I fell asleep watching TH-cam on my iPad, and this video was playing when I woke up yesterday. It was so laid back & mellow. Good thing it was Saturday, I might have missed work.
@@markmcclellan8421 if you invert the audio wave and play the non inverted wave with the limiter on one you will hear the difference between the two audio files, meaning you will hear only what your limiter is doing
@Gabriel Freitas Get two tracks, one under the other. One before limiting, second after limiting. Invert the phase of one of the two tracks. Some DAWs have invert phase button on every track, or you can use some plugin that has that feature. Inverting phase makes the waveform flip upside down. When it went up, now it goes down. Two identical tracks when flipped will cancel themselves. It's like having +10 and then adding flipped -10, the result is 0. When one track will be limited, everything will be canceled apart from the work made by limiter.
Soldat don’t forget a null test will only fully cancel if the waveforms are identical, any random modulation will be different so you will also hear that too, try inverting the 2 bounces of the same mix, EG if there is a reverb with an LFO of some kind in it you’ll hear that difference.
This was a very interesting presentation, but kinda sucks that he got rushed for time towards the end. I feel like he possibly had a plethora of more information he could have provided. I would love to see a "full length" un-rushed presentation of this material. It would be great if Alan could record the full version and upload to youtube. I feel like it's an important message to get out there.
Totally agree, towards the end I felt like he was going to talk about how this would translate to genres where compression and saturation have become 'part of the sound', but it was then just rushed over quickly without much information regarding this side of loudness. I think especially today this is actually a very interesting topic to talk about since a lot of electronic genres do rely on smashing things. Still, this was a great video though, it actually showed in a pretty easy and understandable way what the whole new loudness normalization means and what it does. However since I come from a background of 'heavier' electronic music and grew up making Hip Hop, I do wished he continued more on the topic of what it means for those genres and how to tackle it.
@@artfxdnb I guess i've infered what's the point. Even though we do make music that is comp-heavy, that's part of sound design, while we can change some of the decisions on the mix/master part such as limiters or comps or saturator to gain volume, and light final limiting. I'm just trying to do so.
It seemed to me the message was "Forget everything you knew about making yourself louder. That exploit has been patched, now all you're doing is making yourself sound small and shitty." The beauty of it is that the solution is to return to _doing nothing._ Use whatever dynamic range you need to use. Compress because you want a certain sound, not because you think it's required. Then stop. It might seem that taking a step out of the workflow would actually be welcome, but like reverb on a bad vocalist, it's possible to hide a lot of sins behind compression. If you have been, now you have to fix them.
@@mal2ksc compressors can finally go back to their intended use case. We love to put compressors on every single track, but now we don't HAVE to get it to a certain point. We can actually put as much as the track needed and maybe even take it off because it wasn't necessary. This opens up so much more options for creativity gotta love it.
He deserves a grammy. One of the best explanations on this topic. 30 minutes was too short. I wish he had the time to really show how to use those loudness meters / plugins.
@@IcidLink Sadly, some labels (UMG) have been compressing the crap out of even acoustic jazz releases in the 21st century. John Scofield's recordings come to mind. His Blue Notes from 1990-1992 with Joe Lovano had a very high dynamic range. He reformed this band around 2015 and put out an album on Verve that sounds much less "alive" than digital recordings released 25 years earlier, with considerably lower dynamic range. Small labels are still releasing jazz albums with great sound, however.
My mind is blown. I feel like one of those soldiers who's been hiding in the woods and didn't know that the war is over. Thank you Alan! What a great presentation, wish I could hear the rest!
11:06 is the key point, 12:19 pop dynamic rang, 12:57 peak normalization, 14:12 loundness normalization, 16:43 new targets, 24:19 different level demo track, 25:13 Level all the same, 26:10 what happened, 28:00 VU is getting back!
Having known Alan for 40 years, I was very happy to see him again, if only virtually. Alan's musicality has always brightly shown through on every project we have worked on together. Alan, thanks for sharing your insight into a complex, changing and sometimes confusing subject.
They limit different harmonics in the track in order to create its "Sound" essentially blending all the frequencies together. Some limiters focus on high end transients better where others focus on the midrange or lower midrange. Thats why some limiters can sound warm or bright. So they are all doing the same thing essentially just balancing the track out differntly if it makes sense. Turn up the music and when you feel it, then you will hear the difference.
I always mix with a VU meter. Everything gain staged to -18db on the individual tracks before I mix and a VU Meter on the master buss calibrated to -14db and the records really do sound so much better when the final mix is done.
@@yassinetalbi5884 couple of reasons. 1) your mix will never come out too hot, plenty of headroom 2) less work in mastering needs to be done compressing and limiting to get up to commercial levels while still giving them lots of room to work 3) your plugins on the master buss are still being hit in their sweet spot and your not blowing out compressors and EQ's with level
@@yassinetalbi5884 I find anything less than -14 on the mixbuss results in an extremely quiet mix and any noise floor will be present in the final product (when gaining tracks to -18). And anything above -12 to -14 results in peak levels reaching close to zero and doesn't leave mastering a lot of room to work
@@yassinetalbi5884 A VU meter calibrated to -14db will read 0 when your RMS (Average) level is -14dbfs. Your peak level will be determined by how dynamic your mix is, instruments used, compression, etc. A real variety of factors. As a general rule of thumb though you're probably right and your peaks will fall between -10dbfs and -5dbfs, give or take a bit
Amazing! The part where he shows off the different limiters and what they're doing "under the hood" blew my mind. Really helpful stuff, it is always appreciated!!
Forced loudness leveling basically wins the loudness wars after 3 decades of pop that has been limited so much it ended up pretty much dead sounding and with unbearable kickdrum pumping. So we finally get our dynamic range back. Proves me right for _not_ compressing the shit out of my mixes even at the cost of them sounding quieter on a standard stereo system. If you want it louder, use the friggin' *volume* dial.
I think the idea he was trying to convey is that forced leveling is not an issue that has to relate to loudness if working with transient shaping and balancing so that the track doesnt feel so quiet.
@@jasonsteves734 essentially yes. The dB range has been more limited, but it does not mean we are incapable of combatting that issue by realizing loudness does not necessarily translate to a higher dB. It also has to do with the frequencies boosted, the transient shaping, etc. I think it is better for us because especially things like classical music had such a dynamic range that volume levels changing so fast made it impossible to listen to certain songs without having to change the volume part way through
The thing that most forget it s that compression/limiting should be done by the end user. If they are in noisy environment or need to listen to it at nights, at low levels, the limiting should be more aggressive. And when they are listening in quiet environment with more sound pressure, limiting should be non aggressive. The volume knob is actually the wrong adjustment for most needs. If you ever had the "pleasure" of mixing bespoke to one room, one system, one environment, it becomes quite obvious. You can get to perfection when it is used only in one place and the environment is static, doesn't change (for ex, art installations, theater etc). What most users need is not just gain, it is both limiter and gain. Luckily, more and more customers do have dynamic range control, sadly they are very simple but it already makes our job so much easier. You do not have to master for headphones used in a sub/metro, the listening device has thought of that and does the final "crush" itself. What we need is a standard that all manufacturers need to follow.. Just simple none/mild/medium/strong limiting defaults, done the same way in each device. If we had that, we could easily give our mixes and masters more room the breathe and we can test the results in a standard limiter to make sure it works with all the defaults.
To anyone else who has been as obsessed as myself about finding that version of "Swing! Brother, Swing!" It's Catherine Russell. From her album Harlem on my Mind. Listen and soak up the mix and the jazz.
Coming from a film audio background, I have been using minimal compression in my music projects as well, out of practice and frustration with the modern limited sound, so this gives me some serious hope!
