Thanks for this video. After uploading some songs to spotify, which I mastered to have an average LUFS level of -14 o -13, because they normalize each track to that level anyway, it was still apparent my tracks were a little bit lower than many other commercial tracks. I looked everywhere for videos explaining why this happens and how to work around it but none were found. They just would stick to the script "everything is normalized to -14 LUFS playback so you dont need to compress more your track" without addressing the obvious fact that even with normalization many spotify songs have different perceived levels. You are the man.
Took me sooo long to get loudness right, Like over 10 years. It’s such an accumulation of things it’s hard to teach someone how to get Loud clean punchy mixes as it stems back to choices you make early on many different things. Gone are the days I used to throw an L2 limiter on the master bus and push it 😂. Now I generally get pretty clean -7 on rock and -6 on dense electronic without distortion or pumping. I never thought I could be able to do that even like a year ago and can now dial it back a bit to be safe.
Would love a metering video. Especially how and what metering to use during gain staging, before and after analog plugins, setting static mix levels, bus processing, and mastering, etc. Also, would love your thoughts on managing and optimizing crest factor for increased perceived loudness and how to measure and monitor that properly. (Sorry, that’s a lot of stuff, but would be very helpful-this stuff gets confusing).
I've been trying to mix and master my first track after previously outsourcing it and having real problems with getting it to what Spotify recommends without ruining it, so this is just what I needed to hear! I'm going to revisit it with just my ears and a reference track and try uploading it to see what happens...fingers crossed!
Thank you for saying it how it is! I agree. It changed my life when I just started compressing and more important *limiting* my masters to around 8-10 LUFS
Please do an episode on metering! One of the less sexy sides of mixing/mastering that rarely gets talked about and I am 100% here for it. Also, I certainly can't call myself a professional mastering engineer but I have mastered a number of projects for friends and I usually deliver two masters. One at -1.5db peak level for streaming and one at -.2db for CD. I didn't arrive at these peak levels through trial and error, but through a recommendation from Eric Valentine which I figured is safe. Is delivering two masters necessary? Also, do you have limiter recommendations? I have been loving the Brainworx XL V2 personally.
@@RiffMusic1970 Honestly, I am not totally educated on this so don't quote me on this. But I think it's to leave headroom for the intersample peaks that may become problematic when converting the file using different codecs. So a track with -.2db of headroom might clip when Spotify or whomever converts it to their codec. This is a problem. So leaving 1.5db of headroom gives room for digital services to do their thing without negatively affecting the music. I don't know if this is 100% how it works, but this is how I understand it. I may be mixing and mastering buzzwords, though. If you want to see Eric Valentine's video I referenced, watch the episode on his channel about mastering. I think it also leaves headroom for your client to convert the file for themselves. If they want to create mp3s of their songs they can do so without introducing artifacts from the conversion process. I think
I work on mainly rock stuff and I tend to land somewhere between -6 and -9 LUFS, even with a lowered ceiling on the limiter. I really don't think about it too much. I use my ears to what sounds best for a particular track. I make mix/mastering decisions for my music, not the streaming platforms. Future-proof your work... these platforms will undoubtably change their recommendations down the line.
Quick question for anyone out there... I've worked with 3 different mastering engineers now and all 3 mastered my tracks to right up to 0.0 db. They were all great at their job and I'm sure they knew what they were doing, but can pushing the peak to "0" cause some distortion on some playback systems? I thought -0.3 db (or -0.2 or -0.1db) was more common. Anyone have any thoughts on this or have any info? Thanks!
Thx man, clear and well to the point. Two questions (between a million others!): - if doing a mastering level of a track with "True peak" meter ON, do we still have a to go for the -1db "safety zone" for streaming services (recommended)? I know the inter-sampled peaks mostly can bother us adding distortion when transcoding to lossy formats. - If the LUFS reading is above -14.0 for a master, then Spotify wants a True Peak value under -2.0. But what if we go -0.5? Any penalty from them on this? For the LUFS i get i they normalize, but what about the "True Peak reading"? Thx for your time and your video! And yes i think you can go more nerdy! Gently...
Well, this still isn't that clear. Suppose you have a song that sounds good at -14 and also at -8 and everything in between, what are you going to choose? You get more dynamics with -14, so then that? It seems like the pros aren't doing that, and are just going for louder. Why? Is it just like you said, that the quieter sections of that louder master will sound louder on streaming services? Or does it have nothing to do with streaming services at all? It seems the rule is "as loud as you can get without losing fidelity." Why? For vinyl, or CDs, or radio?
My thoughts on loadness for mastering. Use your ears and make a master that sounds as good as possible and make it as load as you possible can. How much you should compress and if you should use a limiter at the end of the master chain really depends on genre (the more you use compression and limiting, the loader the master can be). I recently remastered a recording I made in 2001. I just used my ears and made an analysis after the mastering was done, it ended up around -9 to -10 LUFS, which somehow is the "normal" level for a rock/hardrock record. Jazz is lower, metal is higher, but the important thing is to be in se same ballpark as your competitors. The band Fata Morgana is not a metal band so I think I did it right. Feel free to listen (just click on me). If you like guitar-driven hardrock with more complex arrangements but still melodic, you should defenetely listen. Remember, your ears are going to tell you whats right. not numbers! // Peter, guitarist in Fata Morgana
I measured a few songs off of the new blink album. I think they won the loudness wars. I measured -3 to -4 lufs! Everything was square waves off and I can hear digital distortion.
Probably the best video you have ever posted about music production. I love the honesty, bravo. The streaming services need to do better loudness analysis before penalties matter. Obviously, distortion for a music audiofile is outside of all this conversation.
