"If the recommended loudness unit levels are exceeded the services will quite often turn the music down leading to horrible limiting artefacts and a generally unpleasant sound." Only if those things already existed in the master. Normalisation is almost always just a clean gaining down of the original master. Unless someone is using the loud setting on Spotify of course, as that is the only setting on any service I know of at least, that uses a limiter. Even then that is to bring a lower level master UP, not down to that louder 11 LUFSi. Which lends more to the argument of definitely mastering hotter than -14LUFSi, not the other way around. It's fairly moot anyway as less than half of Spotify users actually have premium and thus access to those normalisation settings, a lot of playback systems don't have normalisation at all like on 3rd party players like Sonos systems and even of those that have access to those settings it remains unclear how many people are really aware of them never mind change them. The only figure i've seen is from Ian Shepherd saying only 17% of users change this setting (although this is unsubstantiated), which considering only about 46% have access to it to begin with is an incredibly small number. Master to make it sound good. Worrying about the streaming services normalisation specification is counterproductive and if anything, is more of a floor, not a ceiling, in terms of the loudness of your master.

I can watch this repeatedly - so it sinks into my brain and is fused there.

this clear explanation helped me a lot to place in order all this knowledge even when I work everyday as an electronic thecnician since I work with outboard gear

@PresentDayProduction

Жыл бұрын

I’m glad to know it helped! :)

I totally understood all of this and definitely was not left foaming at the mouth. Now, where’s my paracetamol … 🤯😳😰

Great video as always! But 0 dBSPL is the smallest sound pressure a human can hear only at 1kHz, that's the reference point! For example according to the Fletcher-Munson curves a 50Hz sine wave should be at around 45 dBSPL to be audible.

I prefer to work in DB-FU

You’re so articulate! Thx! 😀

@djalexia

2 жыл бұрын

Simply the truth! Keep the good work up! Oh! Hope that Focal will send you the EVO so you can finally give your audiences a good ol’ review. If I may take a liberty to suggest, see if PreSonus are willing to send you a pair of PreSonus E. XT7 or XT8 I know already components wouldn’t be top notch but believe me I truly enjoy the sound of those speakers… I got them as trial and ended up keeping them. Anyway… till next time.😊

The explanation of true peak was frigging good. That’s been somewhere in the back of my mind “I ought to learn what this is some day” as I routinely pick it in limiters etc knowing it is a better measurement. Given my digital engineering background and (not very successful) signal processing studies it was a bit of humble pie to learn that this mysterious audio term was something pretty close to my own home turf :)

Thank you James

@PresentDayProduction

10 ай бұрын

Cheers David! :)

Great job, very well explained! It's probably worth mentioning the K-weighting curve, which is the one LUFS (originally LKFS, Loudness, K-weighted, full scale) use to make the measurement behave closer to what the human hearing does in terms of frequency response. But again, great job in making such a hard topic so clear and understandable!

@PresentDayProduction

2 жыл бұрын

Cheers Stefano, well pointed out! :) thanks for watching

Awesome. This is exactly what I needed to hear. Thanks for the very helpful explanation and a fantastically well-produced video. Thank you, James. 🙂❤👏

@PresentDayProduction

2 жыл бұрын

Cheers Deano!!!

Brilliantly explained sir! If that was off the top of your head then you're a seriously sharp young man... I understand some of this but certainly not all! great job anyway.....

Awesome! Thanks!

Great video! I wish you had touched on Bob Katz’ K-weighting system but this covers all of the essentials for most situations and then some!

Material that exceeds the streaming platforms’ reference level will not exhibit any limiting artefacts when turned down - no limiting is involved, merely an overall gain adjustment. Also: Apple Music (and others including Spotify IIRC) recommend a true peak ceiling of -1dBFS (or -2dBFS on louder material), not -0.1 dBFS.

@PresentDayProduction

2 жыл бұрын

I think he said -1dB ceiling, if not, that's what he meant. But I run everything through Apple's encoder and turn the ceiling down until the red lights go away. Mark x

@PANTECHNICONRecordings

2 жыл бұрын

@@PresentDayProduction Cool - I heard -0.1 I think, but if I did, I’m sure it was probably a slip of the tongue!

@PresentDayProduction

2 жыл бұрын

James here - definitely said -1dBFS! Probably just my internet voice not being on-point ;)

@PANTECHNICONRecordings

2 жыл бұрын

@@PresentDayProduction Watched again, and yes, do did say -1. I stand corrected!

I’ve worked in the high end home audio (HiFi to those stuck in 1970 … poke poke) for over 20 years. One of the devices I never wanted a consumer to get their hands on was a dB meter. Most had no concept of its use and would then tell me how to do my job. If they had watched this video they may have annoyed me a tad less.

