Great review Matty! I agree, AI can be a useful tool but it's important to never value what it's telling us over our own ears and musical preferences. I'm so glad you pointed out how a bunch of major label mixes still have peaks. When I went though the process of getting certified for "Mastered for iTunes" it was hilarious to me that the standards they were saying I had to go by were broken by every major release on their platform with that badge 🤣 Happy New Year man, keep up the great content!
omg they have peaks. It's a pity that new commercial songs have no peaks and the gain is at 98-99.8 decibels with hard bass, closed sound and distortion like Taylor Swift or Billie Eilish, I can barely listen also their high frequencies. Also in Greece, many songs have flat waveforms and gain at 99decibels, they also cut many higher frequencies. Apart from specific artists that make songs in normal waveforms and gain about 96.
Thank God for you Matty, so many UA-camrs are still deluded by the -14 LUFS myth for Spotify. I worked for a significant record label that has had billions of plays, and all their tracks, especially in the last 7 years, I have been uploaded everywhere at around -8 LUFS. And they do lots of different genres. Why do people like Wyste and others keep pushing false information. Why don't they use Izotope's new plugin and they will see loads of tracks are at -9 or -8. No-one goes near -14. What is wrong with them?
It wont sound *worse if you master above the recommendations But it may affect the dynamics. Like if you uploaded a track theoretically at -7 LUFS and -1db true peak then you may get normalized to -14 LUFS and -8db true peak. Giving you 6db of average dynamic range When you may prefer to have access to the full 12LUFS of dynamic range. Its not going to be obviously apparent, in most situations, especially on headphones and smaller montiroing situations... but it may affect the force that impacts your body on larger sound systems such as in a car or something. Where dynamics are felt through the body and more important to the experience. All the tracks i've sent to streaming recently are like -7 or -8 recently and i think they sound fine. But i do acknowledge that it restricts how much dynamic range i can have within the platform, i just dont think its so important for the music i make. Of course this os overly simplified, as the integrated LUFS dont tell you the exact dynamics but rather averages.
This is all great-these reference plugins are super useful. However, I rarely see videos explaining how to actually fix the issues detected with the plugin and reference track in a practical way. If you’ve already made a video like this, or if anyone knows of one, please share the link in a reply! I’d love to see a tutorial focused on how to use these tools to improve our mix and bring it closer to the sound of the chosen reference mix, from a practical, problem-solving perspective. A hands-on guide on addressing these issues would be incredibly valuable!
Correct me if i am mistaken but comparing a release with a stream service track are 2 seperate things. Like comparing a mastering for cd with a mastering for vinyl. Indeed you can compress your mix in your taste and as loud as you like, but the -14lufs is not just a myth. The only thing that -14 is pointing out is that you can achieve a certain loudness with less limiting/compression and this would be streamed almost the same. Surely an overcompressed song would sound a bit louder but will have less dynamic range and shaped transients.
So all the streaming services have a normalization control that you can turn on and off. ( just wrote an article about it here mixandmastermysong.com/how-to-disable-normalize-on-apple-music-and-spotify/) If you turn it off you can then hear the level that the song was actually mastered. In my opinion if you send a master in that is -14 LUFS you are gonna take a massive L as far as loudness goes. Cause they only seem to turn things down that are too loud. Also if you measure a few songs with the loudness off you will see that 99% of songs are still mastered around -7 to -9 LUFS. Sad thing is it's not for the quality of the song. It's more because of people trying to keep their job. See the Producers want the song loud when they are trying to get it on the album. So it starts off loud. Then the mixer can't send the mix back lower cause when the client compares they will be like it doesn't sound better if it's lower. So then you have a loud ass mix going to the mastering engineer. And guess what. lol. He gotta make it a DB louder to make sure his sounds better too. Now this isn't allways the case. But it does happen way more than not
@@mixedbymatty Surely most of the engineers master at -9 lufs more or less depending on the genre and track and the reason for that is to be competitve. But loudness war nowdays affects production in a way of structuring a track to be loud than being creative or authentic. In any case if you are aiming for a stream release I think is a good approach to follow the streaming specs, in case you can get the track louder without smashing the mix this would be great. Out of curiosity though, how do you measure a streaming track?
Thanks, Matty. I just got both of these plugins. This video helped me understand what AI thinks is a proper mix and what the real world considers acceptable. 👍🏼
True peaks are important when converting to any codec used by streaming platforms, the process can overshoot and create distorsion by being over 0 DbFS , hence the -1dbtp "rule" , it's just a safety net...