I mainly mix hip hop and RnB. My clients typically like a super loud mix so I tend to get between -10 and -6. I use the god particles limiters and find that it doesn’t complete squash my transients and dynamics as much as what I used to use, the L2 limiter. Gain staging helps tremendously
I've been mastering a piano based track which has a lot of aggressive key velocities mixed with some very soft touches also. If I get my loudness level anywhere within -6bd of 0, the higher octave piano notes played on the heavy side rattle my eardrums. I've done a ton of eq notching and finally decided that it doesn't really need to be at 0. That doesn't mean I don't have a lot to learn, but I'm starting to think loudness level depends on what's in the track
Yeah, loud transients on full range piano can wreak havoc on small iPhone and laptop speakers. It’s often better to master these kinds of pieces quieter because of it. That one instrument causes more trouble in this regard than any other in my experience.
This is a great video on LUFS and mastering loudness. My thing is making sure that my mix sounds balanced b4 I go to my 'My Mastering mode". I've been using a True Peak mastering VST that has an Output Dim adjustment. Maybe a mix example where you then take that mix to master it would be cool.
Thanks for the info! I'd love to dive deeper with more details and a 'nerdy' video on mastering and loudness. Could you also recommend the best plugins for this?
Problem with streaming platforms, they butcher the audio quality and let's face it, too many Dummy phones playing thru their speaker and many don't even care about quality . But KOODOS to you from a Musician who cares about Quality WAV😁👍🏻
Thank you for this great episode! 😍🤗 Very well done! Super useful information you packed in there. And I wouldn't mind if you'd go even more technical. I would like that! 😁
thanks for this. i've asked this question many times and could'nt find a real answer. all i found was -14 and im like WHAT. Then i found that EDM -7 and nothing else until this with -8 for hip hop. thank you. i was getting to -10 and it just didnt quite feel as loud so now i know.
The most comon -8 Lufs cause they will mess it up if you try to turn it down and if you just stay with-8 LUFS You will be in the industry standard so they tell me besides why would we make different levels for different platforms. Just bang it but make it clear I say.
Can you comment on recent drum and bass tracks specifically the neurofunk style currently being as loud as -3 LUFS with True Peaks of +2dB, yet not sounding too distorted? How do the mastering engineers achieve this?
i did not understand this part - when streaming services turn down ur master level, why do the quieter parts still sound louder than a generally quiet song? is it because the louder song's quiet parts are normally louder, or is it because streaming services only turn down the loudest parts?
This is the mistake I made initially. Mastering to a particular LUFS level. Reason some tracks sound louder is cos they have generally shoot for much louder on the LUFS reading. Sometimes -7 to -8 LUFS. Its not necessarily 'Louder' just more dense.
Very helpful. But I'm still getting my volume pushed down when I publish on spotify or youtube. I'm mastering at a reasonable volume, matching with other professional tracks, and even putting my true peak at -1, but I still get my song much quieter than others (when publishing). What am I doing wrong?
Justin I've got questions if you get a shot on bouncing things down after the Master. I've seen some casts on dithering and I sort of get it...but are Mastering engineers saving all these copies of single songs? Like dithered, not dithered, WAV, mp3...and do you have a system for the way you name each version. Thanks for all of the content its fantastic!!
.wav files or .flac files are going to sound the best, because they are lossless. .MP3, .ogg and .aac are going to sound less good because they are lossy formats.
@@johnb6723 my question is more so referring to playback systems and such...like aren't tracks for CDs dithered more than tracks for streaming...for example(I thought I heard that, but I'm certainly no expert hence the confusion!)...and are engineers saving them in all these different ways or how's that work
Intersample Peaks: I have found out that most of the hot mastered hit records showing inter sample peaks up to +4dB. Looks like nobody in this field taking care about any headroom below zero and nobody uses true peak limiters. What do you think about that ? Could show some examples if you like.
True peak is meaningless since once it hits the converters it cannot exceed 0. In the digital world headroom is infinite so meters reading above true peak in the daw are ignored by 99% of professional engineers
Yes! In the real world, you can have some significant occasional intersample peaks and not have an issue. However, if you have a lot of very hot intersample peaks, it can potentially cause issues when converted. Listening through potential end codecs with a plugin like the ADPTR Audio Streamliner or the ones made by Nugen Audio can let you know if this will actually be an issue. After mastering hundreds upon hundreds of projects per year for the last 8 years or so, there has been exactly one time where we heard new distortion added after conversion occurred and decided to re-upload the album. But it made it so that I always check now! Simply lowering the peak level on that record from -.4 to -.8 did it. It's true-not a lot of people use true peak limiters these days, because most limiters sound better in their normal mode to most people who use them. HOWEVER, the new bx_limiter True Peak, only has a true peak mode and sounds really good that way. Its detection circuit is a bit different than the norm, and I've found myself starting to use it for hotter material. Very nice. Hope that helps, -Justin
@@SonicScoop Thanks Justin, great reply. I just wanted to point this, because almost every TH-cam ”engineer” says: True Peak Limiting is a must and even better bis 1dB of headroom. But if you discover current modern chart music, nobody cares in the real world. I have tested the bx TruePeak Limiter as well and compared to FF Pro-L. In my cases Fabfilter got me better results. So i´m not a big fan of the bx Limiter, but i like the Sonible Smart Limiter. This one works quite well for me. Cheers
i made some tracks mastered at abbey road studio for an album and they were showing +0,4 db true peak !! i sent the engineer a message, he answered me if it sounds good to your ears don t worry about true peak and it sounds amazing so... and i watched a video too about mastering with the engineer who mastered bad guy from billie eilish and he said: never use a true peak limiter... they all sounds bad... so in the real world nobody cares about true peak because i think that if true peaks stays around 1 db or maybe 1,5 db i m not sure that we can hear any difference.
I only presently work on my own songs, none of which fit into the very small Genre List used in this video. Maybe some of mine could be squeezed into the "pop" area, but it would be nice to see what the "blurred Genre of Country" would be an acceptable level, and no one ever seems to bother to talk about worship/CCM areas. I use two ADPTR plugins, Metric A/B, which is THE BEST I have seen (you can also upload 16 reference tracks), and its Streamliner, for comparison which also covers Codec, , along with Waves WLM Plus 5.0 loudness meter. Since I have Studio One Pro 5, it has a super duper Project Page module, the ONLY DAW with a built-in mastering module, in which you can tweak a song by going back to the original track to do that for separate tracks. It would be SO NICE if all these services would come together and choose ONE STANDARD LUFS.
I guess im a DR nut....I just did a mix at -17 LUFS and I really like the space....so fun to crank up those wide DR mixes...I trust folks to use their volume knobs...probably controversial I know. ;) Nearly every album ive spent a ton of time listening to had relatively high DR's.....I like to hear things really bounce around. Thanks for the great content!
How do you sort a mix within a genre that requires a lot of low end? I'm making a doom metal personal project, and one of the things I've run into is that when the low end sounds appropriate for the genre, it's destroying my master bus compressor. With guitars tuned this low, it's basically like having 5 basses in my mix once both sides are double tracked. I'm making space where I can, but sometimes the soup just sounds better. Is the solution to just not compress the guitars at all and ride the faders?
Good length & explanations 👍 (but yeah, get nerdy! :) Are there downsides of sending to Apple/Spotify at -2 peak? Like, can’t you just give a +2 (0db) copy to clients etc & send the -2 peak to Apple? Does it sound better/worse? Is it not just normalized?
I've seen comments about pushing beyond -14 LUFS, with some going as high as -6 LUFS. But what's your experience with releasing masters on streaming platforms? Personally, I've found that going above -14 LUFS can lead to issues, especially on Spotify, where it seems to degrade the sound the most. Apple Music handles it a bit better, and TH-cam fares the best. Still, even at louder levels, the master doesn’t sound as good as it does in your DAW or on your speakers. Also, commercial tracks often sound way louder than those mastered at -14 LUFS. Some believe that streaming platforms, particularly Spotify, might have a bias against independent releases compared to commercial ones.
All of these services should simply be turning down louder masters, so there should be no additional "degradation" from any of these services. However, you may be listening to Spotify in a lower quality audio mode. You can update your audio quality settings with the instructions at the beginning of this video: th-cam.com/video/L4pf29_u1W8/w-d-xo.html Ironically, TH-cam has the lowest quality playback of those 3 services if you put all three in their highest quality settings. It maxes out at 256kbps and can also do some subtle HF rolloff above 15k. (But it's barely noticeable.) So there may be other factors at play to look into. And yes! Commercially mastered tracks can sound louder at the same LUFS rating, but part of this will be because of frequency balances. Although the scales used in LUFS try to compensate for the quirks of human hearing, this is imperfect, and brighter will sound louder, while excess low end that's found in a lot of amateur mixes with make things sound quieter at the same LUFS reading. Also, there are little tricks that can be done, such as baking in quieter sections to bring down the integrated LUFS rating of a song, allowing is louder sections to sound louder than a song that's the same LUFS reading the whole way through. I hope some of that makes sense! -Justin
Great video. Two questions: 1. This -14 to -8 LUFS is a what the LUF meter should read POST mastering, not pre, correct? 2. How does the "Loudness Penalty" play into this? I've mastered stuff at -12 yet the Loudness Penalty website says i'll get hit by up to -4? Any answers would be helpful. Thanks!
Uh-oh haha, I must have accidentally uploaded the not-finished-being-edited version! Thanks for letting me know. Out of the thousands who have heard it so far you're the first one to tell me. My goodness! That's embarrassing :-) Thanks! -Justin
Usually you’d lean on the quieter side, -13 to -12, but there are cases of modern, loud production aesthetics, even in these styles. Depends on how processed vs natural you want things to sound. The Johnny Cash American recordings for instance are QUITE loud, even when it’s just a solo guitar and voice. If you want that kind of tone and aesthetic, more limiting is part of it. Ultimately, you should study records you really love the sound of. Hope that helps! -Justin
Great video. At the end, i didn't understand: you recommend setting the limiter with the true peak of -1 for not extremely loud music (Pop and Rock: -13 to -10 LUFS)? I make rock...
With a true peak limiter, you should theoretically be able to go all the way up to just about 0 as it is catching the intersample peaks. Though may as well set it to -0.1 I guess? If you're not using a true peak limiter, I usually find that setting the output to anywhere from -0.4 to -1.0 will be enough to keep the intersample peaks from causing any meaningful issues, depending on how hot you are mastering. This does require a little bit of trial and error. A tool Like ADPTR Audio's Streamliner can help you determine this by looking at intersample peaks and hearing what the codecs will actually do to your master on streaming. Hope that helps! -Justin
Do you know if it's possible to master audio for delivery to streaming services like Curiosity Stream in Premiere or Resolve? I would love to find a good tutorial about that process. Thanks for the help.
The loudness of the master should be THAT which is RIGHT for the song,the and the genre. IF THIS MEANS WHATEVER STREAMER WILL TURN IT TDOWN, SO BE IT. The recommended True peak level is the only value that you should adhere too for the different streamers.My experience is that the higher the sample rate and bit file that you deliver the file, the better the sound quality of the song when it appears on the platform, mix and master for for what what believe is right for how the song and bedamned for what the streamers do.KEEPING THE SONG AND INTEGRITY IS THE MOST IMPORTANT THING, MIX AND MASTER FOR THE RIGHT SOUND, if the result comes out very high then ask yourself if you wanted it so loud and maybe let it breath, look at you average LUFS afterwards and make decisions. And Try FIDEF, THIS CAN GET YOUR SONGS TO SOUND BETTER AND ENERGETIC WITHOUT SO MUCH COMPRESSION OR LIMITING ( My little tip of the day…..
Hey, Justin! You did a podcast a month ago arguing against the need for super high resolution audio, and I totally agreed with your stance. However, I was thinking while watching this episode that perhaps using a sample rate of 192kHz or higher might actually help to prevent inter-sample peaks while mastering due to a more accurate representation of the wave shape. What are your thoughts on that? Has anyone done any research on that?
