Wolfgang Lonien

As you might know in case you’re reading this blog once in a while, my DAW (digital audio workstation) of choice is the free and open source tool called ‘Ardour‘. And almost 2 weeks ago, the project announced its latest release 6.3, with the list of changes here.

And today I finally found the time to download (I’m a supporter, otherwise that would cost you at least 1$) and to try this version, which I did on Windows (I normally use it on Linux, but on that platform I always wait for new software being added to the repositories – too much to explain here and now). And I tested this new feature, the new Loudness Analyser, with a cool song from a few friends from Wikiloops which I had downloaded but not worked on yet (so it doesn’t have a bass) – hear Marc’s original remix and keyboard add here if you wish:

Ok; so I’ve put this track in a shared drive which the machine can “see” both from Windows and Linux, so first of course I started Ardour 6.3:

Then I added the track above (“Sunny Garden”), and checked the tracks loudness with right-clicking on it. Result:

As you can see, the integrated loudness is -15LUFS, and the True Peaks are +0.9dB – so integrated is perfectly fine depending on what you’re about to do with it, but the peaks are a bit hot and even distorted.

You’ll find the new Loudness Analyser tool in the master channel/bus of the DAW, on the right hand in the following screenshot, above the master fader where it says ‘LAN’ and ‘0.00dB’:

If you click on that LAN tool, a dialog with an explanation opens:

And if you further click on ‘Analyze’ it will show you its default setting which is EBU R 128:

Now EBU R 128 is for the European TV standard loudness which as you can see is -23 LUFS (integrated), with True Peaks of -1dB. And while that is perfectly fine and always recommendable in case you want to send something to a broadcasting station, most of us don’t do this, but rather use some streaming services to upload to – all of which have different settings to which they’ll reduce your track in case it’s too loud for them. Here’s a list of choices you have:

You see that for instance Youtube which is currently selected would (and will) reduce such a track to -14 LUFS and to -1dBTP – and under the ‘Measured’ column you can also see that for Youtube the integrated (average) loudness could even be higher, but since the true peaks aren’t -1 but +0.9 it will reduce the total gain by an amount of -1.93dB. If we do that ourselves here we can at least check the result *before* uploading it elsewhere…

I always use -16 LUFS and -1dBTP as my target, which is what the choice of the ‘Apple Music’ streaming service would also do. So all of the tracks of all of my albums in Wikiloops (which *I* remixed last) have that same loudness, and to you, dear listener that means that you won’t have to always look for the volume knob in case you’ll hear one of those in the car or elsewhere… and note that the loudness reduction for *this* track would be absolutely the same as when deciding on Youtube as the ‘target’ – because of that peak of +0.9dB somewhere (and note that peak also shown in both the channel and master strips in this screenshot:

So the reduction in this case is also -1.93dB. So, ok – I applied this as my setting for this song and export, which leads to this changed part in the master channel:

So during exporting of the song, I checked that I want another analysis *after* that export (to see the result), with setting the checkmark here:

And after the exporting is done, I see the new analysis:

So now the track is at -16.9LUFS and -1dBTP which is fine, unless of course you even care for that last 0.9 dB of integrated loudness – in that case you’d have to go back to the channel and find that peak marked in yellow in the left channel of the above screenshot, correct that a bit (like with an automated fader down of a dB or so over it), and repeat – it’s an iterative process if you really want to be the ‘master’ engineer of your track(s).

Anyway, I’ve listened to that normalised track with my calibrated headphones again, and be assured that Marc’s (and Oliv’s and Martin’s) track still sounds wonderful.

And what I also find wonderful is that even in a no cost (or low cost with the 1$ for the download as minimal selectable amount) environment we now have tools like these which really make life a lot easier even for us non-technicians. So for today, thank you for reading, and I’ll let Zuleikha (an older photo of her) greet you from the background of my Windows partition: