter is actually compressing the mix so the mix has enough "wall/brick" feeling at high energy
One would have to hypothesize : so if a streaming service sets their loudness threshold lower, doesn’t just mean the loudness zombies will just try to be the loudest person at that level. Or maybe they just crank that limited -0.1 as usual and let the streaming services dial it back as they will . Either way, I am grateful that this has happened. Yes I love EDM power, and there is a time and place for boom- but it’s nice to see people have to work harder to create contrast in their music and use variation in volume via LUFS as a way of achieving it. Instead trying a different limiter to sqHash a signal into an inky black bar of boom lol All this means is better variety across the board
Great video as always man ! Still you talked about low dynamic music for festivals club etc.. and high dynamic for streaming services, but why is that ?
One problem is that UA-cam Music premium service doesn't compensate for volume. All tracks are at original volume. It's one of the reasons I use the service but it still makes you think about the loudness more than if it was a truly universal practice.
Unfortunately, I think most services do a file that's too loud (so yeah, don't get it too loud). Something which I think is important and oftentimes disregarded is that some platforms only reduce the loudness when it's over the limit; they don't boost it when the loudness is below the level (UA-cam doesn't, probably 'cause you don't want those fail videos with people screaming to be that loud hahahah). On the loudness penalty, when they put the "--" indication instead of a value, that means the file is not too loud, and therefore will say as it is, quieter than other music (unlike Spotify).
So when you master above -14 LUFS to try to make it louder, likely with a limiter/maximizer/clipper, then you’re really just squashing the dynamic range, because it’s ultimately gonna get turned down anyways. I think rather than focusing on the loudness, or LUFS, you should focus on dynamic range, and just how it sounds/feels. If anything, aim for -14 so that the normalization doesn’t affect it. What are your thoughts?
Super helpful video... the last part of just making two exports instead of multiple ones was super helpful. I think what is different these days is that the loudness war was getting ridiculous and the way music was meant to be listened to was getting confused. Then I think it because easy for everyone to make loud mixes anyway and then the services stepped in and thought "let's try to just make the music speak for its self"... It also did a bot of a "fuck you" to those people obsessed with it being simply "loud"
Does anyone know if they treat tracks in classical categories differently? Those tracks are likely to be more like -20 LUFS, and turning them up that much would probably ruin them, whether they’re limited or clipped or whatever, but is that really happening to classical tracks too? (Read other comments, answer seems to be that they do NOT turn volume up, only down? Makes sense actually. Wont kill any dynamics you care to include, but will tone down the loudness wars.)
yep, for example all the current modern metal releases are at like -5 sometimes even -4 LUFS. In this genre, heavy limiting and hardclipping makes the sound - you get everything "in your face".
I am pretty sure that spotify is still not measuring in LUFS but in replaygain which is really a problem. Also they claim to turn up the volume with a compressor wich is mostly not good as well.
Thank you Wytse! This will save me a lot of time trying to make all my tracks have the same exact loudness! I recently signed up with SoundCloud which, BTW, nobody is talking about. Not even Loudness penalty has SoundCloud in their loudness analysis. Does anybody out there know why? Keep up the great videos!
@@mekullag lol thanks well still way back when I first encountered Lufs, I thought perhaps it was a German measurement as it seemed more accurate?? ( i know thats bad right just assuming) so Ive always said it kind of with that German/Russian l-U-f extra emphases on the U :)
Yeah it does come across on UA-cam. The dynamic one sounds better, quite obvious on my headphones, but oddly enough it also sounds louder to me, meaning LUFS is not entirely accurate. There are other things that unfortunately make LUFS inaccurate. Try comparing a LUFS normalised saw wave, sine wave and square waves for example.
it can get the itunes figure quite wrong sometimes, eg I compare a quieter file to a louder one, and the quieter one has MORE of a loudness penalty for itunes.
well if you master for cd, the more dynamic one will sound better, always I've always thought that EQing is most important in mastering, not that making it louder part as in case of cd the end user controls the volume, not the mastering engineer
hello all! it is really weird that i have been analysing youtube audio lufs volume and different youtube channels. there is a music that are (integrated lufs) about -7 to -11 lufs in youtube videos so i think whats going on ? some/mostly tracks are about or close to -14 lufs in youtube but some music tracks are way more louder than youtube's loudness reference -14 lufs. some youtube channels have very loud and louder than youtube's -14 lufs settings (integrated lufs) and the loud tracks that i have been analysing are -7 to -11 lufs mostly...so all people can't get same lufs settings and thats really annyoing. some big channels upload a music track and that is be like -7 lufs and when small youtube channel upload a music track it will mostly be going to youtube's -14 lufs setting and it is really annyoing! youtube lufs volume and not all people are equal...sorry for my english...😀😎🥳 I noticed it when I analyzed the volume in youtube with different music tracks and small + big channel 😥😥🤔🤔 and this issue is hardly talked about at all...