Hey Jarrod, that's a fair question! Using a higher sample rate wouldn't totally eliminate this issue, but maybe it would decrease the magnitude of it? I'm not sure. But it's plausible. Would have to look into it. That said, there is already an option to basically do this, regardless of what sample rate you are working in: "True Peak" limiters use oversampling in the detector circuit to help eliminate intersample peaks. Of all the ones I've tried so far, I like the sound of the bx_limiter True Peak. But other brand's limiters will often have this mode. They do tend to sound a little different in their True Peak mode. You'd still need to address the intersample peak problem with a higher sample rate, but maybe they wouldn't be quite as high? I'd really need to do more research and experimentation on it. But honestly, it's not that big of a deal either way, because just turning down the peak level slightly or using a true peak limiter) tends to fix this without issue, and would still likely need to be done at a higher sample rate anyway. Hope that helps! -Justin
Hi, Where are you getting your client references? Are you using Spotify (which has been modified) are you using the commercial CD, digital download? If Spotify, are you normalizing the file to get a proper master equivalent? I would think the reference needs to be at the actual level to have any relevance.
Yes, you used to be able to turn off volume normalization in Spotify. Not sure if you still can. But you can definitely turn it off in Apple Music and other places. That is important. It's another reason TH-cam is not great for referencing. -Justin
Not for music releases, no. The FCC has no jurisdiction there. The case is different in broadcast, where everything is played back at the same average level. Streaming services too, normalize levels on playback by turning down the loudest tracks. But the real world experience is that even after this occurs, louder masters can and often will still be perceived as louder. Hope that helps! -Justin
If your music sounds really bad at -18 LUFS then something is wrong. That's super quiet. I'd recommend checking your low end. You may be piling in a ton of low frequency sub energy that you can't hear in your room, but that's eating up all your headroom. Hope that helps! -Justin
@@SonicScoop also it's my vocals that sound bad. My beat sound great when I push it. I'm using the neuman tlm102 to record and my voice is a little higher pitched for a male. But im going to try what you recommend first then maybe I'll try digging around to see what else it can be
For some genres and styles that's perfect! For some it's too loud and for some it's too quiet :-) But if that range works well for what you're doing, keep at it! -Justin
this is the problem of outdated sound engineers, they still carry the old-fashioned Audio CD attitude. They think they are the ones who created the streaming service, If the streaming service engineer has already made -14lufs is the limit, why does the sound engineer need to break it? do you need to download the audio before you listen to it? that's why I said, if there is still an attitude of loudness war, it is better to put the song in the CD store. If you want to really compete in streaming services, try mixing and mastering that sounds loud and clear without changing the -14 lufs, if you succeed, you are a very great sound engineer in this era.
-8 to -6 on my final masters for rock anything really. 99% of the songs I ref are fucking loud. Overall Eq plays a huge part. Sometimes when I think it needs to be louder, it realy needs more top end.
It'd probably get you a good amount of views, but he seems to be pretty smugly opinionated about a lot of things so it might just be an exercise in frustration. If you do have him on, ask him to show his mixing and mastering credits. I can't seem to find any. @@SonicScoop
I usually don't comment videos but this is probably the best video I've seen on loudness. Great job!
Go nerdy, go technical cause it is not time wasting. All information is a blessing. The more we know the merrier.
amen
Thanks for this video. After uploading some songs to spotify, which I mastered to have an average LUFS level of -14 o -13, because they normalize each track to that level anyway, it was still apparent my tracks were a little bit lower than many other commercial tracks. I looked everywhere for videos explaining why this happens and how to work around it but none were found. They just would stick to the script "everything is normalized to -14 LUFS playback so you dont need to compress more your track" without addressing the obvious fact that even with normalization many spotify songs have different perceived levels. You are the man.
Took me sooo long to get loudness right, Like over 10 years. It’s such an accumulation of things it’s hard to teach someone how to get Loud clean punchy mixes as it stems back to choices you make early on many different things. Gone are the days I used to throw an L2 limiter on the master bus and push it 😂. Now I generally get pretty clean -7 on rock and -6 on dense electronic without distortion or pumping. I never thought I could be able to do that even like a year ago and can now dial it back a bit to be safe.
That's reassuring AF to hear.
Would love a metering video. Especially how and what metering to use during gain staging, before and after analog plugins, setting static mix levels, bus processing, and mastering, etc. Also, would love your thoughts on managing and optimizing crest factor for increased perceived loudness and how to measure and monitor that properly. (Sorry, that’s a lot of stuff, but would be very helpful-this stuff gets confusing).
Agreed! Justin, this would be very helpful!
I've been trying to mix and master my first track after previously outsourcing it and having real problems with getting it to what Spotify recommends without ruining it, so this is just what I needed to hear! I'm going to revisit it with just my ears and a reference track and try uploading it to see what happens...fingers crossed!
More technical and nerdy. Don’t hesitate.
And thanks a lot. Message get’s through “loud” and clear.
Thank you for saying it how it is! I agree. It changed my life when I just started compressing and more important *limiting* my masters to around 8-10 LUFS
Please do an episode on metering! One of the less sexy sides of mixing/mastering that rarely gets talked about and I am 100% here for it.
Also, I certainly can't call myself a professional mastering engineer but I have mastered a number of projects for friends and I usually deliver two masters. One at -1.5db peak level for streaming and one at -.2db for CD. I didn't arrive at these peak levels through trial and error, but through a recommendation from Eric Valentine which I figured is safe. Is delivering two masters necessary? Also, do you have limiter recommendations? I have been loving the Brainworx XL V2 personally.
Interesting. What would be the disadvantage to delivering -.2 to a streaming service?
@@RiffMusic1970 Honestly, I am not totally educated on this so don't quote me on this. But I think it's to leave headroom for the intersample peaks that may become problematic when converting the file using different codecs. So a track with -.2db of headroom might clip when Spotify or whomever converts it to their codec. This is a problem. So leaving 1.5db of headroom gives room for digital services to do their thing without negatively affecting the music. I don't know if this is 100% how it works, but this is how I understand it. I may be mixing and mastering buzzwords, though. If you want to see Eric Valentine's video I referenced, watch the episode on his channel about mastering.