@@mixedbymatty We don't have any control on the process when we deliver our masters to spotify, youtube , soundcloud,etc... They can apply absolutely anything they want to conform their "rules" , so better safe than sorry ;) When you hear distorsion on the platforms, it's too late.
This is so true and finally somebody explained it in the right way. Almost all major records are still heavily loud and a lot of them are (true) peaking even up to +4dB. -14LUFS for Spotify means nothing until it is not obviously distorting. I can agree that -14LUFS sounds a bit weak, even on Spotify when comparing with other songs. Well said here.
but if you upload a song louder than -14 intLufs you have to set max Tp not over -1dbTp, right ? so how to use truebalance and true level using reference instead of presets ?
If you really want to have more doubts about this flashy plug-in. Take a song any song, in my case I used a Rick Braun jazz song and run that as your source and as your reference. Since they are identical it should think everything is perfect, but inexplicably it finds things that are off. That tells me this plugin is pretty stupid, kind of like chat GPT how it lies. So much for AI
Agree. Remember AI is a "machine", though very complex. The machine is programmed by people. So, what you get in an AI, is firstly the opinion of those people (which may not be same as yours), secondly the soul-less-ness (is it a word?!) of a machine. I think if you can rely on your ears and like your judgments, then no need to be sold to a machine, but for the beginners who are less seasoned, the machine can give them some baseline to tweak on.
streaming audio is a pain, I ignore its demand and release from a balance using my ears. These plugins are helpful for balance, but shouldn't be considered to make spotify happy.
@@mixedbymatty This is very interesting. I'm squarely in the learning camp. Logically if a little part that is clipping is subtle, or just the tip of a loud waveform or area, it'll probably be indiscernible. I've never heard anyone suggest this before. Common sense, really though, when you think about it. But wouldn't a brick wall limiter or clipper accomplish the same thing? Thanks for your efforts here!
@@mixedbymatty Quite a few engineers will use a clipper in conjunction to a peak limiter to get the volume up. However, I will say that it's all about taste. The mastering process can squash the crap out of the mix (Pretty common in todays music; pop, metal, hip hop, etc) so use this tool as a way to get more information about your music but always use your ears as mixing and mastering is objective to the engineer.
8:15 Just mad as if 0.1 LUFS make a big difference. However some pros clip on purpose, for example in ableton you can clip to get a certain sound without worries :) ua-cam.com/video/Ggz8CqzwNmw/v-deo.html
08:50 i think it is made to make masters and not run other people's masters to fiddle with how the analyzer works. yes you have the point here but read the thing again:
You dont suggest to push the true peak, but this whole video was saying to go over the true peak? Then theyre great plugins, yet they arent telling you the truth... This is why i watch White Sea Studios channel..
I'm no expert but the reason I believe that you are seeing true peak in excess of 0 dbfs is due to the fact that the sample rate of the audio vs the sample rate of the DAW are different. If you run a 48Khz track on a daw that is playing back at 44.1Khz, the tru peak will exceed 0dbfs. However, if you run the same track on a DAW with 48Khz playback, it will not exceed 0 dbfs.
if we wanna be technical then mastering the mix technically did it before any of those two but brand loyalty sucks and having competition is not a bad thing for the consumer
_"Ozone. Rip off"_ Nah . . . Mastering the Mix made Reference, the Ozone copied Reference with Tonal balance control, and then Sonible copied them both with true:balance.
Great review Matty! I agree, AI can be a useful tool but it's important to never value what it's telling us over our own ears and musical preferences. I'm so glad you pointed out how a bunch of major label mixes still have peaks. When I went though the process of getting certified for "Mastered for iTunes" it was hilarious to me that the standards they were saying I had to go by were broken by every major release on their platform with that badge 🤣 Happy New Year man, keep up the great content!
Haha. Yes, I don't understand people being worried about staying at -14 LUFS. lol. Happy New Year to you too man!
omg they have peaks. It's a pity that new commercial songs have no peaks and the gain is at 98-99.8 decibels with hard bass, closed sound and distortion like Taylor Swift or Billie Eilish, I can barely listen also their high frequencies. Also in Greece, many songs have flat waveforms and gain at 99decibels, they also cut many higher frequencies. Apart from specific artists that make songs in normal waveforms and gain about 96.
But if mixes are loud doesn’t it get turned down in streaming services
Changing the level doesn’t change the spectrum. Way better than ozone IMO
Thank God for you Matty, so many UA-camrs are still deluded by the -14 LUFS myth for Spotify. I worked for a significant record label that has had billions of plays, and all their tracks, especially in the last 7 years, I have been uploaded everywhere at around -8 LUFS. And they do lots of different genres. Why do people like Wyste and others keep pushing false information. Why don't they use Izotope's new plugin and they will see loads of tracks are at -9 or -8. No-one goes near -14. What is wrong with them?