I fully agree with thi comment. We all know this information in this video and still I m very grateful for all the free advises you give, thanks so much, keep doing this, but I'm a realist too!! We already make sure we master around -12.5 lufs and stay under final peak -1, but all the tracks I released so far were not as loud as the tracks coming from the big music factories, so I think there is another proces going on besides the so called algoritmes, something with "money" I would guess. Can you comment on that?
I heard that I have to send a track to Spotfy at -1db peak. That means my last brickwall limiter should have a -1db celling? I always use a celling of -0,3 dB. Please, I would love to understand that.
Even with all of this loudness compensation going on I still think (it sounds to me) they are running the loudness of the commercials higher than the program content. Just like the volume of the commercial in this video, I turned the volume down because of how loud it was and back up after the commercial. Is this actually happening or is it just me? Have you ever tried measuring the difference?
"Loudness" (amount of compression and saturation etc) should just be an artistic choice. I'm glad it moved in that direction instead of just being a marketing trick.
yet i'm hard pressed to find anything that's overcompressed
@@dutchdykefinger Oh I like compression/saturation just as much as the next guy. But I think it's clear that at a certain point in time things got really out of hand.
Andrew "not only is the loudness war over, I won it" Scheps did some infamously over-compressed/saturated mixes. And I'm even a big fan of Andrew Scheps' work, but at the time he was just a little 'too good' at complying to such marketing demands.
Many remasters of old albums also sadly fell prey to the loudness wars. Sure they sounded louder at lower volumes (and thus 'better' in most circumstances), but significantly worse when just really sitting down to listen to it.
the reason for -1dB headroom is because it may clip after being converted to the codec of platform and avoids intersample peaks
1.5 db for mp3
I guess my question is why they cant analyze the volume first then convert.
@@dutchdykefinger every mp3 I've ever looked at does not have any headroom. Have you found one or why do you say 1.5db of headroom since literally noone does that, can't find a single mp3 that has it. So if you do that then your track is just gonna be 1.5db lower than literally anything else in your library
AboveEmAllProduction that’s why the headroom is necessary, the conversion literally eats it up
@@Tiban no , not really, it does eat up alot of it, but not 1.5db. and if you already used True peak limiting in the mastering chain, theres only ever about 0.5 - 0.9db extra created from the mp3 conversion. and if you look at your local mp3s; most of them seem to go way over 0dbfs; thats why my point is valid too - your master WILL be quieter if you use true peak limiting ASWELL as -1.5db of headroom.
Sounds to me like loudness wars are over and streaming services and the consumer won. As it should be. Dynamics for the win. 💯
NO!!! They are not over until record companies stop the madness. Most new music that I buy is still ridiculously loud. It will take time for the companies and THE ARTISTS to figure it out. Unfortunately, this may take YEARS....
@@TheMirolab This is because, very often, A&R people are not gain compensating when they are reviewing music, as well as many other people with a say in the big money music company chains. Many of them don't know and don't care about this quality concern and let themselves be fooled by loudness.
@@TheMirolab Yeah the artists not understanding audio at a deeper level will be the problem.
A friend of mine uploaded a quieter track in spotify: it was really quiet (made on purpose hoping that spotify turned up the volume). That did not happen, i think spotify and other stores do NOT turn up volumes. just down
That is also what they told me
I knew UA-cam do this, just trying it out of curiosity a couple months ago
I think this is quite likely true otherwise they'd need a limiter in their normaliser
Remember u can still mark "normalize" option in Spotify settings and check them again
Make sure the option for level matching is turned on, and make sure you're not using the web client, as it doesn't even offer that option. But yeah, you can't just expect to upload a quiet master and have Spotify turn it up an limit it till it sounds as loud as everything else, because there are so many reason why one track sounds louder than another that can't be caught with a LUFS meter.
The loudness wars are not over.
Mastering is such a misunderstood art. Great video as always - thanks!