I think it also leaves headroom for your client to convert the file for themselves. If they want to create mp3s of their songs they can do so without introducing artifacts from the conversion process. I think
thank you
Thanks so much Justin! Awesome information ... I love pretty much all of your Videos that I've seen so far.
I work on mainly rock stuff and I tend to land somewhere between -6 and -9 LUFS, even with a lowered ceiling on the limiter. I really don't think about it too much. I use my ears to what sounds best for a particular track. I make mix/mastering decisions for my music, not the streaming platforms. Future-proof your work... these platforms will undoubtably change their recommendations down the line.
Quick question for anyone out there... I've worked with 3 different mastering engineers now and all 3 mastered my tracks to right up to 0.0 db. They were all great at their job and I'm sure they knew what they were doing, but can pushing the peak to "0" cause some distortion on some playback systems? I thought -0.3 db (or -0.2 or -0.1db) was more common. Anyone have any thoughts on this or have any info? Thanks!
I 100% need more of this
Thx man, clear and well to the point. Two questions (between a million others!):
- if doing a mastering level of a track with "True peak" meter ON, do we still have a to go for the -1db "safety zone" for streaming services (recommended)?
I know the inter-sampled peaks mostly can bother us adding distortion when transcoding to lossy formats.
- If the LUFS reading is above -14.0 for a master, then Spotify wants a True Peak value under -2.0. But what if we go -0.5? Any penalty from them on this?
For the LUFS i get i they normalize, but what about the "True Peak reading"?
Thx for your time and your video! And yes i think you can go more nerdy! Gently...
Thanks Justin, please go full nerd mode on this, I think it's really important but not covered whole lot on TH-cam in such details!
Thanks Justin, awesome episode!! It”s really helpful information. No. 1 mastering video on YT 👍
Well, this still isn't that clear. Suppose you have a song that sounds good at -14 and also at -8 and everything in between, what are you going to choose? You get more dynamics with -14, so then that? It seems like the pros aren't doing that, and are just going for louder. Why? Is it just like you said, that the quieter sections of that louder master will sound louder on streaming services? Or does it have nothing to do with streaming services at all? It seems the rule is "as loud as you can get without losing fidelity." Why? For vinyl, or CDs, or radio?
This was, once again, most excellent!! Thank you Justin!
My thoughts on loadness for mastering. Use your ears and make a master that sounds as good as possible and make it as load as you possible can. How much you should compress and if you should use a limiter at the end of the master chain really depends on genre (the more you use compression and limiting, the loader the master can be). I recently remastered a recording I made in 2001. I just used my ears and made an analysis after the mastering was done, it ended up around -9 to -10 LUFS, which somehow is the "normal" level for a rock/hardrock record. Jazz is lower, metal is higher, but the important thing is to be in se same ballpark as your competitors. The band Fata Morgana is not a metal band so I think I did it right. Feel free to listen (just click on me). If you like guitar-driven hardrock with more complex arrangements but still melodic, you should defenetely listen. Remember, your ears are going to tell you whats right. not numbers! // Peter, guitarist in Fata Morgana
Please bring on the nerdy esoteric metering episode.
I measured a few songs off of the new blink album. I think they won the loudness wars. I measured -3 to -4 lufs! Everything was square waves off and I can hear digital distortion.
Hey Justin! Thanks for talking about this stuff!! Good stuff!! Please don't be afraid of going too nerdy and technical!!
Please do an episode on meters. Would love your common sense explanation. Thanks
Probably the best video you have ever posted about music production. I love the honesty, bravo. The streaming services need to do better loudness analysis before penalties matter. Obviously, distortion for a music audiofile is outside of all this conversation.
The more info, the better. WE ARE THE NERDS.
I mainly mix hip hop and RnB. My clients typically like a super loud mix so I tend to get between -10 and -6. I use the god particles limiters and find that it doesn’t complete squash my transients and dynamics as much as what I used to use, the L2 limiter. Gain staging helps tremendously
I've been mastering a piano based track which has a lot of aggressive key velocities mixed with some very soft touches also. If I get my loudness level anywhere within -6bd of 0, the higher octave piano notes played on the heavy side rattle my eardrums. I've done a ton of eq notching and finally decided that it doesn't really need to be at 0. That doesn't mean I don't have a lot to learn, but I'm starting to think loudness level depends on what's in the track
Yeah, loud transients on full range piano can wreak havoc on small iPhone and laptop speakers. It’s often better to master these kinds of pieces quieter because of it. That one instrument causes more trouble in this regard than any other in my experience.
This is a great video on LUFS and mastering loudness. My thing is making sure that my mix sounds balanced b4 I go to my 'My Mastering mode". I've been using a True Peak mastering VST that has an Output Dim adjustment. Maybe a mix example where you then take that mix to master it would be cool.
I really appreciate the detailed, clear, well edited explanations. Thank you.
I loved this.Thank you for the nuanced explanation.Great ! I think you explained things just right for me.
Thanks for the info! I'd love to dive deeper with more details and a 'nerdy' video on mastering and loudness. Could you also recommend the best plugins for this?
Here are some deeper nerdier ones :-)
th-cam.com/video/y4NlW5S8AsE/w-d-xo.html
th-cam.com/video/85-GmZ9YF7k/w-d-xo.html
Hope that helps!
do u have a video on Masters for Dynamics and not loudness like they did back in the day when they did not smash the life out of a record?
Problem with streaming platforms, they butcher the audio quality and let's face it, too many Dummy phones playing thru their speaker and many don't even care about quality .
But KOODOS to you from a Musician who cares about Quality WAV😁👍🏻
This is incredible advice. Thank you.
I've slowly begun to realize... Justin is the TRUTH
Yeah, get nerdy, get down tonight 🎉
Thanks 😊
This was lovely. More please! Maybe with a mini white board to illustrate things like intersample overs?
Thank you for this great episode! 😍🤗 Very well done! Super useful information you packed in there. And I wouldn't mind if you'd go even more technical. I would like that! 😁
Yes we need a lot of technical data, it's the core of our jobs!