LUFS out Loud(ness). I shall henceforth direct any relentless gain merchants I encounter to this video, in the vain hope that they'll finally grasp why greener is cleaner.

Very useful video and a rare topic. I think I last read about different db scales more than 25 years ago, in an article about companies trying to decept customers by using "friendly" db scales. Mainly for signal to noise ratio in pro-sumer products. LUFS wasn't a thing back then, although radio and TV probably already measured it. Radio was slamming every track to death, but independant wouldn't get airplay anyways and we were in the heights of the loudness wars. Excellent video, again. Thanks!

@PresentDayProduction

2 жыл бұрын

Glad to know you found it useful! Thanks Peter! :)

@secretelitemusic

2 жыл бұрын

There was a trend among cheap (& nasty) speaker and amp manufacturers back in the 20th century to rate their kit in PMPO (Peak Music Power Output) rather than WRMS. Hence the profusion of 2kw PMPO car subs and amps that were actually about 50WRMS. The coils on the subs would frequently unspool before the cones tore themselves to pieces. Any time a cheapskate manufacturer can manipulate data to confuse consumers for profit, they most certainly will. Caveat emptor when it comes to tech spec sheets, even from "reputable" manufacturers, because even they may be using selective comparisons to dupe the credulous.

very well done

Here's what I learned, although this was splendidly explained it is a compelling case for paying someone who does this all day, every day. It is the limit of my tolerance for technical knowledge. I love writing and recording and mixing but it ends there. Mastering can remain a dark art for me.

Thank you

Hopefully people won't intepret the LUFS thing as "you should aim for -14LUFS for Spotify" as they often do, because that's not what you should do.

The good news is that kit like Clarity M (pictured in the title Image) , Smart Limit, Brickwall HD, and their ilk, can give you some quality visual feedback on the subject, and even some streaming service presets to work with. Your ears can easily finesse the fine detail if you've allowed sentient machines to do all the basic hard work. We lufs U Skynet. edit: If it sounds terrible when all of your loudness settings are perfect, it may be time to return to the scene of the crime. Namely, The mix.

hi, I really love your channel, in fact, your channel is the only mixing&, mastering channel I watch I have a question I watched your kali lp review I liked how honest you are I bought kali audio in5 about one month ago based on the reviews but I really want to know what do you think about the kali in series thank you for the time you but on the videos see all appreciate it. your fan from Jordan

@PresentDayProduction

2 жыл бұрын

Hey there! Unfortunately we haven’t actually heard the IN-8! We would like Kali to send us a pair but I haven’t had a chance to ask yet, so we will see how that goes!

Great video and absolutely perfection when it comes to explaining dB.

Ironically, the horrible limiting artifacts appear when your master is below the normalization threshold, in which case it gets turned up with bad limiting. If it's above, it's just gained down. The takeaway: make it sound good and don't worry about LUFS

Thank you so much for another awesome video! When you say the master is set to -1dB do you mean -.1? This always confused me when I use the izotope mastering limiter.

Noice! Big ups guys.

Cool story, I was once looking at flats, and really wanted to move to one because the postcode ended in 0DB.. but the flat was shit, so regardless of the cool postcode, it was a no

@almightytreegod

2 жыл бұрын

Did every building have super loud music inside yet unnatural silence on the streets?

@1337murk

2 жыл бұрын

@@almightytreegod lol

As ‘decibel’ is a mathematical function, quite literally a fraction on a logarithmic scale, dB has no meaning. In fact 3dB has no meaning either. When we say add 3dB, the word ‘add’ is what gives it meaning. Likewise a + sign. +3dB has meaning. Yes, it is confusing but really doesn’t need to be understood in mathematical or historical terms. Just remember that a number followed by dB is a multiplier of whatever it is that’s being measured. 20dBu denotes a voltage of 10x the agreed standard of u=1vAC, or 10vAC. It gets even more complicated because AC voltage can be described as peak or average but that’s not important here. 6dB, or very close to it, represents 2x our measure of sound. 10dB is 3.16x our measure and 20 is 10x. 40=100x and the familiar logarithmic function continues. 60x = 1000. When we say RT60, we mean Reverb Time (decay) for 60dB loss, or one thousandth the original sound’s measure. That’s as good an indicator as any that sound needs a logarithmic scale or we’d be burdened with very large numbers. What IS important is to know what the references mean and how they apply to your job. As audio engineers, VU and dBm can be effectively ignored as they no longer measure anything of importance to us. The ones that matter are: •The letter ‘u’ refers to AC voltage, often 1vAC RMS. This is a technical measurement useful for comparing equipment capabilities. -6dBu would indicate .5vAC. •FS or Full Scale. OdB FS simply says this is the adopted ceiling for those of us working with digital audio (that’s all of us) and it is a relative newcomer. We measure our audio signals almost exclusively as a value of dB below it. -20dBFS means we can increase the signal voltage 10x before running completely out of headroom. SPL or Sound Pressure Level. This is a created scale with an obtuse and unimportant reference of absolute pressure called a Pascal. But 94dB SPL, which happens to be 1 Pascal, has real meaning to our ears. We can create a sound that measures that value and hear what it sounds like. Incidentally, microphone sensitivity is expressed in a number of dB below 1 Pascal. Handy to know. The math is not important, but the references and key numbers are. When the reference is missing, you are not being given enough information to make a judgement. Btw, your description of dBu is actually the proper description for dBm (1mW dissipation in a 600 ohm load). This was in wide use in the days of analog telephony and carried over to audio in the old days when we actually added a 600 ohm resistor at the load end of every single path which resulted in a ton of wasted energy. dBu was the result of eliminating this 600 ohm load and the u actually comes from “unloaded”, a measure of relative voltage independent of impedance.