Hi folks. Just to let you know the standards and algorithms for LUFS/LKFS measurement are available online. LUFS uses the k-weighted filter for it's measurements. It's a two-band filter with a low-cut on the bottom and high-shelf on the top that helps to mimic the perceived loudness curves of our ear (Fletcher-Munson curves) . The filter co-efficients are available in the Rec. ITU-R BS.1770 (page 4) paper, on MATLAB documentation, and also if you dig into the JS- on the actual Loudness Penalty website. It isn't a perfect system for measuring loudness but was made as a one-size-fits-all as it is easy and inexpensive for software and hardware engineers to implement into their tools. I've had clients which try to get me to 'trick' it and there isn't really a way to do it without destroying a perfectly good mix- either by boosting loads of bass or keeping high frequencies to a minimum.
You are 100% spot on my dude. I used to be OBSESSED with being as loud as other people in my field and TBH it made the finishing process for every song a NIGHTMARE. I would export 10,0000 times making tweaks and this and that. Now I just make it, master it, export it and boom. So much more simple.
this is exactly how i am right now. did you just fix it by following his steps in this video?
For all newbies that get caught by numbers : try to understand the loudness perception before loudness algorithms! An orchestral piece playing smoothly won’t sound the same as heavy low end trap beat even when they both read -14 lufs (standart for many streaming services) so, master your music as loud as it allows you to do so without sacrificing dynamics and tonality! Happy mixing
I have started to make -14 LUFS so UA-cam won't do anything and Spotify will boost roughly 3dB. Some soft limiting and checking levels are good in Youlean, and if needed for own mp3's or so, just add compressor/limiter plugin in chain after Youlean. Also one Youlean after that to check how levels are, and bang! You can save song with -14LUFS muting last compressor, and then activate it and save louder version.
Also, when mixing tracks, use that compressor to check how it sounds after compressing so it will be good. If only mixing without compressor, it may cause trouble when doing final compression. Just check everything is good using compressor on/off.
Your hair has a lot of dynamic range. I will give you a plus.
Needs some clipping in my opinion. 😂
@@adl0815 should give it some color though
you guys aren't serious LMAO
Great Video, really appreciate you demonstrating the importance of digital mastering to us (Engineers/Producers) 👍💯
Something people also need to understand:
DON'T master to specific numbers either. Not only you're not focusing on delivering the best sounding master you can, but also the platforms algorythms are constantly changing, so an "optimized master" today, won't be optimized tomorrow.
Btw dude, remember the -1dB mentioned by the online platforms is not the dB we commonly know, but the True Peak/Inter Sample Peak, which is often higher than the common dB. The rule is "never go above -1dB True Peak" written as -1dBTP.
Another important fact that a lot of people seem to forget is that a lot of genres were born and/or thrived during the "loudness war". For example, EDM. Sure, dynamic is better in a lot of cases, but for genres like this the consumer is already used to a really limited sound, because that amount of limiting is an element of that genre I just want anyone to name 1 EDM song below -8 lufs.
I liked what you said at 11:32 about not making different volume level versions...I had thought about it but then...only one master is sent to a distro like CD baby. Thank you for the Videos!
Glad you put this video out. I've never understood why so many people master to -8 Lufs, when it's just going to get turned down anyway.
As far as mastering for different streaming sites, I don't that either. I usually just make one file that meets the requirements of two different sites. That is, I master for the highest lufs (-12 lufs required by (I can't remember who)) and for a true peak of -1.5 (from Spotify, I believe). Mastering a file that meets the requirements from the different sites ensures you won't have any trouble no matter where you send it. The only caveat in this is that I believe some sites require a smaller dynamic range.
Clear, informative and encouraging! 👍🏻
Very sensible! I'm still amazed at how some masters feel louder than others even when the Lufs read the ame. I guess THAT'S where the real skill of engineering is...
Honestly it's magic I'm trying to figure out right now ;-;
Please use a Spotify ripper program and compare your actual master to the version of your master ripped from Spotify.... come on !! you're the guy to do this.
You can't rip off Spotify, it's all encrypted
@@RodrigoroRex there’s ripping methods , someone ripped my tracks off Spotify for me to compare to my own masters. The Spotify versions held up well sonically but were all 7-9db lower in volume than what I had uploaded. Loudness wars over for sure.
Great tutorial! I love the absence of bullsht around the tutorial itself! You're very straight forward and everything you say holds value to the tutorial! Thanks for that.
On TV there is a lot of times when the volume of the TV show is very low. Because of that you bump up the volume. When the commercial comes on it has normal volume. Because you increased the volume, because of the TV show, the commercial is now super loud.
I talked once with an old Mastering engineer , he said that literally the streaming service are going to trun the fader down or up to much their levels, but in extreme loudness its gonna get compressed.. great video my friend
can you make a video about psychoacoustics, and how to get a "perceived louder" master?