For Me Also Same -10 To -13..Because That Lufs Give Best Volume Sound..
thanks for this. i've asked this question many times and could'nt find a real answer. all i found was -14 and im like WHAT. Then i found that EDM -7 and nothing else until this with -8 for hip hop. thank you. i was getting to -10 and it just didnt quite feel as loud so now i know.
Super video. Thanks!!
Awesome! Thanks for watching.
-Justin
The most comon -8 Lufs cause they will mess it up if you try to turn it down and if you just stay with-8 LUFS
You will be in the industry standard so they tell me besides why would we make different levels for different platforms. Just bang it but make it clear I say.
This topic has become almost ideological, which makes it hard to find practical info. Thanks as always Justin for the clarity.
...and definitely, a similar practical video on metering from YOU would be grand, for the same reason!
Regarding headroom and intersample peaks. Would you do anything differently if you had a true peak limiter?
Can you comment on recent drum and bass tracks specifically the neurofunk style currently being as loud as -3 LUFS with True Peaks of +2dB, yet not sounding too distorted? How do the mastering engineers achieve this?
i did not understand this part - when streaming services turn down ur master level, why do the quieter parts still sound louder than a generally quiet song? is it because the louder song's quiet parts are normally louder, or is it because streaming services only turn down the loudest parts?
This is the mistake I made initially. Mastering to a particular LUFS level. Reason some tracks sound louder is cos they have generally shoot for much louder on the LUFS reading. Sometimes -7 to -8 LUFS. Its not necessarily 'Louder' just more dense.
@@soundbeasttv so its the perceived loudness which comes from the added compression?
@@kingillmusic yeh. That and saturation. Anything to bring that RMS level up.
@@soundbeasttv ok, appreciate the response!
Very helpful. But I'm still getting my volume pushed down when I publish on spotify or youtube. I'm mastering at a reasonable volume, matching with other professional tracks, and even putting my true peak at -1, but I still get my song much quieter than others (when publishing). What am I doing wrong?
Justin I've got questions if you get a shot on bouncing things down after the Master. I've seen some casts on dithering and I sort of get it...but are Mastering engineers saving all these copies of single songs? Like dithered, not dithered, WAV, mp3...and do you have a system for the way you name each version.
Thanks for all of the content its fantastic!!
.wav files or .flac files are going to sound the best, because they are lossless. .MP3, .ogg and .aac are going to sound less good because they are lossy formats.
@@johnb6723 my question is more so referring to playback systems and such...like aren't tracks for CDs dithered more than tracks for streaming...for example(I thought I heard that, but I'm certainly no expert hence the confusion!)...and are engineers saving them in all these different ways or how's that work
24:34 bro, mastering is all about being nerdy! lol Thnx a million for this invaluable video!
Yes please, go extra nurdy into technical derails of mastering & related subjects, it's all good stuff.
This was great and confirmed some of my observation. How would you master in general ambient music?
Intersample Peaks: I have found out that most of the hot mastered hit records showing inter sample peaks up to +4dB. Looks like nobody in this field taking care about any headroom below zero and nobody uses true peak limiters. What do you think about that ? Could show some examples if you like.
True peak is meaningless since once it hits the converters it cannot exceed 0. In the digital world headroom is infinite so meters reading above true peak in the daw are ignored by 99% of professional engineers
Yes! In the real world, you can have some significant occasional intersample peaks and not have an issue. However, if you have a lot of very hot intersample peaks, it can potentially cause issues when converted. Listening through potential end codecs with a plugin like the ADPTR Audio Streamliner or the ones made by Nugen Audio can let you know if this will actually be an issue.
After mastering hundreds upon hundreds of projects per year for the last 8 years or so, there has been exactly one time where we heard new distortion added after conversion occurred and decided to re-upload the album. But it made it so that I always check now!
Simply lowering the peak level on that record from -.4 to -.8 did it.
It's true-not a lot of people use true peak limiters these days, because most limiters sound better in their normal mode to most people who use them. HOWEVER, the new bx_limiter True Peak, only has a true peak mode and sounds really good that way. Its detection circuit is a bit different than the norm, and I've found myself starting to use it for hotter material. Very nice.
Hope that helps,
-Justin
@@SonicScoop Thanks Justin, great reply. I just wanted to point this, because almost every TH-cam ”engineer” says: True Peak Limiting is a must and even better bis 1dB of headroom. But if you discover current modern chart music, nobody cares in the real world. I have tested the bx TruePeak Limiter as well and compared to FF Pro-L. In my cases Fabfilter got me better results. So i´m not a big fan of the bx Limiter, but i like the Sonible Smart Limiter. This one works quite well for me. Cheers
i made some tracks mastered at abbey road studio for an album and they were showing +0,4 db true peak !! i sent the engineer a message, he answered me if it sounds good to your ears don t worry about true peak and it sounds amazing so... and i watched a video too about mastering with the engineer who mastered bad guy from billie eilish and he said: never use a true peak limiter... they all sounds bad... so in the real world nobody cares about true peak because i think that if true peaks stays around 1 db or maybe 1,5 db i m not sure that we can hear any difference.
@@michelrostaing2729 that's right Michael, only most of the TH-cam guys cares about true peak limiting. But It does not reflect the real world ;-)
Many pop records go over 0 true peak when I download versions of them. How do they get away with this?
Because it doesn't matter much most of the time? :-) As long as it's not excessive you'll rarely hear it as an issue.
-Justin
Would be nice to talk a bit about how to use K system meters while mastering or mixing.
Stefan, there are a number of YT vids on that subject. There is also the EBU standard now also popular.
What level of mixdowns do you usually get sent in for mastering, what level would you recommend?
What level should my master fader be at before I master.. should I bounce-3db on the output fader 🎚 or leave it zero ?
Very helpful! Thank you.
You can go deeper on subjects. Thx again.
