Wrong.
'dynamic range':
there is no official definition for it, and it may be confused with the dynamic range of a recording medium, which is basically the difference between the highest and lowest level it can handle.

The term 'dynamic range' is reserved for the measure of signal-to-noise ratio of a recording medium.
This is why today we talk about "loudness range" as established by "EBU 3342" document and not dynamic range which, as a term, leaves many incorrect interpretations.

"(*= because they are using only analog gear and process it using real mastering 1/2" tape machine etc)" This simply is not real.

 

Last edited: Nov 2, 2018

Peak normalization in post-production increases signal-to-noise ratio.
Peak normalization during rec increases dynamic range of the recording medium.
Loudness normalization is always destructive ? YES and NO.
YES: when the level of your music is lower than the streaming-service target.
No: when the level of your music is higher than the streaming-service target.
P.S.(SoundCloud don't loudness-normalize but encode to mp3, so be aware of TP mastering at -1/-2 dbTp

 

Last edited: Nov 3, 2018

Crest Factor is totally indipendent from RMS and LUFS. Crest Factor is relevant only to the wave-form. There is no direct link between crest factor and loudness

The crest factor is the difference between the peak level and the RMS wave-level over the course of the song. Intuitively, it measures the amplitude of the emerging 'peaks' in the audio stream. It tells you how much a waveform is compressed.
More compression generally means a lower crest factor.
Let's me explain:
Imagine you're listening to some music. You want to it louder. You walk to the volume control, and simply raise the volume. By doing so, you increase the signal's RMS, increase its peak level, and leave its crest factor untouched. We'll call that the 'first loudness paradigm'. Suppose now that you've got a region in your DAW that peaks at 0dBFS. You can't raise its volume in the traditional way, or it's going to distort. But you can insert a limiter, and lower its Threshold slider. By doing so, you still increase the signal's RMS, but this time its peak level remains stable and its crest factor gets reduced. That's what we'll call the 'second loudness paradigm.

Normalization(Peak):
1)-don't modify crest-factor.
2)-Modify RMS level and LUFS.
3)-Increase signal-to-noise ratio.

 

Last edited: Nov 2, 2018

Can you answer these questions?

1) Do both big stars and unknown artists upload to online stores(itunes etc) their pop songs mastered to as low as about -12 LUFS?

2) At the same time, CDs are still being sold (and pop songs on CDs are mastered to rather louder levels like to -6 or -5 LUFS)
So do artists who want to sell their song both on itunes and on a CD make two versions of mastering? (1st version around -12LUFS for online stores, and 2nd version e.g. -5LUFS for CD?)

 

This is essentially the same question and the answer is of no use to anybody. If you post to iTunes at -6LUFS, iTunes will crank you down to -16LUFS when they stream your music, similarly if you upload at film levels (-23LUFS) iTunes will crank you up to -16LUFS. So whatever you do, you're coming out over the wires at -16LUFS. What Pros do will depend more on the expectations of their genre rather than the expectations of businesses that adhere to AES standards for streaming music.

The choice is yours. You will sound better mastered to -16LUFS/-14LUFS Integrated and undergo little cranking than you will going out too hot and getting cranked down. Would a label pay for that difference in mastering? If you're Taylor Swift, hell yes, if you're The Grondfurklers From Planet Zoob on a single EP budget, hell no; you'll be mastered to whatever sells most units in your target market.

Bear in mind that you cannot maximise for vinyl the way you can for digital. You must keep more headroom for MP3 than you do for CD and all and every DPP image you order will all cost you an exact multiple of a single mastering charge at any ME who actually knows what they're doing.

My advice: master to -16LUFS, -1dBTP. That'll sound good whatever you do with it and use somebody, anybody, else to make your master.

 

This is not true. At all.

First of all, Apple uses it's own proprietary algorithm. People keep repeating numbers like -16LUFS and -14LUFS, but I have not been able to find evidence the algorithm actually even uses LUFS. No doubt it uses something similar though.