This right here. I know the answer is that the perceived loudness has to do with the mixing, but i would love this being discussed more as its something i feel a lot of people don't talk about.
Create headroom by taking out frequencies that aren't adding valuable information to the track. That's generally gonna be in the low-mids, in my experience - the infamous "smiley face curve". Any couple of dBs you can scoop, means a perceived increase in every other frequency you didn't scoop.
I recommend you watch this video:
ua-cam.com/video/EiRMYoqU3ys/v-deo.html
In a streaming setting where loudness is "normalized", all songs are adjusted to have the same integrated LUFS value so they have the same perceived loudness. Outside of these platforms the songs with the higher RMS or LUFS value will be perceived as louder and you can achieve this through heavy compression and careful level mixing.
In the video I recommended you can see how being slightly above the target value of a streaming platform can make a song sound more interesting than the other songs with higher LUFS value because it is more dynamic.
I don't know why but to my experience, it seems songs are still being mastered to reach the -8 LUFS target (or louder) regardless of the streaming platforms loudness control.
I'm no proper mixing or mastering engineer so don't take what I wrote as gospel.
The Sonnox Inflator is a great plugin for that, with a limiter can give you that feeling of louder without gaining true peak i think theres other plugin that does that (Reviver) but basically is a “saturation” technique so when you go to the mastering process you can gain even more volume.
@@tired.mp3 u cant have 10 mixes for each streaming service. Just one.. and no they don't normalize unless its turning it down, never up. So make sure your LUFS are hitting in the ball park of commercial music.
Technically, back then, the consumer is the mastering engineer if they're the one in charge of turning volumes up or down so it would be audible/wouldn't blow their eardrums off.
Lots of Mastering Guides and gurus online said to master to -12 LUFS. So I pushed my limiter with a -3 dB headroom. However, all other tracks were louder than mine, even with normalization. So I pulled some reference tracks into my DAW and measured them. They were -7 and -5 LUFS. 😐😨😳
Had a talk with a friend of mine who's a Mix Engineer for Sony, Universal & Warner. He's like: "Yeah, nobody master at -12 LUFS, you want to do it minimum at -10 LUFS, then let streaming services normalize it."
Turns out to lower your loudness overall makes your track louder than making it dynamically hit -12 LUFS.
Blew my mind, why don't people teach you that? I've probably spent 10 years thinking my production was bad due to wrong mix volume 😷
By watching this video, I´m really happy to see that there´s still a "hope" for this new generation, and the great values and achievements we did in the past, our experience and knowledge will go on. Congratulations, just one more subscriber. Even at my age, I want to learn more and more.
It is good practice to be at minus 3db peak limit to allow for aliasing avoidance issues and levelling stuff the sites do across the board, much better for them to boost the volume if needed, on your track, rather than squash it down in volume for these reasons.
2 Toughts on This:
1) I know many people who have disabled loudness norm. in spotify, due to the reason that tracks from EDM Artists like Hardwell, KSHMR and so on transfere their energy best if their is no processing going on. And there is also no loudness normalisation on beatport releases - so if you want to sound as loud as the "stars" you have to be somwhere around -6 to -5 LUFS.
2) I think its way more important to check the artefacts which compressing (wav. to acc. or mp3.) creates. Those artefacts can create somer louder signals and then you get even more "penalty" then you should get. Be Aware of this. You can check it with ozone for example.
Just my thoughts on this topic.
Haven't tried this myself, but you can have a louder chorus/drop if you turn down every other section by a few dB as the overall average loudness is less
Sounds almost top down
11:05 Amen!😇
The guidelines are great, thank you. At the end of the day one has to use one 's ears according to what is best for the song and depending on what material it hosts inside of it. If one is not willing to feel out how loud a collection of files should be mixed and thereafter mastered at as one file then one won't find it's sweetspot. Don't think the sweetspot is not loud enough, it is, then when mastering it's all about where you can take that sweetspot to a louder one if needed according to what processes the material can handle ect.
Don't mix at mastering levels either, leave that headroom for your dynamics that will later trigger your mixbus chain and or mastering chain. The mix's headroom takes hands with what mastering chain can achieve without destroying the music. Personally I find it easy to hear good or bad pushback from any device or plugin used when making music or pushing that music upwards.
I hope my attempt at commenting in a non technical way helps.
Here are some tips for anyone who can't decide how should they master their song:
- If you think most people will be listening on headphones in a bus, train or a public place, or at lower volumes, lower dynamics will benefit your song because they will hear details better even if the noise floor is really high.