This is very true and informative video. As a producer myself being in the industry for over 8 years now, the massive change of how " Mastering " is perceived definitely varies from artist to artist. Do not forget that every genre, every specific sound deserves a completely different touch in order to make it stand out as the producer wishes it to be in result, so even if the particular style is meant to be very powerful, it does not necessarily mean people have to make it sound loud. I see the beauty of mastering in actual manual control, where every sounds is carefully crafted, tweaked and monitored so the result will sound as best as possible without having to simply let the automatic settings bleed it out. So the whole auto-normalizing might be helpful only to producers who do not master at all so in the very least their tracks do not blow the listener's speakers, but other than that, it sure does damage the quality if that same technique is done to " Already Mastered " track as it becomes less pronounced. But more of a reason to get it in the best quality so that goes away
@@Jz-sv1ju Normalization being applied by streaming services is not at all the same thing as mastering. Alan Silverman, in this very video, specifically states that loudness is subjective and influenced by many more variables than simply "level", and "level" is the only thing normalization is really affecting. Mastering itself creates tonal and dynamic changes in a piece of music that can be quite pleasing, and every track should be mastered, but depending on the genre some tracks will need very little in the way of dynamic reduction, and some will absolutely depend on that dynamic reduction to give it the sound that genre is known for. The point of all of this is that if you master your music above target LUFS levels set by various streaming services, it will be turned down by those streaming services, and that past a certain point you will have reduced dynamic range without the benefit of any perceived loudness on these streaming platforms. Normalization does not replace mastering, it sets a ceiling for it and discourages the continuation of the loudness wars.
you watched the video yet you still don't appear to know how loudness normalization works. It does not affect the quality of your track, mastered or not.
I wish this were longer - I could listen to this guy for hours. If there's one good thing to come out of streaming its that it has pulled the rug from under the loudness war.
This all works to a certain extent. All my albums were mastered at -9 LUFS. Then I learned more about this topic so I decided to master my last album at -14 LUFS. Result? On Spotify, my last album sounds quieter than the previous ones. Now I think I've done everything right, I also involved an external mastering engineer to make sure everything was right. But if I could come back in time I would master my last album at -9 LUFS as well.
Thanks so much for sharing this story Fabio. To a degree, it depends to a degree on the genre, EQ curve and style of mixing, I think. If the mix is already pretty compressed loud, turning it down won't make it louder on streaming services. Alternately, if there is a lot of low end in the track, that can make it sound relatively quiet compared to other tracks at the same LUFS level that are brighter. There are other factors I could go on about here. As a mastering engineer myself, I always try to do my last pass of level setting by ear, comparing to relevant references. For me, meters are a guide and never the final decision maker. Hope that helps!! -Justin
Just to be clear, did you master your tracks peaking at -14dbfs or you mastered in -14LUFS. 80% of the time people get confused with this so im just making sure. Because a track mastered at -14LUFS cant really be lower in loudness if a -9dbfs master is normalized to -14LUFS. If you really did mastered in LUFS, most likely your 1st album was more dynamic in the 1st place compared to the 2nd one, A highly compressed and crushed peaking at -14dbfs will always sound quiter than a highly dynamic track mastered in -14LUFS I mean after all -14LUFS can have parts peaking at -6dbfs compared to a crushed -14dbfs whos limited to just -14dbfs. So i just want to know if its really -14LUFS or just metered in -14dbfs.
@@pelbriks wow tha'ts weird now I guess. Was this spotify? Was the -9LUFS master also peaks at -0.2dbfs? if its, then I could guess there is something wrong with spotify with their algorithm now, thats a bummer. cause both should be just as relatively loud at -14LUFS. cause im assuming the dynamic rage would be just the same(of course other things takes into account). Thanks for the tip, I guess people should be careful about spotify cause it has that problem. Honestly in my country I usually have a problem with this specs cause networks here are still using the old standard, so at some point using LUFS doesnt even matter to them, they just want something that peaks at -3dbfs mono!. They dont even care about EBU R128, I actualy caught one network engineer who doesnt even know EBU R128 guidelines!
Greg Rodriguez III I’ve no idea 🙂 but yes there’s something weird for sure. I dig Justin comments though, maybe more bass in the mix can lead to a different apparent loudness
It's very heartening to see this finally happening. My only concern is that the next loudness war will be fought on perceived loudness and give us harsh, mid-heavy mixes and masters. But I think that will be harder to get away with.
I have started turning my masters down. I mostly only record my own music with the odd project for other people and I use Ozone 9 for mastering. I have consciously started mastering to a lower level to retain dynamics in the music, especially on a ballad. To my ears, it sounds better and if you want it louder, you can just turn it up :)
This was by the far the most interesting thing I've watched all week, maybe in the last month. I don't even mix or produce. I just love music. Damn shame to cut it short.
THE FUTURE OF MASTERING proceed to talk about history and present. When things were starting to get to the actual subject of the video it ended. Please give this guy more time!
They should either start their own label or negotiate their contracts better. I don't pity adults having everything done for them and then complaining they're not multi-millionares.
and that's too much in their eyes. They will keep raising the price just to upload to streaming platforms. At one point, it will be art for those that can afford it, making no money from it but paying to have their art heard and gain prestige merely from the popularity.
Alan is truly one of just a few in the world who are absolutely the best. If only his kind of genuine expertise was the norm in in the world we’d be so much better off. Bravo Alan!
this was great and overall informative, but I hope he realizes that on Spotify users can disable the normalization, which basically makes it where the difference in LUFS from track to track will vary accordingly. he was kinda speaking on it as if this is a global permanent change, but it’s definitely not. still useful to be aware of how it works when it’s enabled... I believe it defaults to being enabled and most ppl probably don’t change the settings but it’s definitely possible to disable
True! Most people will never touch or even look for that button. But it is there, hiding in the shadows. Anyone willing to look for it, is probably the kind of person who is comfortable using their volume knob though :) -Justin
99.9% of consumers definitely have no clue that it even exists, let alone what that even means when they see it. But yeah, just looked and it’s there! Shhhh! Don’t tell anyone!!! 😉
15:40 THIS is what we had since after World War 1. It was called the VU Meter. Zero was about two-thirds up the scale, with a peak positive value anywhere from 6 to 18dB VU, and negative values from -12 to -infinity. It was when digital recording began to accompany and ultimately, replace analog recording, and Peak metering replaced RMS, that any semblance of standardized levels went out the window. Hopefully, loudness metering(LUFS, or RMS VU meter emulation), will restore that relative normalcy
Gregaras The Greg bahahahaha. What do you think limiters are? You’re squashing a signal which guess what? Creates distortion... not clipping. Distortion does come from clipping but it also comes from saturation, compression, using lots of drive (which is increasing the signal beyond 0db then reduces the loudness so that it doesnt blow your ears and speakers) do you even mix?
That chart showing cinema dynamic range explains a lot. There are so many movies, that when watched at home are terrible. The actors mumble at such a low volume (in that fake Hollywood gruff voice), what they're saying is inaudible. I always rant in my head 'what kind of sound engineer moron has done this'. It ruins the movie in a home environment. I think I need a compressor on the TV output.
And then comes sound effects, car chases, music, etc and your ears are bleeding. But barely hear what is been said. Same goes to most TV programs/shows and also so many youtube videos. Music is many times way too loud compared to conversations. Have to keep adjusting volume.
This is a common problem. The "night mode" on a lot of digital receivers is a compressor. My DVD player also has a dynamic range compression feature tucked away in a menu.
I really don't think so, the difference is distortion yes, but because the source envelopes are a different shape to the target envelopes you never hear this type of distortion, because you are never comparing the envelopes. So to say you can hear the distortion is pretty much like subtracting a song played on the piano, from one played on the guitar and listening to the only the harmonic differences (Which could be interesting.. now I think about it 🤔).
@@stuartwilson4960 hmm im not sure if your right. He literally states that we are hearing what the limiter is clipping out while still blending in the clipped information into the original track. Am I mistaken? I can time stamp his statement if you wish.. and I, unfortunately, believe him more than you lol
@@Hugoknots If you listen to the difference of two digital signal processing functions then you would literally hear distortion, but how is that a surprise? It doesn't mean there is distortion there, it means the difference sounds horrible. Just like you applied two limiters with different envelopes to a pure sine wave with some resulting phase offset, the difference between the two will sound impure and horrible, however listening to the results of either independently will sound perfectly pure and non distorted. Also, some of these limiters are multi-band, so you should and would expect harmonic differences between the limited and non limited versions. Which is also not distortion, it is just purely a different envelope shape than the original.
The distortion is definitely in the mix. When the mix still sounds clean it just means the distortion volume is so low, that you can't here it conciously
@@nathanconholio1863 I don't think you can understand, if you had two versions of a just a pure sine wave, at different levels and different temporal envelopes, both independently would have REAL phase distortion maybe (0.000%) however, differentiating them will produce something that sounds like distortion, because you hearing the overtones on a set of envelopes. It is not distortion, just a different level of envelopes and harmonics.
17:35 throwing shade at Rick Rubin ahahaha!!! I am happy that music I have mixed and mastered since 2008 has honored Dynamics/maintained transients. My stuff doesn't sound dated because of good practices.
@@Ascoundrel agreed. Rick may be past his prime and make very questionable engineering decisions but to say that he is a bad producer would be incorrect. His portfolio speaks for itself.
This has been great. New challenges for musicians where one simply can’t afford not to up technical understanding of sound and sound editing’s latest developments.
Every genre has its own signature dynamic range. Ironically, only when rushed at the very end does Alan mention this, so this important concept isn't given the weight it should be. Too many people hear presentations like this and walk away thinking "I should mix and master to -14 LUFS or -12 LUFS, to take full advantage of the dynamic range at that level", when that isn't really the case at all. For jazz and other quiet genres? Sure! For EDM or Pop? Nope. Your song will sound thin and tepid and weaksauce if you do that.
exactly, it was only at the end when he was finally about to talk about what the next step should be to help solve this then he was cut off and it ended smh
Excellent video. Concise explanations. This change in allowable loudness was one of the primary reasons I created the service called: Spectrelayer. The industry was changing and the time had arrived that allowed advanced tools (such as AI) available to optimize, & repair material damaged by the "Loudness War". And it is STREAMING where the schema I created really shines. For a fraction of the cost of remastering - Spectrelayer uses advanced heuristics & "machine hearing" under the guide of a competent engineer to act as a POST-mastering stage to rebuild the audio released by the studio into a more dynamic, less distorted & more tonally balanced end waveform. The result is a track that is much less likely to be reduced in loudness by the streaming service and as a result will play with better overall sound quality & superior dynamic range - than the studio release would have without Spectrelayer. Visit & hear for yourself! Spectrelayer Rocks
thank you very much for this, as a mastering guy myself, I have the problem that I mostly master DJ music, and DJ's want it loud, but streaming don't need this. so, in the end, we will need to make 2 different masters.
Robert Babicz Stoaked to see you around, you just mastered my album few days ago and I can’t tell you how happy I am to had this done by a true professional 😁 🎶 It sounds super!!!
Did you know blinkers in cars are the same? The sound used to be mechanical. Now it's just there cuz that's what we expect and it also prevents us from leaving the blinker on.
Absolutely such a great video about the history of Loudness till now. "They have taken away Level control but given you more headroom". This went off like a bright light bulb in my head. Fantastic.
The loud mastering is only useful for smartphones that limit volume to avoid lawsuits by parents of deaf teenagers. For an audiophile, ANY limiting at all is a departure from reality. I got to hear a 2" RCA master tape on my system and I was stunned at how REAL it sounded. Playing these recordings on a hi dynamic range system is an experience like hearing the band in your room. Now today's music is so compressed that it is fatiguing to listen to. White space is so important in music. Without it, it loses the element of surprise which keeps it interesting.
Limiting can make sense when producing/playing live elecronic music, I killed a couple of PAs myself because the owners didn´t expect they can be melted by synthesisers.
@@7k7yub7 That's a problem of an inadequate PA, not your problem. Recordings with compression are fatiguing to listen to. Having heard a master tape recently, I was astonished at how much more "live" and fatiguing it sounded compared to the release on CD.
@@basspig obviously you are right, I just wanted to make an argument that RIGHT sound engineering is important and that in the wild there is more then black and white.
Absolutely, my friend! My ears find hard limiting and heavy compressing very offensive, especially when used on entire mixes. I just love sound of instruments with natural texture and space. For the typical material I work on I find that around -16 LUFS is the "liberation point" were the peaks are just naturally sitting below full scale (particularly if I've done good mic placement and instrument input), with maybe just a couple stray single moment peaks. It feels liberating to stop worrying about head room and just focus on the aesthetic of the mix artistically, and not have people asking why the mixes are so "quiet" when I chose to ignore the loudness war. Sad that it took so long to circle back to the gloriously large dynamic range we should have been able to keep from almost 40 years ago!
@@musicman3569 People used to say that about my Bridgeport Symphony Orchestra recordings.. absolutely ZERO compression, for a totally "you are there" experience. In pop music, a little compression on individual instruments like kick drum or bass guitar can be used to fatten certain sounds, but overall compression is the devil.
This is very important to know. This explains why some videos are getting quiter after uploading it on TH-cam. I also noticed that the sound on youtube seems sometimes much quiter on mobile phones than if you are watching on PC. I'm not really sure if it's because of the smartphone or probably the loudness is normalized much more down for devices where you are usually using headphones.
This was really good and Alan is a master of his craft. To me it begs a question - how many of todays mastering engineers not in their 50s even know the right way to do it? I bet most pop producers will still smash the shit out of their music - and let Spotify just reduce it, vs going through this whole reverse learning curve to arrive at a master that is delivered at the place it should have been all along.
What a great presentation. I’ve learnt a lot about defining loudness and its cost over the last two years, but the breakdown of the three limiters inducing distortion was particularly eye opening.
@@kopczas Oh, and don't forget Death Magnetic and 13, among many others. Corey Taylor even stated that he never wanted to work with Rubin ever again after Slipknot recorded Vol.3: The Subliminal Verses.
Great video and info. There is some impracticabilities to this though. People still buy and download tracks, so the LUFS of those tracks have to compete with their playlist of other downloaded tracks. Therefore you have to still master for loudness. In an ideal world, streaming services and music distributors could allow you to upload 2 versions of the track. One for streaming and one for download, but that's not the case. Also the sound of the limiter and squashed dynamic range is a factor that gives music a 'modern sound'. It's what people have become used to.
Great notes Graham. I agree! I tried to speak a little bit to that last point of yours in my intro. For my own mastering, I always do my final level setting by ear against great sounding references, rather than against the meters alone. But what Alan's remarks and examples do for me is make it so that I don't really ever strive to be "the loudest" of all the references. Rather, I try to find a spot that sits well across platforms, and across all relevant comparisons. That's a good way to go in my opinion. There are so many factors that effect loudness, including timbre, which meters don't always do an ideal job of compensating for. But they are a helpful aid in getting things roughed in to start. There is the possibility as well of having different files for different platforms and mediums, but this is probably out of scope for most smaller artists and likely isn't worth the trouble in many contexts. Hope that helps! -Justin
@@SonicScoop Thanks for that Justin and for your excellent effort with your channel. I wasn't aware you could have different files for different platforms. All my clients always tell me they don't have that option (Tunecore etc). But like you said, it's probably not applicable at the lower end of the market, which is where they are. I must confess, I never use VU meters (I'm not sure if I should?). I always ensure there's no clipping along the route, then when I come to master, I drag in 3 or 4 reference tracks and A/B the various frequency bands so I know I'm in the ball park and use an LUFS meter, to adjust. The problem with mastering your own mix (as you've pointed out in previous videos), is that you're still in the same room, with the same room anomalies. Using reference tracks helps to somewhat negate them.
Thank you, this is a wonderful presentation of a real problem and the solution. I personally like to go for a -12LUFS when mixing and in mastering not going much louder. And it works perfectly. Still there is some resistance still from engineer that spent a better part of their lives learning how to smash the songs and now they are "lost in headroom". This was a real revolution and I love it!
22:33 How it was shown in the presentation was a bit missleading, as he suggested, that the "cghr, cghr" distortion is audible in the mixed signal. It's not true actually, since it's just a Fourier transform that subtracts the peak amplitude from the signal. You hear the sum of both frequencies, so there is no "cghrr" emmited by the speakers nor it is present in the waveform. In order to hear this distortion you would need to have it shifted out of phase or reversed in polarity. Of course limitting can sound artificial and too heavy limitting is terrible, but it's not due to added, audible distortion as it was presented in the talk. You could argue, that any amplitude lowering "creates distortion", since you need to subtract the signal that is already there. I.e. - when you calmly mute a guitar string with your palm, you actually add a lot of terrible frequencies on top of the string vibration, which in sum just results in a silenced sound. If you somehow played back this "muting" action in reversed polarity it would sound like a terrible noise, not at all as a gentle string muting.
This is not true there is no Fourier Transform involved, just phase cancellation of one signal with a like signal that is inverted aka null/nill test as he said. The cghr, cghr distortion is probably clipping, as most of these "limiters" are really clippers. well a clipper after a limiter stage anyway. He didn't show the envelope settings for any of them so the more distorted examples where probably limiting less and clipping more (longer attack?). The distortion would be audible as well in the original signal it wouldn't be in the resulting null/nill test otherwise.
@@nathanisherwood5695 As far as I understand these algorithms they use Fourier transform to find the base frequencies which after a superposition cancel the complex waveform with an inverted signal. 1st a Fourier transform to generate the base frequencies, then the subtraction of the signal (or addition with different polarity depending on how you'd like to phrase it). My formulation in the comment wasn't correct indeed, as I suggested, that the Fourier transform does the subtracting.
@@bzqp2 no there is no fourier transform... why use a fourier transform frequencies have nothing to do with this process... a linear phase eq or a multiband compressor that has linear phase crossovers yes a kind of FFT is used here (to correct phase) but this example is only single band limiting. its not frequency dependent it applies the same to all frequencies.. its just a gain adjustment and nothing more :)
@@nathanisherwood5695 The whole point that was made in the presentation is that gain adjustments produces actual waveform with reversed polarity with regards to the original signal. Gain adjustment has to be exactly on point with the frequency if you want to only affect the amplitude of the signal. To paraphrase - if you want to lower the volume of something, you need to produce such superposition of sine waves which exactly mirrors the original waveform.
@@bzqp2 no that wasn't the point at all... it does not produce an inverted waveform (reversed polarity)... he did that deliberately after the processing just to do the null test. and no i don't know why you think you need to generate a negative sine wave to lower the volume of a sine wave no you just apply a scaling multiplier to all sample values... ie.. if you want to half the volume of a signal (sine, program or otherwise) you just multiply every sample value by 0.5 (or divide by 2).
You're right about everything but every time i listen to those argument, i want to say that there is some music who need this compression and this limiting even with loudness normalization. Those style of music are played on the dancefloor and need to be powerfull and loud as possible, just listen the youtube file of a lot of edm, dubstep or hardmusic, they are around -5 lufs integrated and even more for certain songs. Yes the loudness war is over but some style of music will need this loudness as far as they live and i've never hear that on any topic or video. I thinks it's usefull to precise those things for producer who produce those style and listen to great engineers like this guy.
Amazing video. I've been doing this since 2002. Years have nothing to do with experience though. I learned a LOT from his presentation. For the past 10 years I have worked in radio. I have my own radio show on AM radio (I'm a unicorn, I know). I have called every bit of my music mixing and mastering knowledge into how I prepare shows for air...and it's nice to know I was on-track all along without fully knowing it. For example, where other hosts obliterate their audio to 0db, I purposefully restrain myself on compression and limiting. My "mastered" show peaks at -1.0db while I keep 12db average minimum of dynamic range throughout. Why? Because in radio, the source audio will run through a gauntlet of compressors and limiters before it ever reaches the antenna. I purposefully leave headroom to compensate. So by the time it gets broadcasted, it's not "overly" compressed and limited by the time it reaches listeners' radios. My show is mastered for radio. I feel like the very nature of mastering for radio has actually kept me ahead of the streaming curve without realizing it. My podcast uses the same audio normalized to -1.0db with around 12db of dynamic range. Turns out, my rule-of-thumb 12db of dynamic range is pretty darn close to ideal for streaming for largely similar reasons attributed to radio broadcast. The streaming algorithms "average" everything out between sources just like radio has been doing for a century. What's old is new again. I just think it's all so fascinating.
![Mick Guzauski Masterclass on Pop and Funk Mixing [ft. Jamiroquai]](http://i.ytimg.com/vi/cFx4gBSwM3g/mqdefault.jpg)
![PURPEECH - นี่ฉันเองคนที่... (It's me) [Official MV]](http://i.ytimg.com/vi/NuhHd0hOOg8/mqdefault.jpg)
14:00 the Streaming services don't seem to care much about the listener's health when they play their ads though.
Lucas N yeah how the hell can they blast those at near double-volume
you can install thing called adblock and stop whining.
So in the end the streaming services win the loudness war no matter what lol
@@petesmith9475 Which you can't really do on smartphones. Also you could just pay for the service.
@@kurtgerd7886 Patched apps, but yeah. Pay for it.
My girlfriend called me halfway through this and I never felt so much interrupted in my whole life
Hahahaha
A quiet guy speaking softly about loudness. 👍🏼
And exactly what he was saying, less is more. Shouting means no one listens It's that subjectivity he talks about. And that equates to compression.
David Paul PRACTICE WHAT YOU PREACH
Walk softly and carry a big limiter.
You folks have no clue what the word means
God. Yes....
i haven't been paying attention to what he's saying but the sound of his calm voice has improved my mastering skills
yes
this comment needs an award!😹
haaa😂😷👌🏽
@Don’t Ask What Kind Of Music absolutely!
I fell asleep watching TH-cam on my iPad, and this video was playing when I woke up yesterday. It was so laid back & mellow. Good thing it was Saturday, I might have missed work.
Phasing tracks to hear what Limiters are doing is a cool tip
What do you mean by phasing tracks? Do you mean putting tracks out of phase? Or putting them through a phaser? Or what?
@@markmcclellan8421 if you invert the audio wave and play the non inverted wave with the limiter on one you will hear the difference between the two audio files, meaning you will hear only what your limiter is doing
@Gabriel Freitas Get two tracks, one under the other.
One before limiting, second after limiting.
Invert the phase of one of the two tracks. Some DAWs have invert phase button on every track, or you can use some plugin that has that feature.
Inverting phase makes the waveform flip upside down. When it went up, now it goes down.
Two identical tracks when flipped will cancel themselves. It's like having +10 and then adding flipped -10, the result is 0.
When one track will be limited, everything will be canceled apart from the work made by limiter.
Soldat don’t forget a null test will only fully cancel if the waveforms are identical, any random modulation will be different so you will also hear that too, try inverting the 2 bounces of the same mix, EG if there is a reverb with an LFO of some kind in it you’ll hear that difference.
@@pacman_pol_pl_polska this is exactly how its done. thanks for clearing it up, a lot more detail then my 1st comment.
This was a very interesting presentation, but kinda sucks that he got rushed for time towards the end. I feel like he possibly had a plethora of more information he could have provided. I would love to see a "full length" un-rushed presentation of this material. It would be great if Alan could record the full version and upload to youtube. I feel like it's an important message to get out there.
Totally agree, towards the end I felt like he was going to talk about how this would translate to genres where compression and saturation have become 'part of the sound', but it was then just rushed over quickly without much information regarding this side of loudness. I think especially today this is actually a very interesting topic to talk about since a lot of electronic genres do rely on smashing things.
Still, this was a great video though, it actually showed in a pretty easy and understandable way what the whole new loudness normalization means and what it does. However since I come from a background of 'heavier' electronic music and grew up making Hip Hop, I do wished he continued more on the topic of what it means for those genres and how to tackle it.
@@artfxdnb I guess i've infered what's the point. Even though we do make music that is comp-heavy, that's part of sound design, while we can change some of the decisions on the mix/master part such as limiters or comps or saturator to gain volume, and light final limiting. I'm just trying to do so.
It seemed to me the message was "Forget everything you knew about making yourself louder. That exploit has been patched, now all you're doing is making yourself sound small and shitty." The beauty of it is that the solution is to return to _doing nothing._ Use whatever dynamic range you need to use. Compress because you want a certain sound, not because you think it's required. Then stop.
It might seem that taking a step out of the workflow would actually be welcome, but like reverb on a bad vocalist, it's possible to hide a lot of sins behind compression. If you have been, now you have to fix them.
@@mal2ksc compressors can finally go back to their intended use case. We love to put compressors on every single track, but now we don't HAVE to get it to a certain point. We can actually put as much as the track needed and maybe even take it off because it wasn't necessary. This opens up so much more options for creativity gotta love it.
The loudness war claimed another victim, lol
He deserves a grammy. One of the best explanations on this topic. 30 minutes was too short. I wish he had the time to really show how to use those loudness meters / plugins.
indeed, I would like to see hours of this
He’s actually got quite a few grammys
Alan masters ALL of our recordings and he is a certifiable bad ass through and through!
Listen to what I’ve created n mixed my self does it sound loud enough.
th-cam.com/video/WiILtL0dzUQ/w-d-xo.html
Now I'm impressed!
I even heard the difference samsung phone 😆
@@DIVERGENCE777
Yes??
That dynamic range of the jazz track really made it sound interesting and pleasant to the ear others were boring
Yeah Jazz is amazing in this regard. The problem is the most modern Pop Music is Over Produced Quite bland Music with no dynamics whatsoever
@@IcidLink Sadly, some labels (UMG) have been compressing the crap out of even acoustic jazz releases in the 21st century. John Scofield's recordings come to mind. His Blue Notes from 1990-1992 with Joe Lovano had a very high dynamic range. He reformed this band around 2015 and put out an album on Verve that sounds much less "alive" than digital recordings released 25 years earlier, with considerably lower dynamic range. Small labels are still releasing jazz albums with great sound, however.
My mind is blown. I feel like one of those soldiers who's been hiding in the woods and didn't know that the war is over. Thank you Alan! What a great presentation, wish I could hear the rest!
had this EXACT thought. EXACTLY.
Really cool for the NBA commissioner to give audio engineering advice on his free time
Lolllll
as reggie watts would say: "... and there is a profoundness... to this loudness... that is especially hard... to come down with"
11:06 is the key point, 12:19 pop dynamic rang, 12:57 peak normalization, 14:12 loundness normalization, 16:43 new targets, 24:19 different level demo track, 25:13 Level all the same, 26:10 what happened, 28:00 VU is getting back!
Having known Alan for 40 years, I was very happy to see him again, if only virtually. Alan's musicality has always brightly shown through on every project we have worked on together. Alan, thanks for sharing your insight into a complex, changing and sometimes confusing subject.
You can tell someone is a true soundengineer when he wears sonic protection on stage while giving a seminar... Thumbs up for him!!
It feels like a decades-old exploit has been patched. Finally.
Loudness war was the worst low moments in music History.. now we won. We gained headroom
Now everything is leveled except the ads :/
'bout freakin' time!!11!
:-)
Yes. Still a problem in live performances.
Still a problem in nightclubs
Still a problem, sadly, just listen to most contemporary music.....but yeah, maybe something is changing
As a mix and master engineer, this is a conversation i am continually having with artists. This is a great talk, thank you so much for sharing
"all the limiters sound completely different" me sitting here not hearing a single difference lmao
Me too exactly hahaha
You guys must not mix music lol
I had the same and my thinking was that I am listening through youtube and on a bose speaker.
Mattéo Perron if you were there in the auditorium, you probably would. But across TH-cam, no way
They limit different harmonics in the track in order to create its "Sound" essentially blending all the frequencies together. Some limiters focus on high end transients better where others focus on the midrange or lower midrange. Thats why some limiters can sound warm or bright. So they are all doing the same thing essentially just balancing the track out differntly if it makes sense. Turn up the music and when you feel it, then you will hear the difference.
Yoooo, my mastering engineer learned under Alan Silverman and speaks VERY highly of him. Listen to everything this dude says.
I did not know that Michel Foucault was still alive and that he knew this much about music!
it's actually Michael Pollan
Strums & Picks no shit Sherlock
@@Kizip a mushroom told me
...I thought that was a pretty good one. Cheers.
I always mix with a VU meter. Everything gain staged to -18db on the individual tracks before I mix and a VU Meter on the master buss calibrated to -14db and the records really do sound so much better when the final mix is done.
why -14 db on master buss ?
@@yassinetalbi5884 couple of reasons. 1) your mix will never come out too hot, plenty of headroom
2) less work in mastering needs to be done compressing and limiting to get up to commercial levels while still giving them lots of room to work
3) your plugins on the master buss are still being hit in their sweet spot and your not blowing out compressors and EQ's with level
@@yassinetalbi5884 I find anything less than -14 on the mixbuss results in an extremely quiet mix and any noise floor will be present in the final product (when gaining tracks to -18). And anything above -12 to -14 results in peak levels reaching close to zero and doesn't leave mastering a lot of room to work
@@OdinOfficialEmcee Oh yeah ! sorry I just thought for a second we were talking about Peak volume not db VU.. that would be around -6 dbfs right ?
@@yassinetalbi5884 A VU meter calibrated to -14db will read 0 when your RMS (Average) level is -14dbfs. Your peak level will be determined by how dynamic your mix is, instruments used, compression, etc. A real variety of factors. As a general rule of thumb though you're probably right and your peaks will fall between -10dbfs and -5dbfs, give or take a bit
What a time to be alive gentlemen, this video really gave me a good mood.
Amazing! The part where he shows off the different limiters and what they're doing "under the hood" blew my mind. Really helpful stuff, it is always appreciated!!
Forced loudness leveling basically wins the loudness wars after 3 decades of pop that has been limited so much it ended up pretty much dead sounding and with unbearable kickdrum pumping. So we finally get our dynamic range back.
Proves me right for _not_ compressing the shit out of my mixes even at the cost of them sounding quieter on a standard stereo system. If you want it louder, use the friggin' *volume* dial.
I think the idea he was trying to convey is that forced leveling is not an issue that has to relate to loudness if working with transient shaping and balancing so that the track doesnt feel so quiet.
@@3llipsis181 Do you mean that the process just has changed for the same outcome?
@@jasonsteves734 essentially yes. The dB range has been more limited, but it does not mean we are incapable of combatting that issue by realizing loudness does not necessarily translate to a higher dB. It also has to do with the frequencies boosted, the transient shaping, etc. I think it is better for us because especially things like classical music had such a dynamic range that volume levels changing so fast made it impossible to listen to certain songs without having to change the volume part way through
right! And on the radio it's still overcompressed anyway
The thing that most forget it s that compression/limiting should be done by the end user. If they are in noisy environment or need to listen to it at nights, at low levels, the limiting should be more aggressive. And when they are listening in quiet environment with more sound pressure, limiting should be non aggressive. The volume knob is actually the wrong adjustment for most needs. If you ever had the "pleasure" of mixing bespoke to one room, one system, one environment, it becomes quite obvious. You can get to perfection when it is used only in one place and the environment is static, doesn't change (for ex, art installations, theater etc). What most users need is not just gain, it is both limiter and gain.
Luckily, more and more customers do have dynamic range control, sadly they are very simple but it already makes our job so much easier. You do not have to master for headphones used in a sub/metro, the listening device has thought of that and does the final "crush" itself. What we need is a standard that all manufacturers need to follow.. Just simple none/mild/medium/strong limiting defaults, done the same way in each device. If we had that, we could easily give our mixes and masters more room the breathe and we can test the results in a standard limiter to make sure it works with all the defaults.
To anyone else who has been as obsessed as myself about finding that version of "Swing! Brother, Swing!" It's Catherine Russell. From her album Harlem on my Mind. Listen and soak up the mix and the jazz.
Am I an audio engineer? No.
Am I a music producer? Nope.
Do I even do anything related to sound? Not at all.
Did I watch the whole thing?
Oh hell yes.
Pointless
23:27 is when it all came together for me. This is an important message and I'm glad Alan shared it with all of us.
Coming from a film audio background, I have been using minimal compression in my music projects as well, out of practice and frustration with the modern limited sound, so this gives me some serious hope!
This is very true and informative video.
As a producer myself being in the industry for over 8 years now, the massive change of how " Mastering " is perceived definitely varies from artist to artist. Do not forget that every genre, every specific sound deserves a completely different touch in order to make it stand out as the producer wishes it to be in result, so even if the particular style is meant to be very powerful, it does not necessarily mean people have to make it sound loud.
I see the beauty of mastering in actual manual control, where every sounds is carefully crafted, tweaked and monitored so the result will sound as best as possible without having to simply let the automatic settings bleed it out.
So the whole auto-normalizing might be helpful only to producers who do not master at all so in the very least their tracks do not blow the listener's speakers, but other than that, it sure does damage the quality if that same technique is done to " Already Mastered " track as it becomes less pronounced.
But more of a reason to get it in the best quality so that goes away
Scranton music official- So is there any point to mastering at all if it will be affected and if the standard will automatically do it then???
@@Jz-sv1ju Normalization being applied by streaming services is not at all the same thing as mastering. Alan Silverman, in this very video, specifically states that loudness is subjective and influenced by many more variables than simply "level", and "level" is the only thing normalization is really affecting. Mastering itself creates tonal and dynamic changes in a piece of music that can be quite pleasing, and every track should be mastered, but depending on the genre some tracks will need very little in the way of dynamic reduction, and some will absolutely depend on that dynamic reduction to give it the sound that genre is known for.
The point of all of this is that if you master your music above target LUFS levels set by various streaming services, it will be turned down by those streaming services, and that past a certain point you will have reduced dynamic range without the benefit of any perceived loudness on these streaming platforms. Normalization does not replace mastering, it sets a ceiling for it and discourages the continuation of the loudness wars.
So when you're doing production tutorial?
you watched the video yet you still don't appear to know how loudness normalization works.
It does not affect the quality of your track, mastered or not.
Please release a remaster of Secret Curiosity
OH WOW. It didnt even feel like 30 mins I thought he was just getting started!!! Agreed with everyone else: we need more of him!
I feel like having caught this lecture at the 6-month point of my own personal deep dive into engineering, I am very lucky.
WOW! Never heard exactly what a limiter was doing. This is amazing
I wish this were longer - I could listen to this guy for hours. If there's one good thing to come out of streaming its that it has pulled the rug from under the loudness war.
This all works to a certain extent. All my albums were mastered at -9 LUFS. Then I learned more about this topic so I decided to master my last album at -14 LUFS. Result? On Spotify, my last album sounds quieter than the previous ones.
Now I think I've done everything right, I also involved an external mastering engineer to make sure everything was right. But if I could come back in time I would master my last album at -9 LUFS as well.
Thanks so much for sharing this story Fabio. To a degree, it depends to a degree on the genre, EQ curve and style of mixing, I think.
If the mix is already pretty compressed loud, turning it down won't make it louder on streaming services. Alternately, if there is a lot of low end in the track, that can make it sound relatively quiet compared to other tracks at the same LUFS level that are brighter. There are other factors I could go on about here.
As a mastering engineer myself, I always try to do my last pass of level setting by ear, comparing to relevant references. For me, meters are a guide and never the final decision maker.
Hope that helps!!
-Justin
Just to be clear, did you master your tracks peaking at -14dbfs or you mastered in -14LUFS. 80% of the time people get confused with this so im just making sure. Because a track mastered at -14LUFS cant really be lower in loudness if a -9dbfs master is normalized to -14LUFS. If you really did mastered in LUFS, most likely your 1st album was more dynamic in the 1st place compared to the 2nd one, A highly compressed and crushed peaking at -14dbfs will always sound quiter than a highly dynamic track mastered in -14LUFS I mean after all -14LUFS can have parts peaking at -6dbfs compared to a crushed -14dbfs whos limited to just -14dbfs. So i just want to know if its really -14LUFS or just metered in -14dbfs.
Greg Rodriguez III it was mastered at -14LUFS with peaks at -0.2dbfs (that’s where the ceiling of the limiter was set)
@@pelbriks wow tha'ts weird now I guess. Was this spotify? Was the -9LUFS master also peaks at -0.2dbfs? if its, then I could guess there is something wrong with spotify with their algorithm now, thats a bummer. cause both should be just as relatively loud at -14LUFS. cause im assuming the dynamic rage would be just the same(of course other things takes into account). Thanks for the tip, I guess people should be careful about spotify cause it has that problem. Honestly in my country I usually have a problem with this specs cause networks here are still using the old standard, so at some point using LUFS doesnt even matter to them, they just want something that peaks at -3dbfs mono!. They dont even care about EBU R128, I actualy caught one network engineer who doesnt even know EBU R128 guidelines!
Greg Rodriguez III I’ve no idea 🙂 but yes there’s something weird for sure. I dig Justin comments though, maybe more bass in the mix can lead to a different apparent loudness
It's very heartening to see this finally happening. My only concern is that the next loudness war will be fought on perceived loudness and give us harsh, mid-heavy mixes and masters. But I think that will be harder to get away with.
I hope all mixers who work with classical music listen to this.
I have started turning my masters down. I mostly only record my own music with the odd project for other people and I use Ozone 9 for mastering. I have consciously started mastering to a lower level to retain dynamics in the music, especially on a ballad. To my ears, it sounds better and if you want it louder, you can just turn it up :)
This was so quick but learned so much! I could watch another hour of him talking more in depth
This was by the far the most interesting thing I've watched all week, maybe in the last month. I don't even mix or produce. I just love music. Damn shame to cut it short.
A must watch for engineers and musicians. stick with it through the history, get to around 17:50, and keep going.
What the point? It's rather a history lesson. Nothing helpful for actual mixing or mastering.
THE FUTURE OF MASTERING proceed to talk about history and present.
When things were starting to get to the actual subject of the video it ended. Please give this guy more time!
"Ten million dollars a day from streaming"
And the actual artists see a pittance from that.
Don’t Worry Jay Z, Beyoncé, Kanye West, Tailor Swift make enough money already
The music industry is the devil's office.
They should either start their own label or negotiate their contracts better. I don't pity adults having everything done for them and then complaining they're not multi-millionares.
and that's too much in their eyes. They will keep raising the price just to upload to streaming platforms. At one point, it will be art for those that can afford it, making no money from it but paying to have their art heard and gain prestige merely from the popularity.
Alan is truly one of just a few in the world who are absolutely the best. If only his kind of genuine expertise was the norm in in the world we’d be so much better off. Bravo Alan!
this was great and overall informative, but I hope he realizes that on Spotify users can disable the normalization, which basically makes it where the difference in LUFS from track to track will vary accordingly. he was kinda speaking on it as if this is a global permanent change, but it’s definitely not. still useful to be aware of how it works when it’s enabled... I believe it defaults to being enabled and most ppl probably don’t change the settings but it’s definitely possible to disable
True! Most people will never touch or even look for that button. But it is there, hiding in the shadows.
Anyone willing to look for it, is probably the kind of person who is comfortable using their volume knob though :)
-Justin
99.9% of consumers definitely have no clue that it even exists, let alone what that even means when they see it. But yeah, just looked and it’s there! Shhhh! Don’t tell anyone!!! 😉
I think it still turns down the loud stuff though. Just doesn’t bring up quieter songs. I’ll have to try it again and measure.
Damien Lewis I did the experiment before my original reply lol. It does not turn down the louder songs. It leaves all songs in their original form.
I tried it and changed it back because I hear a lot of different Genres and it was jarring that every second Song was quieter than the one before
15:40 THIS is what we had since after World War 1. It was called the VU Meter. Zero was about two-thirds up the scale, with a peak positive value anywhere from 6 to 18dB VU, and negative values from -12 to -infinity.
It was when digital recording began to accompany and ultimately, replace analog recording, and Peak metering replaced RMS, that any semblance of standardized levels went out the window.
Hopefully, loudness metering(LUFS, or RMS VU meter emulation), will restore that relative normalcy
Need more from him. Please make another video with him but even longer and showing how we can master
The ads that interrupted this TH-cam video were so much louder!
Love how we get to hear exactly what the limiter is doing
i think that was clipping not limiting
Gregaras The Greg bahahahaha. What do you think limiters are? You’re squashing a signal which guess what? Creates distortion... not clipping. Distortion does come from clipping but it also comes from saturation, compression, using lots of drive (which is increasing the signal beyond 0db then reduces the loudness so that it doesnt blow your ears and speakers) do you even mix?
@@Creator_Veeto_PAEACP sorry, didn't delete my dumb comment
Loudness is EXACTLY why I first got into mixing as a curious nerdy hobby. After this video I feel like I truly understand what drove me into this.
Wow there are some serious thoughts in this.
I'm rethinking loudness and looking forward for great, natural sounding music!
Jaw dropping, thanks!
This guy explains stuff so beautifully that I understood this without knowing anything about music.
That chart showing cinema dynamic range explains a lot. There are so many movies, that when watched at home are terrible. The actors mumble at such a low volume (in that fake Hollywood gruff voice), what they're saying is inaudible. I always rant in my head 'what kind of sound engineer moron has done this'. It ruins the movie in a home environment. I think I need a compressor on the TV output.
dope
It's because the sound job is rushed and under budget. Good, fast and cheap pick two.
I was thinking the excact same thing, good point brother.