Wyste?
The idea of Spotify specific mastering is absolute bs. I always upload a -8ish and never find issues with distortion or anything
@@jasoncruizer Exactly. So many UA-camr audio gurus are deluded. Even Dan Worrall.
@@alexbreugelmans8448 Wytse Gerichhausen @ White Sea Studio
It wont sound *worse if you master above the recommendations
But it may affect the dynamics.
Like if you uploaded a track theoretically at -7 LUFS and -1db true peak then you may get normalized to -14 LUFS and -8db true peak. Giving you 6db of average dynamic range When you may prefer to have access to the full 12LUFS of dynamic range.
Its not going to be obviously apparent, in most situations, especially on headphones and smaller montiroing situations... but it may affect the force that impacts your body on larger sound systems such as in a car or something. Where dynamics are felt through the body and more important to the experience.
All the tracks i've sent to streaming recently are like -7 or -8 recently and i think they sound fine.
But i do acknowledge that it restricts how much dynamic range i can have within the platform, i just dont think its so important for the music i make.
Of course this os overly simplified, as the integrated LUFS dont tell you the exact dynamics but rather averages.
This is all great-these reference plugins are super useful. However, I rarely see videos explaining how to actually fix the issues detected with the plugin and reference track in a practical way. If you’ve already made a video like this, or if anyone knows of one, please share the link in a reply! I’d love to see a tutorial focused on how to use these tools to improve our mix and bring it closer to the sound of the chosen reference mix, from a practical, problem-solving perspective. A hands-on guide on addressing these issues would be incredibly valuable!
Correct me if i am mistaken but comparing a release with a stream service track are 2 seperate things. Like comparing a mastering for cd with a mastering for vinyl. Indeed you can compress your mix in your taste and as loud as you like, but the -14lufs is not just a myth. The only thing that -14 is pointing out is that you can achieve a certain loudness with less limiting/compression and this would be streamed almost the same. Surely an overcompressed song would sound a bit louder but will have less dynamic range and shaped transients.
So all the streaming services have a normalization control that you can turn on and off. ( just wrote an article about it here mixandmastermysong.com/how-to-disable-normalize-on-apple-music-and-spotify/)
If you turn it off you can then hear the level that the song was actually mastered. In my opinion if you send a master in that is -14 LUFS you are gonna take a massive L as far as loudness goes. Cause they only seem to turn things down that are too loud. Also if you measure a few songs with the loudness off you will see that 99% of songs are still mastered around -7 to -9 LUFS. Sad thing is it's not for the quality of the song. It's more because of people trying to keep their job. See the Producers want the song loud when they are trying to get it on the album. So it starts off loud. Then the mixer can't send the mix back lower cause when the client compares they will be like it doesn't sound better if it's lower. So then you have a loud ass mix going to the mastering engineer. And guess what. lol. He gotta make it a DB louder to make sure his sounds better too. Now this isn't allways the case. But it does happen way more than not
@@mixedbymatty Surely most of the engineers master at -9 lufs more or less depending on the genre and track and the reason for that is to be competitve. But loudness war nowdays affects production in a way of structuring a track to be loud than being creative or authentic. In any case if you are aiming for a stream release I think is a good approach to follow the streaming specs, in case you can get the track louder without smashing the mix this would be great. Out of curiosity though, how do you measure a streaming track?
@@lucaslaboudes2919 have to worry about losing dynamics as well
Great video & tips, thanks Matty !
Thanks, Matty. I just got both of these plugins. This video helped me understand what AI thinks is a proper mix and what the real world considers acceptable. 👍🏼
Thanks for confirming this. I've also analyzed tons of tracks and can confirm exactly what you said.
Audio lens?
True peaks are important when converting to any codec used by streaming platforms, the process can overshoot and create distorsion by being over 0 DbFS , hence the -1dbtp "rule" , it's just a safety net...
Only if you can hear it.
@@mixedbymatty We don't have any control on the process when we deliver our masters to spotify, youtube , soundcloud,etc... They can apply absolutely anything they want to conform their "rules" , so better safe than sorry ;) When you hear distorsion on the platforms, it's too late.
Just set the threshold to -1 people
You can't overshoot 0dB.
@@davelordy they referring to Apple music limiting at -1. Also, you can overshoot 0db when it converts back to analog.
Why is the car full?