- Higher dynamics basically will be more exciting but it won't be as much fun when listening in bad conditions.
- For CD masters, loudness war still continues, so you should be aiming for maximum loudness (Unless it's slow, jazz, classical etc. because their song structures benefit from dynamics too much)
FACTS!
Damn! So relevant and so to the point I really hope it comes across for everyone else. For me, you really helped to see it all a bit more clearly. I'm mostly self taught as an engineer thru necessity (which sounds stupid when I say it, but all engineering is equations mostly combined with a little trial and error) ... So as a musician / producer / composer (/engineer) this is more helpful than even you know lol ... You make me feel a little more sane in a crazy world of consumer driven "MORE" culture ... Things that are quiet actually make us listen more, because we have to. Creative use of dynamics are the "secret sauce" a lot of us producers and artists seem to miss for the first few years. If the only good thing to come out of the streaming revolution is the end of the loudness wars then maybe it was worth it ..... Maybe?
Thanks again for another razor sharp upload my brother in loudness lmao
ps i wish we could all JUST master "commercial" music to around -14 LUFS regardless of CD's and P.A. systems with too much headroom (ironically because of inconsistent mastering levels and buzzing DJ's in their twilight years with blown out eardrums lol)..
Currently Spotify mastering spec is -2db true peak if the track is louder than -14 LUFS and -1db true peak if below -14 LUFS.
10:20 I think they want the tracks to be -1db at peak is because when the file gets compressed to a codec it gets a little louder and so it doesn’t go over 0db
Yup I've had tracks peak when converting to mp3 from 0.3 db true peak
yeah, it's unpacking the mp3 before it goes to DAC. sometimes you can get close to 2db gain from decoding mp3 to pcm.
Nice video and good tips those you share. With all the due respect, there are few things I don't agree with:
1) some streaming platform (for ex. Spotify) have the option to disable the normalization. I think that in a short time almost all platforms will make it as an option so that no genre of music will be penalized.
2) I would prefer to master my track without thinking about streaming platforms. I make dubstep and the modern songs have loudness like -1 LUFS and peaks up to +5dB. I prefere that my track is penalized by UA-cam and not by a live gig. The risk of "yawning audience" at a dubstep concert is really too scary.
3) if music is still art, it's the artist that decide, not a streaming platform. People can download the mp3 version and listen to utra loud music. But if we start to master according to Spotify, make them shorter and fast hooking for Tik Tok, and so on... I think we lost the meaning of music.
This is just an exchange of opinions, I totally respect your point of view, so please, don't consider this as an attack!
Ok... Correct me please, If I get this wrong... According to my understanding the human ear hears loudness mostly via the average volume encountered.
Above is the reason why a more compressed file always sounds louder in comparison to the less compressed one ie. a track by -10 LUFS with 8dB dynamic range sounds louder than a track by -10 LUFS with 15 dB dynamic range.
Right? So to achieve loud results in comparison with the rest of the bunch means that the quieter parts should sound louder in comparison to the peaks ie. compression should be used... Naturally in a way that it doesn’t make the sound all too flat.
Getting the very last part (compressed, but not flat) from the above is the true skill of a mixing (and in some cases the mastering) engineer.
Right?
Defintly the prefer the dynamic version! The super compressed version gives me shivers just by looking at the waveform. It’s a solid black bar! Absolutely horrible that mastering engineers do this to the listeners. As a listener it makes me cry just seeing that hard clipped loud to the max version. I am glad the loudnesswar is no more and it also should never come back. As a hobbiest I would not want to have that sort of result and go back to the drawing board and mix it again from scratch. Sadly a lot of radio stations trash the sound that includes Veronica which I listen at work.
It's no surprise that the plugin shows +2.x dB at around 9 mins. into the video since your v2 master item had been turned down by 2.4 dB and it was already peaking a few 1/10 of a dB below 0 dB to begin with. This means that its peak level after turning it down by 2.4 dB was probably around -3 dB, thus, provided that the mix complies with the service's loudness regulations, they will normalize it to peak at 0 dB. No "secret" limiting or compression involved - just linear gain. If you don't overdo your mastering loudness-wise, you can upload fairly loud mixes peaking at 0 dB and it will not get touched by the algorithm.
