Link to the original start of the discussion which includes links to articles: https://www.vegascreativesoftware.info/us/forum/post-your-vegas-creations--109464/?page=10#ca756543

It's extremely important to differentiate "too loud" (implying the perceptual response to volume) from "Loudness Standards," which are a set of defined matrices based on Gain (volume), True Peak readings, Compression (Loudness Range), and Short/Long Integration.

So if we say something is "too loud," it begs the question, "to who?" Is it too loud for EU or US broadcast, for iTunes, for Theatrical Surround, or for Gramps? Four questions, four answers.

It is also necessary to ask the question about one's intentions for their internet audio, and broadcast too for that matter. "Competition or Good Audio?" A pretty fundamental question really.

So, if we remove judgment entirely from the equation, we get an example like this. Here we run into a lot of gray areas, but some are clearly defined. For instance, the EU broadcast standard for Integrated Loudness is 24 LU, (the equivalent of dB increments on a volume scale). Also, one suggested LRA (Loudness Range) for narration and voiceovers is 12 dB, and raising the LRA with less track compression has the collateral effect of raising the Integrated number, sometimes dramatically, but without lowering the Gain! Neat stuff, but it takes some study. I've been on this since A/85 was just a proposal, and I'm still just dipping my toes in the water.

If Spotify, and possibly Youtube are kicking in the limiters at 10-14 LU, I would be more careful of uploading anything hotter than that, as well as if I was doing it for broadcast. Slammed -1dBFS "music" tracks are soon to be a thing of the past, and I'm glad to see it happen.

In my earlier tutorials, I leveled everything for Youtube at conservative A/85 audio levels, and it is definitely too soft to compete on Youtube five years later.

So, to counter the tendency to judge the "too loud" comment as being critical, here is my admitted version of "too soft," in anybody's playbook. And please forgive my tendency to play "levels cop," @michael-harrison, it's been a rather "confining" couple of months for me.

Love, Gramps

@Musicvid @NickHope Thanks for the input. Another set of tools for the production bucket

Back around 2000 AD I was interested in making in-house teevee spots with the DC10+ with its infamous Philips/Zoran chip and hardware MJPEG encoder. After having demos bounced back a couple of times, I became interested video/FM standards for ATSC1.

As I pointed out in another thread. check out the online Loudness Penalty Analyzer to see how much YT, Spotify, iTunes and others will (aledgidly) lower a file's volume. FWIW, for web base projects, I use -16 LUFS narrative and -14 LUFS for music (plus or minus 2.0 LU if I am not too neurotic)

"let the streaming services turn it down"
 - Their leveling algorithms are not quality oriented. Jut ask any pro mixer or mastering engineer.

 

With youtube you can see if they turn your video's volume down. Here are 2 examples. The first is a kid deliberately trying to make loud noises blowing and tapping into a microphone. It appears YT has normalised the volume to 45% due to content loudness being 7.0dB, HOWEVER, his audio still peaks. Integrated LUFS shows it to be -7.

The 2nd video is the opposite in terms of audio. it's an ASMR video which is very quiet. It shows YouTube is not reducing the volume and 'content loudness' is -12.8 dB (LUFS value -27dB). I would guess if you don't want youtube to touch your audio and want it to be as loud as possible you should aim for a content loudness of 0.0dB How do you do that? Well from what the internet is saying you'd likely use -14 LUFS (i) but I have not investigated that

Had a look at 1 more video , guy talking to camera, no YT normalisation, content loudness -2.7dB , LUFS = 16dB. So far that 14dB between 'content loudness' and LUFS is a constant

@Former user That's really helpful. And I learnt something important. I checked my recent music release, which I mixed/mastered (in Reaper) at -13.2dB LUFS Integrated / -1.2dB True Peak. I was surprised to see that YouTube seems to think it has 3.7dB headroom before they would quieten it (assuming I'm interpreting these numbers correctly):

Then when I checked the rendered file, I found its loudness to be -17.7dB LUFS Integrated / -5.6dB True Peak. I'm trying to find out why that drop of 4.4dB happened during render (a Reaper issue). But the fact that 17.7-3.7=14 indicates that YouTube could well be using -14dB LUFS Integrated as their standard.

[Edit: Turns out my volume was down because my master fader was down. Doh!]

So anything from iTunes that gets uploaded to YouTube comes with a 3dB safety cushion, but encourages people to stay in that general range.