Without seeing the spectrum, the mix sounds a bit bright and scooped actually
This is so true and finally somebody explained it in the right way. Almost all major records are still heavily loud and a lot of them are (true) peaking even up to +4dB. -14LUFS for Spotify means nothing until it is not obviously distorting. I can agree that -14LUFS sounds a bit weak, even on Spotify when comparing with other songs. Well said here.
+4dB sounds like alot. Could you provide the name of a major release that has +4 dB true peak? I definitely want to get a copy of it.
@@alex_c_d f.e. Chainsmokers Who Do You Love -7 LUFS, +3.4dBTP. Travis Scott Sicko Mode - 7.9 LUFS, + 2.4 dBTP
but if you upload a song louder than -14 intLufs you have to set max Tp not over -1dbTp, right ? so how to use truebalance and true level using reference instead of presets ?
I tried this with True Balance, I took the song I was working and made it the reference... And it was off.
All about personal taste. Nailed it!
And reference tracks
Matty you're the best
If you really want to have more doubts about this flashy plug-in. Take a song any song, in my case I used a Rick Braun jazz song and run that as your source and as your reference. Since they are identical it should think everything is perfect, but inexplicably it finds things that are off. That tells me this plugin is pretty stupid, kind of like chat GPT how it lies. So much for AI
Agree. Remember AI is a "machine", though very complex. The machine is programmed by people. So, what you get in an AI, is firstly the opinion of those people (which may not be same as yours), secondly the soul-less-ness (is it a word?!) of a machine. I think if you can rely on your ears and like your judgments, then no need to be sold to a machine, but for the beginners who are less seasoned, the machine can give them some baseline to tweak on.
thanks matt
great review matt👍🔥
streaming audio is a pain, I ignore its demand and release from a balance using my ears. These plugins are helpful for balance, but shouldn't be considered to make spotify happy.
is this tool for mixing or mastering ?
Doesn't going over true peak result in clipping?
Only if you can hear it
@@mixedbymatty This is very interesting. I'm squarely in the learning camp.
Logically if a little part that is clipping is subtle, or just the tip of a loud waveform or area, it'll probably be indiscernible. I've never heard anyone suggest this before. Common sense, really though, when you think about it.
But wouldn't a brick wall limiter or clipper accomplish the same thing?
Thanks for your efforts here!
It would take a lot consecutive samples to create any kind of audible clipping.
@@davelordy Thanks Dave!
@@mixedbymatty Quite a few engineers will use a clipper in conjunction to a peak limiter to get the volume up. However, I will say that it's all about taste. The mastering process can squash the crap out of the mix (Pretty common in todays music; pop, metal, hip hop, etc) so use this tool as a way to get more information about your music but always use your ears as mixing and mastering is objective to the engineer.
8:15 Just mad as if 0.1 LUFS make a big difference. However some pros clip on purpose, for example in ableton you can clip to get a certain sound without worries :) ua-cam.com/video/Ggz8CqzwNmw/v-deo.html
Matty, to "Let's go" was not convincing..what is going on??
Great video
08:50 i think it is made to make masters and not run other people's masters to fiddle with how the analyzer works. yes you have the point here but read the thing again:
You dont suggest to push the true peak, but this whole video was saying to go over the true peak? Then theyre great plugins, yet they arent telling you the truth... This is why i watch White Sea Studios channel..
Damn. They just stole Izotope’s product name like that? Wth
Izotope, often imitated but never duplicated.
Big facts about AI plug-ings.
true level is bad..always wants the peak lowered
What dude! No just no. Never go over true peak.
Play any top 100 song, and most if not all are over the 0 dB threshold. If you don’t HEAR distortion, you’re completely fine.
@@rjsworld1512 there is difference between 0db and 0dbfs. It’s not possible to go over 0 dbfs it’s always a clipping.
I'm no expert but the reason I believe that you are seeing true peak in excess of 0 dbfs is due to the fact that the sample rate of the audio vs the sample rate of the DAW are different. If you run a 48Khz track on a daw that is playing back at 44.1Khz, the tru peak will exceed 0dbfs. However, if you run the same track on a DAW with 48Khz playback, it will not exceed 0 dbfs.
Ozone. Rip off
Works better and costs less tho
if we wanna be technical then mastering the mix technically did it before any of those two
but brand loyalty sucks and having competition is not a bad thing for the consumer
Sonible does it better
_"Ozone. Rip off"_
Nah . . . Mastering the Mix made Reference, the Ozone copied Reference with Tonal balance control, and then Sonible copied them both with true:balance.
Thanks for a nice explanation
Yep. Use helmet and ankle protection. Never use at home unsupervised!!!