And then comes sound effects, car chases, music, etc and your ears are bleeding. But barely hear what is been said. Same goes to most TV programs/shows and also so many youtube videos. Music is many times way too loud compared to conversations. Have to keep adjusting volume.
This is a common problem. The "night mode" on a lot of digital receivers is a compressor. My DVD player also has a dynamic range compression feature tucked away in a menu.
Brilliant. Well worth sitting down, and putting your headphones on, to listen and learn. Absolutley invaluable.
Oh my days, this leaves me with more questions than answers. Amazing talk, very very insightful...
That was just amazing!! So clear and easy to understand. Loved it!! 😁
22:20 "the limiters little secret" - (mildly) mind blowing demonstration that summarizes the negative effects of current mastering practices.
I really don't think so, the difference is distortion yes, but because the source envelopes are a different shape to the target envelopes you never hear this type of distortion, because you are never comparing the envelopes. So to say you can hear the distortion is pretty much like subtracting a song played on the piano, from one played on the guitar and listening to the only the harmonic differences (Which could be interesting.. now I think about it 🤔).
@@stuartwilson4960 hmm im not sure if your right. He literally states that we are hearing what the limiter is clipping out while still blending in the clipped information into the original track. Am I mistaken? I can time stamp his statement if you wish.. and I, unfortunately, believe him more than you lol
@@Hugoknots If you listen to the difference of two digital signal processing functions then you would literally hear distortion, but how is that a surprise? It doesn't mean there is distortion there, it means the difference sounds horrible.
Just like you applied two limiters with different envelopes to a pure sine wave with some resulting phase offset, the difference between the two will sound impure and horrible, however listening to the results of either independently will sound perfectly pure and non distorted.
Also, some of these limiters are multi-band, so you should and would expect harmonic differences between the limited and non limited versions. Which is also not distortion, it is just purely a different envelope shape than the original.
The distortion is definitely in the mix. When the mix still sounds clean it just means the distortion volume is so low, that you can't here it conciously
@@nathanconholio1863 I don't think you can understand, if you had two versions of a just a pure sine wave, at different levels and different temporal envelopes, both independently would have REAL phase distortion maybe (0.000%) however, differentiating them will produce something that sounds like distortion, because you hearing the overtones on a set of envelopes. It is not distortion, just a different level of envelopes and harmonics.
Defining limiting as a problem is the best thing I've heard during these brilliant 30 minutes. Thank you!
I honestly feel like the entire general notion of what this all entails can coincide with some type of futuristic philosophical evolution...
I was thinking the same thing straight up
this should have as many views as there are musicians in the world. Spread this stuff !!
17:35 throwing shade at Rick Rubin ahahaha!!!
I am happy that music I have mixed and mastered since 2008 has honored Dynamics/maintained transients. My stuff doesn't sound dated because of good practices.
*Rick Rubin is an idiot. I dont know a single engineer whos worked with him who has anything positive to say about him*
@@Ascoundrel agreed. Rick may be past his prime and make very questionable engineering decisions but to say that he is a bad producer would be incorrect. His portfolio speaks for itself.
This has been great. New challenges for musicians where one simply can’t afford not to up technical understanding of sound and sound editing’s latest developments.
I was here in July with my friend from HK that's dope, I actually forgot it was filmed!
Fantastic, thanks! Ahhh, bringing back all that DR and headroom....sweet!
Developing the mixing and mastering ear takes time and practice.
This is hands down one of the best explanations I've ever seen. Every mix and / or mastering engineer alive should watch this.
This is golden knowledge and advice!!! Thank you for taking the time to make it available for all of us!! 🤙🏼
Never heard anyone so accurate about audio mastering.
Every genre has its own signature dynamic range. Ironically, only when rushed at the very end does Alan mention this, so this important concept isn't given the weight it should be. Too many people hear presentations like this and walk away thinking "I should mix and master to -14 LUFS or -12 LUFS, to take full advantage of the dynamic range at that level", when that isn't really the case at all. For jazz and other quiet genres? Sure! For EDM or Pop? Nope. Your song will sound thin and tepid and weaksauce if you do that.
@ListenAndLearned Curses! They're onto me! ^.^
exactly, it was only at the end when he was finally about to talk about what the next step should be to help solve this then he was cut off and it ended smh
Excellent video. Concise explanations. This change in allowable loudness was one of the primary reasons I created the service called: Spectrelayer. The industry was changing and the time had arrived that allowed advanced tools (such as AI) available to optimize, & repair material damaged by the "Loudness War". And it is STREAMING where the schema I created really shines. For a fraction of the cost of remastering - Spectrelayer uses advanced heuristics & "machine hearing" under the guide of a competent engineer to act as a POST-mastering stage to rebuild the audio released by the studio into a more dynamic, less distorted & more tonally balanced end waveform. The result is a track that is much less likely to be reduced in loudness by the streaming service and as a result will play with better overall sound quality & superior dynamic range - than the studio release would have without Spectrelayer. Visit & hear for yourself! Spectrelayer Rocks
This was an incredible presentation - thanks!!
I could listen to his advice for hours. Beautifully explained. That's what years of experience is :)
thank you very much for this, as a mastering guy myself, I have the problem that I mostly master DJ music, and DJ's want it loud, but streaming don't need this. so, in the end, we will need to make 2 different masters.
Robert Babicz Stoaked to see you around, you just mastered my album few days ago and I can’t tell you how happy I am to had this done by a true professional 😁 🎶 It sounds super!!!
@@alejandromagana1554 thank you
Nicely stated. Very complex issue
Can’t believe this is free content, thank you very much
I recently did a comparison of Spotify and CDs in my car using the exact same songs on each medium. Wow. CDs sound so much better.
This was amazing and WELL presented! I could sit at his feet for hours! Thank you for this wholesome and helpful content as always!
one of the best videos of all time
It’s funny that we add vinyl & tape noise to our digital music now with no reason but vintage effect.
Did you know blinkers in cars are the same? The sound used to be mechanical. Now it's just there cuz that's what we expect and it also prevents us from leaving the blinker on.
Absolutely such a great video about the history of Loudness till now. "They have taken away Level control but given you more headroom". This went off like a bright light bulb in my head. Fantastic.
Incredible explanation! Thanks for providing.
And here I am a year later. This one is a classic!
WOW thank you. If I didn't watch this I'd just continue on blasting my tracks and losing all type of dynamics
This presentation is a look at what seems to be a revolution
The loud mastering is only useful for smartphones that limit volume to avoid lawsuits by parents of deaf teenagers.
For an audiophile, ANY limiting at all is a departure from reality.
I got to hear a 2" RCA master tape on my system and I was stunned at how REAL it sounded. Playing these recordings on a hi dynamic range system is an experience like hearing the band in your room.
Now today's music is so compressed that it is fatiguing to listen to. White space is so important in music. Without it, it loses the element of surprise which keeps it interesting.
Limiting can make sense when producing/playing live elecronic music, I killed a couple of PAs myself because the owners didn´t expect they can be melted by synthesisers.
@@7k7yub7 That's a problem of an inadequate PA, not your problem.
Recordings with compression are fatiguing to listen to. Having heard a master tape recently, I was astonished at how much more "live" and fatiguing it sounded compared to the release on CD.
@@basspig obviously you are right, I just wanted to make an argument that RIGHT sound engineering is important and that in the wild there is more then black and white.
Absolutely, my friend! My ears find hard limiting and heavy compressing very offensive, especially when used on entire mixes. I just love sound of instruments with natural texture and space. For the typical material I work on I find that around -16 LUFS is the "liberation point" were the peaks are just naturally sitting below full scale (particularly if I've done good mic placement and instrument input), with maybe just a couple stray single moment peaks. It feels liberating to stop worrying about head room and just focus on the aesthetic of the mix artistically, and not have people asking why the mixes are so "quiet" when I chose to ignore the loudness war. Sad that it took so long to circle back to the gloriously large dynamic range we should have been able to keep from almost 40 years ago!
@@musicman3569 People used to say that about my Bridgeport Symphony Orchestra recordings.. absolutely ZERO compression, for a totally "you are there" experience.
In pop music, a little compression on individual instruments like kick drum or bass guitar can be used to fatten certain sounds, but overall compression is the devil.