I only presently work on my own songs, none of which fit into the very small Genre List used in this video. Maybe some of mine could be squeezed into the "pop" area, but it would be nice to see what the "blurred Genre of Country" would be an acceptable level, and no one ever seems to bother to talk about worship/CCM areas. I use two ADPTR plugins, Metric A/B, which is THE BEST I have seen (you can also upload 16 reference tracks), and its Streamliner, for comparison which also covers Codec, , along with Waves WLM Plus 5.0 loudness meter. Since I have Studio One Pro 5, it has a super duper Project Page module, the ONLY DAW with a built-in mastering module, in which you can tweak a song by going back to the original track to do that for separate tracks. It would be SO NICE if all these services would come together and choose ONE STANDARD LUFS.
I guess im a DR nut....I just did a mix at -17 LUFS and I really like the space....so fun to crank up those wide DR mixes...I trust folks to use their volume knobs...probably controversial I know. ;) Nearly every album ive spent a ton of time listening to had relatively high DR's.....I like to hear things really bounce around. Thanks for the great content!
How do you sort a mix within a genre that requires a lot of low end? I'm making a doom metal personal project, and one of the things I've run into is that when the low end sounds appropriate for the genre, it's destroying my master bus compressor. With guitars tuned this low, it's basically like having 5 basses in my mix once both sides are double tracked. I'm making space where I can, but sometimes the soup just sounds better. Is the solution to just not compress the guitars at all and ride the faders?
Good length & explanations 👍 (but yeah, get nerdy! :)
Are there downsides of sending to Apple/Spotify at -2 peak?
Like, can’t you just give a +2 (0db) copy to clients etc & send the -2 peak to Apple?
Does it sound better/worse?
Is it not just normalized?
ok this helped. thank you
I've seen comments about pushing beyond -14 LUFS, with some going as high as -6 LUFS. But what's your experience with releasing masters on streaming platforms? Personally, I've found that going above -14 LUFS can lead to issues, especially on Spotify, where it seems to degrade the sound the most. Apple Music handles it a bit better, and TH-cam fares the best. Still, even at louder levels, the master doesn’t sound as good as it does in your DAW or on your speakers. Also, commercial tracks often sound way louder than those mastered at -14 LUFS. Some believe that streaming platforms, particularly Spotify, might have a bias against independent releases compared to commercial ones.
All of these services should simply be turning down louder masters, so there should be no additional "degradation" from any of these services.
However, you may be listening to Spotify in a lower quality audio mode. You can update your audio quality settings with the instructions at the beginning of this video:
th-cam.com/video/L4pf29_u1W8/w-d-xo.html
Ironically, TH-cam has the lowest quality playback of those 3 services if you put all three in their highest quality settings. It maxes out at 256kbps and can also do some subtle HF rolloff above 15k. (But it's barely noticeable.)
So there may be other factors at play to look into.
And yes! Commercially mastered tracks can sound louder at the same LUFS rating, but part of this will be because of frequency balances. Although the scales used in LUFS try to compensate for the quirks of human hearing, this is imperfect, and brighter will sound louder, while excess low end that's found in a lot of amateur mixes with make things sound quieter at the same LUFS reading.
Also, there are little tricks that can be done, such as baking in quieter sections to bring down the integrated LUFS rating of a song, allowing is louder sections to sound louder than a song that's the same LUFS reading the whole way through.
I hope some of that makes sense!
-Justin
Idk when I start to get to -9 Lufs it feels gross. I prefer my mixes to be around -11. At least that's what works for the music I'm making
Yeeessss sir this a hard video , yo bro nba youngboy drop new albun ,u can do a mixing masterclass of any song ? 💗💗💗💗 love too this channel
Great video. Two questions:
1. This -14 to -8 LUFS is a what the LUF meter should read POST mastering, not pre, correct?
2. How does the "Loudness Penalty" play into this? I've mastered stuff at -12 yet the Loudness Penalty website says i'll get hit by up to -4?
Any answers would be helpful. Thanks!
hey JC the audio version of this had some 'interesting' unedited moments, you might want to check that.
Uh-oh haha, I must have accidentally uploaded the not-finished-being-edited version! Thanks for letting me know. Out of the thousands who have heard it so far you're the first one to tell me. My goodness! That's embarrassing :-) Thanks!
-Justin
Thank You
I'll take technical and comprehensive, aka nerdy, over under-educated any day. Thank you!
Thank you for tuning in!
-Justin
yes get technical!! excellent video.. You should talk about Dithering too please!
I got you! Right here: th-cam.com/video/VbRao4yHJG8/w-d-xo.html
Thank you for making this Video on this Subject. ☺️George Amodei
If I'm mastering an album that's going to be pressed on vinyl and it's electronic music, how should the volume be adjusted?
What about soft acoustic music? For example just two guitars and voice..
Usually you’d lean on the quieter side, -13 to -12, but there are cases of modern, loud production aesthetics, even in these styles. Depends on how processed vs natural you want things to sound.
The Johnny Cash American recordings for instance are QUITE loud, even when it’s just a solo guitar and voice. If you want that kind of tone and aesthetic, more limiting is part of it.
Ultimately, you should study records you really love the sound of.
Hope that helps!
-Justin
@@SonicScoop that helps a lot, thank you so much for answering!
Great video. At the end, i didn't understand: you recommend setting the limiter with the true peak of -1 for not extremely loud music (Pop and Rock: -13 to -10 LUFS)? I make rock...
With a true peak limiter, you should theoretically be able to go all the way up to just about 0 as it is catching the intersample peaks. Though may as well set it to -0.1 I guess?
If you're not using a true peak limiter, I usually find that setting the output to anywhere from -0.4 to -1.0 will be enough to keep the intersample peaks from causing any meaningful issues, depending on how hot you are mastering.
This does require a little bit of trial and error. A tool Like ADPTR Audio's Streamliner can help you determine this by looking at intersample peaks and hearing what the codecs will actually do to your master on streaming.
Hope that helps!
-Justin
@@SonicScoop I appreciate a lot your answer :)
Fantastic explanation and example ranges - thanks! Re a 'safe' range for jazz, can I ask what you would suggest?
Probably between -14 and -10 or quieter depending on the style of jazz.