WARNING: as much as I love this channel, this is VERY misleading. Master your track at the recommended Spotify volume and then be amazed and depressed when your song sounds significantly quieter than the next one. The default setting in Spotify is normalization but you can turn it off, and they’re not transparent about the process of “normalizing” at all. If you look in the Spotify for artists section on mastering, they give a vague explanation, but it’s clear that it’s not perfect. Bottom line, all the top mastering engineers still master to a -8 RMS for any mainstream style of music (hip-hop, electronic/edm, rock, pop). The art of mastering at this point is being able to make loud masters that are also dynamic. IT IS POSSIBLE. You just have to know what you’re doing. For dynamic music (classical, jazz, folk), yes, make it quieter. But what’s also never talked about is you don’t even use a limiter in these genres usually (altho a lot of contemporary jazz that’s not off blue note or verve definitely is trending towards a more contemporary limited sound, e.g. Brownswood records). My suggestion is use a maximizer and then 2-3 limiters in series as the first thing you do, and then work backwards. Start at your target volume and then make the master fit into that volume. Why? Because maximization and limiting IS compression. You need wayyyyy less compression than you think once you’ve done this, and then you just sculpt into the limiter. Try it. Game changer. Anyway, don’t believe the hype. Great channel, just a misleading video
Hey Justin, heard the same thing. I am wondering, how do you set up your mastering chain? I've been pushing my volumes up with just one limiter but seems like it's not ideal and I should use many limiters in a chain. I am just not sure how much gain to push in each of those limiters. Say I need +10dB increase in gain, do I slap on three limiters, each one giving approx. 3-4dB in gain?
Excellent vid, looking at most of the angles... Its funny when I went back to some of my historically favorite tracks over the years and started looking at them as reference tracks, I found that the ones that were more dynamic and well saturated with a nice central density always sounded better to me than the ultra loud ,peaks flattened out, solid block of sound like frequency dynamic. When i started mastering tracks more for my own ears rather than like trying to compete with some sort of quest for the loudness gold... mind you there is nothing wrong with that either... its about preference I guess... but no matter anymore if we host on these sites , they get the last hands on it...
The only problem I find on uploading quieter songs to Spotify is that you can disable normalization (I do that) so if you listen afterwards a "normal" mastered track it'll blast you away. In the genre I'm into (Metal) you can't risk sounding way quieter than classic tracks if people disables normalization, and I bet many Metal listeners disable this feature. Besides, at least in the genre I'm talking about, records are still being mastered loud in 2020, loud as 6 LUFS. I'm actually mixing and mastering my upcoming album and I've aimed for 7 LUFS, which is loud, but not as loud as some records I've listened this year. In the end, I decided to aim for a loudness level that doesn't squash the hell of my music just for loudness sake, yet that it can stand to the loudness of the records I've heard in the past 15 years, but more importantly, that it sounds great and when Spotify and UA-cam lower the volume, it keeps sounding great.
Hi there,
I come with a question I asked myself today. We all know about loudness for streaming like youtube, spotify, etc etc…
But what about UA-cam, Twitch "live streaming" services ? What volume are they streaming ? Can't find any answer even on supports. Was wondering if that also could be something to take in mind during mastering as there are more and more dj live streamings during this period.
Cheers everyone, and thanks White Sea for your videos, always a good moment to watch !
Greetings,
Gregory.
I was never a fan of the brick wall limiting of the 2000s. I felt like music lost a bit of breathing. I am happy to see dynamic range has returned.
The thing about this is that when people have the "normalize" oprtion off, you get squashed by other songs if you made a -14 LUFS Master, so still even doe you loos some dynamic range mastering loud I thhink is the best option rn
Elim neden sürekli ses seviyesi düğmesine gitsin? Normallestir düğmesi sürekli açık. Ayrıca Spotify her güncelleme de otomatik olarak onu açar. Ve youtube normallestirme kapanmaz.
Thank you for always making videos and studying new topics🙏🏻
4:18 This version sounds so much better, more "3-D" than the squashed version. I would buy a copy of the dynamic one whereas I'd ignore the squashed one because it has "that sound" that is in so many pop tracks. The dynamic version has a lot more depth.
about time more dynamic range for us that care about things sound good great vid
Excellent info! Thank you for such a clear explanation on this topic!
One thing that gets me annoyed is that, yes, we can master towards being close to the streaming standards, BUT, try turning of the loudness standardising on the Spotify desktop app and listen to something like God's Plan vs a track uploaded to streaming standards (my track Richard Carter - Blow Up, was mastered around -13luffs). His track is way louder without the control, so it seems like big studios don't actually care?
Something that really messes with my nerves is that there is no accurate tool measure that accurately, so they don’t even have to touch the volume at all... I would love it if streaming platforms made it more transparent.