This is very important to know. This explains why some videos are getting quiter after uploading it on TH-cam. I also noticed that the sound on youtube seems sometimes much quiter on mobile phones than if you are watching on PC. I'm not really sure if it's because of the smartphone or probably the loudness is normalized much more down for devices where you are usually using headphones.
''RICK RUBIN CAN NOT CALL SPOTIFY AND SAY ''HEY DUDES I DON'T WANT -14 TO BE MY TRACK LEVEL I WANT IT TO BE -6'' 👈 😁 MADE MY DAY
This was really good and Alan is a master of his craft. To me it begs a question - how many of todays mastering engineers not in their 50s even know the right way to do it? I bet most pop producers will still smash the shit out of their music - and let Spotify just reduce it, vs going through this whole reverse learning curve to arrive at a master that is delivered at the place it should have been all along.
Anyone remember the volume knob? Ahhhh nostalgia.
What a great presentation. I’ve learnt a lot about defining loudness and its cost over the last two years, but the breakdown of the three limiters inducing distortion was particularly eye opening.
I was waiting for Rick Rubin's name to be mentioned 🤣🤣🤣
Rick "the butcher" Rubin?? The man that made Santanas masterpiece unlistenable ..
@@kopczas Oh, and don't forget Death Magnetic and 13, among many others. Corey Taylor even stated that he never wanted to work with Rubin ever again after Slipknot recorded Vol.3: The Subliminal Verses.
Worked for RHCP, Johnny Cash etc
@@kopczas Which album are you referring to?
@@oneword7143 Africa Speaks.
Brilliant video with invaluable information! 💯❗️👍🏼
Great video and info. There is some impracticabilities to this though. People still buy and download tracks, so the LUFS of those tracks have to compete with their playlist of other downloaded tracks. Therefore you have to still master for loudness. In an ideal world, streaming services and music distributors could allow you to upload 2 versions of the track. One for streaming and one for download, but that's not the case. Also the sound of the limiter and squashed dynamic range is a factor that gives music a 'modern sound'. It's what people have become used to.
Great notes Graham. I agree! I tried to speak a little bit to that last point of yours in my intro.
For my own mastering, I always do my final level setting by ear against great sounding references, rather than against the meters alone.
But what Alan's remarks and examples do for me is make it so that I don't really ever strive to be "the loudest" of all the references. Rather, I try to find a spot that sits well across platforms, and across all relevant comparisons.
That's a good way to go in my opinion. There are so many factors that effect loudness, including timbre, which meters don't always do an ideal job of compensating for. But they are a helpful aid in getting things roughed in to start.
There is the possibility as well of having different files for different platforms and mediums, but this is probably out of scope for most smaller artists and likely isn't worth the trouble in many contexts.
Hope that helps!
-Justin
@@SonicScoop Thanks for that Justin and for your excellent effort with your channel. I wasn't aware you could have different files for different platforms. All my clients always tell me they don't have that option (Tunecore etc). But like you said, it's probably not applicable at the lower end of the market, which is where they are. I must confess, I never use VU meters (I'm not sure if I should?). I always ensure there's no clipping along the route, then when I come to master, I drag in 3 or 4 reference tracks and A/B the various frequency bands so I know I'm in the ball park and use an LUFS meter, to adjust. The problem with mastering your own mix (as you've pointed out in previous videos), is that you're still in the same room, with the same room anomalies. Using reference tracks helps to somewhat negate them.
I love how he credited everyone 👌earned a sub
The clue is in the name: Compressors. If you squeeze everything else you get a squashed consistency
Thank you, this is a wonderful presentation of a real problem and the solution. I personally like to go for a -12LUFS when mixing and in mastering not going much louder. And it works perfectly. Still there is some resistance still from engineer that spent a better part of their lives learning how to smash the songs and now they are "lost in headroom".
This was a real revolution and I love it!
22:33 How it was shown in the presentation was a bit missleading, as he suggested, that the "cghr, cghr" distortion is audible in the mixed signal. It's not true actually, since it's just a Fourier transform that subtracts the peak amplitude from the signal. You hear the sum of both frequencies, so there is no "cghrr" emmited by the speakers nor it is present in the waveform. In order to hear this distortion you would need to have it shifted out of phase or reversed in polarity.
Of course limitting can sound artificial and too heavy limitting is terrible, but it's not due to added, audible distortion as it was presented in the talk. You could argue, that any amplitude lowering "creates distortion", since you need to subtract the signal that is already there. I.e. - when you calmly mute a guitar string with your palm, you actually add a lot of terrible frequencies on top of the string vibration, which in sum just results in a silenced sound. If you somehow played back this "muting" action in reversed polarity it would sound like a terrible noise, not at all as a gentle string muting.
This is not true there is no Fourier Transform involved, just phase cancellation of one signal with a like signal that is inverted aka null/nill test as he said. The cghr, cghr distortion is probably clipping, as most of these "limiters" are really clippers. well a clipper after a limiter stage anyway. He didn't show the envelope settings for any of them so the more distorted examples where probably limiting less and clipping more (longer attack?). The distortion would be audible as well in the original signal it wouldn't be in the resulting null/nill test otherwise.
@@nathanisherwood5695 As far as I understand these algorithms they use Fourier transform to find the base frequencies which after a superposition cancel the complex waveform with an inverted signal. 1st a Fourier transform to generate the base frequencies, then the subtraction of the signal (or addition with different polarity depending on how you'd like to phrase it). My formulation in the comment wasn't correct indeed, as I suggested, that the Fourier transform does the subtracting.
@@bzqp2 no there is no fourier transform... why use a fourier transform frequencies have nothing to do with this process... a linear phase eq or a multiband compressor that has linear phase crossovers yes a kind of FFT is used here (to correct phase) but this example is only single band limiting. its not frequency dependent it applies the same to all frequencies.. its just a gain adjustment and nothing more :)
@@nathanisherwood5695 The whole point that was made in the presentation is that gain adjustments produces actual waveform with reversed polarity with regards to the original signal. Gain adjustment has to be exactly on point with the frequency if you want to only affect the amplitude of the signal. To paraphrase - if you want to lower the volume of something, you need to produce such superposition of sine waves which exactly mirrors the original waveform.
@@bzqp2 no that wasn't the point at all... it does not produce an inverted waveform (reversed polarity)... he did that deliberately after the processing just to do the null test. and no i don't know why you think you need to generate a negative sine wave to lower the volume of a sine wave no you just apply a scaling multiplier to all sample values... ie.. if you want to half the volume of a signal (sine, program or otherwise) you just multiply every sample value by 0.5 (or divide by 2).
This wasn't good... it was EXCELLENT! Very insightful... Thank you! 🙌🏻
You're right about everything but every time i listen to those argument, i want to say that there is some music who need this compression and this limiting even with loudness normalization. Those style of music are played on the dancefloor and need to be powerfull and loud as possible, just listen the youtube file of a lot of edm, dubstep or hardmusic, they are around -5 lufs integrated and even more for certain songs. Yes the loudness war is over but some style of music will need this loudness as far as they live and i've never hear that on any topic or video. I thinks it's usefull to precise those things for producer who produce those style and listen to great engineers like this guy.
Amazing video. I've been doing this since 2002. Years have nothing to do with experience though. I learned a LOT from his presentation. For the past 10 years I have worked in radio. I have my own radio show on AM radio (I'm a unicorn, I know). I have called every bit of my music mixing and mastering knowledge into how I prepare shows for air...and it's nice to know I was on-track all along without fully knowing it.
For example, where other hosts obliterate their audio to 0db, I purposefully restrain myself on compression and limiting. My "mastered" show peaks at -1.0db while I keep 12db average minimum of dynamic range throughout. Why?
Because in radio, the source audio will run through a gauntlet of compressors and limiters before it ever reaches the antenna. I purposefully leave headroom to compensate. So by the time it gets broadcasted, it's not "overly" compressed and limited by the time it reaches listeners' radios.
My show is mastered for radio. I feel like the very nature of mastering for radio has actually kept me ahead of the streaming curve without realizing it.
My podcast uses the same audio normalized to -1.0db with around 12db of dynamic range. Turns out, my rule-of-thumb 12db of dynamic range is pretty darn close to ideal for streaming for largely similar reasons attributed to radio broadcast. The streaming algorithms "average" everything out between sources just like radio has been doing for a century.
What's old is new again.
I just think it's all so fascinating.