Second, it is a normalization scheme. They don't compress or limit your audio in any way. So if you for example upload very dynamic material with few peaks hitting 0db or near it, they won't slap a brickwall limiter on it to bring it's 'loudness' up to level with other 'louder' material on the service. You can't bring all material up to let's say -16LUFS by adjusting the gain (or you can but that's just clipping). It simply does not work like that.

Member Peek just explained how this stuff works, what normalization means and what it doesn't mean. Read his/her post.

 

You can see that what you claim in the first quote is proven wrong by your second quote.. the one you actually took time to dig up some info to use for reference.

When I spoke of loudness normalization (RMS or LUFS) tampering with dynamics by changing the crest factor of material by way of auto-compression.. I meant exactly that. The crest factor (difference between RMS level and peak level) is a way to measure how dynamic the audio material is. A peak of -1.3dB with an RMS average of -8dB is not as dynamic as a spectrum with a peak of -1.3dB and a -12dB RMS average. Common knowledge, but for some reason people refuse to latch on to the simple things they "already know" if it's for the sake of just being in the mood to argue. More importantly, a song has good tonal balance when the RMS goes up and down with the peak and the crest factor maintains the same gap throughout.

Most audio masters have peaks in the songs that come near zero even when they are dynamic songs and not squashed. There is not much room to toy around with batch loudness normalizers and such varying sources, and the audio streaming sites have learned that lesson. That's why they began to compromise with their approach to process an entire album of songs in the manner as if it was one long song that has the duration of over an hour (so it won't attack the people's material as destructively as individually forcing every song that reaches their platforms to measure at -14dB LUFS after processing)

 

Last edited: Nov 3, 2018

So the mastering experts are now YouTube, SoundCloud, iTunes and Spotify? And normalizing to their standards? Are you effin crazy?

44.1/x 48/x or 96/x PLZ! where x = 24>16

With a 24-bit file, headroom becomes highly midgetized!

Are we all talking about the same thing??

 

Last edited: Nov 4, 2018

Excuse me, but I have only one pair of ears, Dude. Welcome to the real world.

 

www.lufsmeter.com

 

No.

Linking a random youtube video isn't really an argument. It's just linking a youtube video. It doesn't cite any sources or original research. The video is made by Meter Plugs. It's a company that sells mastering software.

If you go their website and open up the manual for their flagship processor, it states:

iTunes Sound Check
This iTunes feature allows you to hear all of our songs at a similar volume. Although Apple has not publicly stated how they perform volume normalization, independent tests suggest that songs are normalized to -16.5 LUFS.

I guess it's borrowing from an article on Sound On Sound magazine or at least referencing the same test:

Apple haven't stated publicly what target loudness value its Sound Check function currently imposes, Bob Katz claims it to be an entirely credible -16.5 LUFS.

There is also for example a paper by Thomas Lund of TC Electronics stating that:

Apple devices, for instance, target –16,3 LUFS when the Sound Check function is enabled.

The paper does not cite a source for this information.

There's also a page by a South African university that says they are actually targeting -18LUFS.

There is also this AES Convention Paper by Dolby Laboratories which states:

...While it is still possible to utilize loudness levels encoded into the payload, systems have been designed that rely on a single value per file. iTunes is using such a value, called “SoundCheck”. Little is known about the specific algorithms used by Apple.

Now why would they be saying silly stuff like that since we know from your youtube video that Apple is targeting exactly -16LUFS? Because we don't know how their normalization scheme actually works. Most reputable sources bother to state that.

Which is exactly what I said.

Spoiler

Apple uses it's own proprietary algorithm. People keep repeating numbers like -16LUFS and -14LUFS, but I have not been able to find evidence the algorithm actually even uses LUFS. No doubt it uses something similar though.

 

Last edited: Nov 4, 2018

Yes, now I understand. Sorry mate

 

I wasn't aware this was legal issue. Good luck.