There is more to it than this because there is still a great deal of loudness differences within any single streaming service. So they don't seem to normalize or limit in the way we may think. There is some unknown variable we don't know yet?
well they are even in UA-cam controlling Loudness but with a Normalization technology or algorythm (call it whatever you want) that is horrendous because i can switch to another video of other years or a videoclip or a show and i instantly hear the different loudness and peaks and distortions, the same occurs in Spotify ETC. So, it is everywhere digitally the control of loudness from 2015 or even 2014 but, it's badly processed we need years for this normalization to be great enough.
Men, you can sound louder with presence, boosting 3k - 6k with control. I heard song bad mixed in spotify boosting 10k-12k and sound quieter. Using normalizer of spotify.
High loudness at festivals? I wouldn't do that. Best advise is always dynamic pushed a bit into loudness territory for festivals and loud volume. Live sound engineers (hi!) correct your sound! Also due to sound restrictions/corrections. Which could lead into lower volume or comb filter effects.
I like this and all the videos!!!!
I think that's right but many electronic music producers still mastering their tracks very loud maybe because it's for the style of the track, but I also know that spotify has a selectable volume normalizing so isn't better to be louder?
great video as usual, man. keep up the good work :)
Master every track to around -10 LUFS and call it a day. In this way you won't be punished by streaming normalisation very much, while ur limiter is actually compressing the mix so the mix has enough "wall/brick" feeling at high energy
thankfully.
5:20 PREACCHHHH!!
One would have to hypothesize : so if a streaming service sets their loudness threshold lower, doesn’t just mean the loudness zombies will just try to be the loudest person at that level. Or maybe they just crank that limited -0.1 as usual and let the streaming services dial it back as they will . Either way, I am grateful that this has happened. Yes I love EDM power, and there is a time and place for boom- but it’s nice to see people have to work harder to create contrast in their music and use variation in volume via LUFS as a way of achieving it. Instead trying a different limiter to sqHash a signal into an inky black bar of boom lol
All this means is better variety across the board
I think you meant 1 -dB True Peak instead of dBFS. -1dBFS could in theory still clip when converted to a lossy format.
Great video as always man ! Still you talked about low dynamic music for festivals club etc.. and high dynamic for streaming services, but why is that ?
One problem is that UA-cam Music premium service doesn't compensate for volume. All tracks are at original volume. It's one of the reasons I use the service but it still makes you think about the loudness more than if it was a truly universal practice.
Unfortunately, I think most services do a file that's too loud (so yeah, don't get it too loud). Something which I think is important and oftentimes disregarded is that some platforms only reduce the loudness when it's over the limit; they don't boost it when the loudness is below the level (UA-cam doesn't, probably 'cause you don't want those fail videos with people screaming to be that loud hahahah). On the loudness penalty, when they put the "--" indication instead of a value, that means the file is not too loud, and therefore will say as it is, quieter than other music (unlike Spotify).
In fact the audio from this video is being adjusted by UA-cam, you can check it in stats for nerds 95% / 95% (content loudness -4.6dB).
So when you master above -14 LUFS to try to make it louder, likely with a limiter/maximizer/clipper, then you’re really just squashing the dynamic range, because it’s ultimately gonna get turned down anyways. I think rather than focusing on the loudness, or LUFS, you should focus on dynamic range, and just how it sounds/feels. If anything, aim for -14 so that the normalization doesn’t affect it. What are your thoughts?
Super helpful video... the last part of just making two exports instead of multiple ones was super helpful.
I think what is different these days is that the loudness war was getting ridiculous and the way music was meant to be listened to was getting confused.
Then I think it because easy for everyone to make loud mixes anyway and then the services stepped in and thought "let's try to just make the music speak for its self"...
It also did a bot of a "fuck you" to those people obsessed with it being simply "loud"
Does anyone know if they treat tracks in classical categories differently? Those tracks are likely to be more like -20 LUFS, and turning them up that much would probably ruin them, whether they’re limited or clipped or whatever, but is that really happening to classical tracks too?
(Read other comments, answer seems to be that they do NOT turn volume up, only down? Makes sense actually. Wont kill any dynamics you care to include, but will tone down the loudness wars.)
It's crazy coz it changes the sound. I create aggressive music and heavy limiting gives the track an edge.
yep, for example all the current modern metal releases are at like -5 sometimes even -4 LUFS. In this genre, heavy limiting and hardclipping makes the sound - you get everything "in your face".