Great Video🎼🕊️
Theres no rules with mixing and mastering! That’s why I aim for -60 LUFS, crank up my amps and bliss to the hisssss.
Lol
😁
On that last question, I definitely prefer technical and nerdy, but sometimes it gets ahead of what I know and needs to be rewinded a bit lol
Do you know if it's possible to master audio for delivery to streaming services like Curiosity Stream in Premiere or Resolve? I would love to find a good tutorial about that process. Thanks for the help.
Hi :D Integrated Lufs or Short term lufs from loudest part?
Hi Daniel, All streaming services go by integrated LUFS, so I’m speaking in those terms too.
-Justin
The loudness of the master should be THAT which is RIGHT for the song,the and the genre. IF THIS MEANS WHATEVER STREAMER WILL TURN IT TDOWN, SO BE IT. The recommended True peak level is the only value that you should adhere too for the different streamers.My experience is that the higher the sample rate and bit file that you deliver the file, the better the sound quality of the song when it appears on the platform, mix and master for for what what believe is right for how the song and bedamned for what the streamers do.KEEPING THE SONG AND INTEGRITY IS THE MOST IMPORTANT THING, MIX AND MASTER FOR THE RIGHT SOUND, if the result comes out very high then ask yourself if you wanted it so loud and maybe let it breath, look at you average LUFS afterwards and make decisions. And Try FIDEF, THIS CAN GET YOUR SONGS TO SOUND BETTER AND ENERGETIC WITHOUT SO MUCH COMPRESSION OR LIMITING ( My little tip of the day…..
I always just shoot for -1.5 on the DAW. Going over 0 distort music. None of the streaming sites have distorted my music by lowering volume.
I need more technical and nerdy PLEASE!!
Go as nerdy and technical as possible… There is no peak level on knowledge!
Hey, Justin! You did a podcast a month ago arguing against the need for super high resolution audio, and I totally agreed with your stance. However, I was thinking while watching this episode that perhaps using a sample rate of 192kHz or higher might actually help to prevent inter-sample peaks while mastering due to a more accurate representation of the wave shape. What are your thoughts on that? Has anyone done any research on that?
Hey Jarrod, that's a fair question! Using a higher sample rate wouldn't totally eliminate this issue, but maybe it would decrease the magnitude of it? I'm not sure. But it's plausible. Would have to look into it.
That said, there is already an option to basically do this, regardless of what sample rate you are working in: "True Peak" limiters use oversampling in the detector circuit to help eliminate intersample peaks.
Of all the ones I've tried so far, I like the sound of the bx_limiter True Peak. But other brand's limiters will often have this mode. They do tend to sound a little different in their True Peak mode.
You'd still need to address the intersample peak problem with a higher sample rate, but maybe they wouldn't be quite as high? I'd really need to do more research and experimentation on it.
But honestly, it's not that big of a deal either way, because just turning down the peak level slightly or using a true peak limiter) tends to fix this without issue, and would still likely need to be done at a higher sample rate anyway.
Hope that helps!
-Justin
@@SonicScoop Thank you for the in depth response!
Hi,
Where are you getting your client references? Are you using Spotify (which has been modified) are you using the commercial CD, digital download? If Spotify, are you normalizing the file to get a proper master equivalent?
I would think the reference needs to be at the actual level to have any relevance.
Yes, you used to be able to turn off volume normalization in Spotify. Not sure if you still can. But you can definitely turn it off in Apple Music and other places. That is important. It's another reason TH-cam is not great for referencing.
-Justin
Vary helpful!!!!
-8 or -9 is def the sweet spot for me ..... mainly do rnb pop style music
The Master should be the same loudness as all other commercially released tracks according to FCC rules/calm act?
Not for music releases, no. The FCC has no jurisdiction there.
The case is different in broadcast, where everything is played back at the same average level.
Streaming services too, normalize levels on playback by turning down the loudest tracks. But the real world experience is that even after this occurs, louder masters can and often will still be perceived as louder.
Hope that helps!
-Justin
More is better. Even break up deeper subjects with part one and part two so everyone is happy.
Go nerdy, thanks !!!
I push my music louder and it sounds like shit. I have to have it down to 18 lufs for it to sound good idk how to fix it
If your music sounds really bad at -18 LUFS then something is wrong. That's super quiet. I'd recommend checking your low end. You may be piling in a ton of low frequency sub energy that you can't hear in your room, but that's eating up all your headroom.
Hope that helps!
-Justin
@@SonicScoop thank you bro
@@SonicScoop also it's my vocals that sound bad. My beat sound great when I push it. I'm using the neuman tlm102 to record and my voice is a little higher pitched for a male. But im going to try what you recommend first then maybe I'll try digging around to see what else it can be
Between -7 lufs and -9 lufs and true peak is -0.1
For some genres and styles that's perfect! For some it's too loud and for some it's too quiet :-)
But if that range works well for what you're doing, keep at it!
-Justin
this is the problem of outdated sound engineers, they still carry the old-fashioned Audio CD attitude. They think they are the ones who created the streaming service, If the streaming service engineer has already made -14lufs is the limit, why does the sound engineer need to break it? do you need to download the audio before you listen to it? that's why I said, if there is still an attitude of loudness war, it is better to put the song in the CD store. If you want to really compete in streaming services, try mixing and mastering that sounds loud and clear without changing the -14 lufs, if you succeed, you are a very great sound engineer in this era.
I'm all about the nerdy stuff. If I don't get it, it will push me out of my comfort zone when I try to understand it.
-2 LUFS only right?
-8 to -6 on my final masters for rock anything really. 99% of the songs I ref are fucking loud. Overall Eq plays a huge part. Sometimes when I think it needs to be louder, it realy needs more top end.
Dan Worrall needs to watch this.
Should I invite him on as a podcast guest?
-Justin
It'd probably get you a good amount of views, but he seems to be pretty smugly opinionated about a lot of things so it might just be an exercise in frustration. If you do have him on, ask him to show his mixing and mastering credits. I can't seem to find any. @@SonicScoop