Well done James, you only half fried my brain! I can remember back in the day when some venues had sound loudness meters, that if you're P.A. (Public Adress) system exceeded a given volume would trip out the stage power. I'm to long in the tooth to remember the level, but we always rated our system amps output in RMS (Root Mean Squared) How does RMS relate to any of the measurements that you described?

@PresentDayProduction

2 жыл бұрын

The RMS is the average safe amount of power a system can deliver without exceeding stress levels. For example, a speaker that has 200W RMS can output 200W of power continuously for a very long time, and can PEAK at maybe 210W for SMALL amounts of time. If it was constantly putting out more power than the RMS then it would fry.

@allen394

2 жыл бұрын

@@PresentDayProduction Thanks James, on reading your reply I remembered that RMS was related to Ohms Law.. Am I right in assuming that the drivers (speaker) output spl relates to your measurements and that the measurements you describe are pre ampllcafion? PS. How's Flop Cats herring, does he flop because of an imbalance in M/S listening?

Oh look, Floppy! :)

LUFS this stuff

Brilliant! My only critique is to try another font with true lower case characters thats easier to read - EuroStile is close to what you are using.

@PresentDayProduction

2 жыл бұрын

Thank you! And I appreciate the feedback - I always use the same font to keep channel branding on-point, but you’re right that this one was probably one that could benefit from a different font. Great critique and gladly accepted!

Everything is fine except the fact that the platforms limit causing a noise or dostorsion, the platforms normalize and do not limit, in the spotofy specs it says that if the song is not high enough it will apply a limitation with the treshold at -1dbfs, to raise the volume, but it is not refers to those that exceed -14 lufs, this may change but to this day I have not found that the information you say is correct, please check this info and let me know if im wrong, thanks

excellent video as always

After watching this my IQ is now over 9001 thank you, conservatively speaking my EQ rounds off around that frequency. Although I prefer DBz , not really a fan of DB GT, but dbfs keeps the clients happy.

Scoobie doobie doo! Thats what I have understood! ;P

The question is do the streaming platforms put the signal through their (mostly crap) codecs and then adjust for volume or do they adjust the volume first and then put it through their codecs…….because if they adjust volume first the it wouldn’t make any difference what levels the original files are sent to them, if the volume is adjusted first then theoretically there should not be any distortion from peak and/or true peak levels, unless, like I said their codecs are just crap, see some of their mobile transmission rates

Exceeding the recommended LUFS will not lead to «horrible limiting artefacts and unpleasant sound» It will just get turned down. Most of the top productions are way louder and may also clip true peak.

@thomasberdal3049

2 жыл бұрын

For example Happier by Marshmello, 930 000 000 views is -6.5 LUFS and peaks at +1.3 db true peak.

Thank you very much! I prefer dB fcu. Flop cat units.

How, exactly, does turning something down add limiting?

@PANTECHNICONRecordings

2 жыл бұрын

It doesn’t.

Meerkat is a Erdmännchen in German , a little earthling……

@murraywebster1228

2 жыл бұрын

A little earth boy actually

Decibels true peak (dBTP) deserve some more attention.

i was here to learn about the new standard "db-wtf" 🤣

Is it ok to use Izotope RX (Advanced) to process a -14LUFS master down to -16LUFS? (From my Spotify master to an Apple master) (Using the loudness processor) Rather than going back to the project, changing my limiter then re-exporting which takes a while. Thoughts?

@thomasberdal3049

2 жыл бұрын

Do not do that. Master to make it sound good, not hitting numbers. If it sounds good at -9LUFS keep it at that level, it is only going to get turned down.

@zubinel9540

2 жыл бұрын

@@thomasberdal3049 Hi, sorry, not sure you've totally understood my question. I run a mini label and have a lot of masters (most of which are already mastered for Spotify at -14LUFS) so re-opening tons of tracks to limit down, test etc will take a long time. All I'm asking is whether Izotope's loudness processor is good enough to bring a master down from -14 to -16LUFS in order to release on Apple music. In my own tests I've tried this and I think it sounds fine, I can't hear any difference, I just wandered if anyone else had done this.

Still waiting on standards paperwork to come back for flopcat units, eh?

@PresentDayProduction

2 жыл бұрын

FlopCat Units are a measurement of weight, energy, volume, speed, and whatever else I can think of!

@valleywoodstudio7345

2 жыл бұрын

@@PresentDayProduction naps!

dnb?

OK, so now I'm confused. How does this relate to James Bond, and does it mean that a DB 7 is better than a DB 5 because it's nearly twice as loud? Plus if Bond is so cool how come it's the baddie who has the cat? These are weighty matters which keep me awake at night!

Great video, but I don't think your section on LUFS is completely accurate. If you exceed the LUFS targets of a streaming service, you are correct that they the will indeed turn your mix down. But this is just like turning down the master fader or adjusting the trim pot. There won't be any unwanted artifacts here. The issue arises when your mix is below the target. In that case, the service will turn up your mix and, in order to avoid clipping, they will run it through a limiter and you don't want that. (NO LONGER THE CASE, SEE EDIT BELOW) This is what is described in Spotify's documentation: artists.spotify.com/en/help/article/loudness-normalization EDIT: looks like Spotify actually changed how they do things! They no longer use a limiter FOR MOST USERS. If you have normalization enabled at the "Normal" setting, they'll only bring up the level to -14dbFS IF THERE IS ENOUGH HEADROOM. So neither exceeding nor being below the target result in artifacting. However, when using normalization with the "Loud" setting they will apply gain to bring up the loudness to -11dBFS. In this case they'll slap a limiter on to avoid clipping. This may result in unwanted artefacting. In any case, being above the target causes no artifacting. Being below -11dBFS may result in unwanted limiter artifacting under certain circumstances.

@janr5845

2 жыл бұрын

Copied directly from the link above: "Negative gain is applied to louder masters so the loudness level is -14 dB LUFS. This lowers the volume in comparison to the master - no additional distortion occurs. Positive gain is applied to softer masters so the loudness level is -14 dB LUFS. We consider the headroom of the track, and leave 1 dB headroom for lossy encodings to preserve audio quality. Example: If a track loudness level is -20 dB LUFS, and its True Peak maximum is -5 dB FS, we only lift the track up to -16 dB LUFS. Premium listeners can also choose volume normalization levels in the app settings to compensate for a noisy or quiet environment Loud: -11dB LUFS Note: We set this level regardless of maximum True Peak. We apply a limiter to prevent distortion and clipping in soft dynamic tracks. The limiter’s set to engage at -1 dB (sample values), with a 5 ms attack time and a 100 ms decay time. Normal: -14dB LUFS Quiet: -23dB LUFS"

☕️🤠👍🏽

But what exactly does "foaming out the mouth" mean?

I’m A-Weighting a plain English for Dummies version of all this techy-babble. In your graph (at 2:08) AIEEEE is the screaming noise made by comic-book Nazi Soldiers & Red Indians when the imperialist white bloke hero shoots them with his Webley. Does Floppy mind being called a ‘mere cat’ at the end..?

@PresentDayProduction

2 жыл бұрын

He’s more than a mere cat, he’s compared the meerkat!

@CheapoCardCompany

2 жыл бұрын

@@PresentDayProduction Was the event comparing Floppy ‘to’ a Meerkat compered ‘by’ a Meerkat?

1st Helloooooo

K ;(

I am an amateur, trying to learn. But I am an electronic engineer, so I know the logarithmic math behind a dB. Well, I tested my home made master where I try to stay at - 14 dB LUFS. I was angry to discover that 100% of today's pop productions play much louder. So? Simple, they do not care to be slammed. They love that. And 100% of these productions sound like shit on my wonderful self built hi-fi system. What are these rules for?

@thomasberdal3049

2 жыл бұрын

They are not rules.

What's up with the audio for this video? Very S-esy & Plosive

@PresentDayProduction

2 жыл бұрын

Thanks for the comment - it didn’t come across that way in editing and every one of our videos has the same dialogue processing method applied, so if it’s slipped through the net then I’m not sure how - hmm!

Flop