I am pretty sure that spotify is still not measuring in LUFS but in replaygain which is really a problem. Also they claim to turn up the volume with a compressor wich is mostly not good as well.
Ooh, wat een nerd ben je toch ook haha; wel grappig en netjes uitgelegd voor de meeste schreeuwers 😉👍
Great explanation, very useful to know
Dynameter is super useful for this kind of metering
What about the perceived loudness? Shouldn’t I be mastering at -6/-9 LUFS?
Thank you
Dynamics are so so important. I love a track with great dynamics. Listen to Billie Eilish for example.
Thank you Wytse! This will save me a lot of time trying to make all my tracks have the same exact loudness! I recently signed up with SoundCloud which, BTW, nobody is talking about. Not even Loudness penalty has SoundCloud in their loudness analysis. Does anybody out there know why? Keep up the great videos!
ozone and a few Mastering chains include Lufs, I thought it was pronounced LOOFS lololol, i appreciate the correction.
Matthew DeWard not sure he‘s an authority on pronounciation 😅
Mekullag 😂😂😂lmao
@@mekullag lol thanks well still way back when I first encountered Lufs, I thought perhaps it was a German measurement as it seemed more accurate?? ( i know thats bad right just assuming) so Ive always said it kind of with that German/Russian l-U-f extra emphases on the U :)
l got a tip from a top mastering engineer who said when it comes to streaming...just master it like you would for CD.
Super informative. Thank you! This helps me a *lot*.
but at the end of the day, as it already happens on youtube, louder stuff catches the listener/viewer's attention more easily...
Erg leerzaam Wytse, dank je wel 👊👍😎
NICE VIDEO THANKS
Thank you - found this very helpful.
This is very useful information. Thank you!
Yeah it does come across on UA-cam. The dynamic one sounds better, quite obvious on my headphones, but oddly enough it also sounds louder to me, meaning LUFS is not entirely accurate. There are other things that unfortunately make LUFS inaccurate. Try comparing a LUFS normalised saw wave, sine wave and square waves for example.
I would like to see more on it!
I do insert Ozone 9 into OBS, job done!
it can get the itunes figure quite wrong sometimes, eg I compare a quieter file to a louder one, and the quieter one has MORE of a loudness penalty for itunes.
This video came at the perfect time
well if you master for cd, the more dynamic one will sound better, always I've always thought that EQing is most important in mastering, not that making it louder part as in case of cd the end user controls the volume, not the mastering engineer
hello all! it is really weird that i have been analysing youtube audio lufs volume and different youtube channels. there is a music that are (integrated lufs) about -7 to -11 lufs in youtube videos so i think whats going on ? some/mostly tracks are about or close to -14 lufs in youtube but some music tracks are way more louder than youtube's loudness reference -14 lufs. some youtube channels have very loud and louder than youtube's -14 lufs settings (integrated lufs) and the loud tracks that i have been analysing are -7 to -11 lufs mostly...so all people can't get same lufs settings and thats really annyoing. some big channels upload a music track and that is be like -7 lufs and when small youtube channel upload a music track it will mostly be going to youtube's -14 lufs setting and it is really annyoing! youtube lufs volume and not all people are equal...sorry for my english...😀😎🥳 I noticed it when I analyzed the volume in youtube with different music tracks and small + big channel 😥😥🤔🤔 and this issue is hardly talked about at all...
I fully agree with thi comment. We all know this information in this video and still I m very grateful for all the free advises you give, thanks so much, keep doing this, but I'm a realist too!! We already make sure we master around -12.5 lufs and stay under final peak -1, but all the tracks I released so far were not as loud as the tracks coming from the big music factories, so I think there is another proces going on besides the so called algoritmes, something with "money" I would guess. Can you comment on that?
Nicely explained. Stay well.
It's another freaking moving target. That's apparently where we are right now.
I heard that I have to send a track to Spotfy at -1db peak. That means my last brickwall limiter should have a -1db celling? I always use a celling of -0,3 dB. Please, I would love to understand that.
Even with all of this loudness compensation going on I still think (it sounds to me) they are running the loudness of the commercials higher than the program content. Just like the volume of the commercial in this video, I turned the volume down because of how loud it was and back up after the commercial. Is this actually happening or is it just me? Have you ever tried measuring the difference?
Spotify are running ads louder and its like the most irritating thing in the world.
But why wouldn't you use the more dynamic version at the festivals too?
Take out like 1 db with wide Q eq around 1500 hz and spotify wil turn it up instead of turn it down.
Awwww we both released mastering videos today! Awwwww so cute! LOL