Hello all,
Some while ago, I have normalised all my audio files with the EBU Normalisation Method (EBU R128 LUFS + dBTP) to -10 LUFS.
Now we want to change the LUFS to -14 , but when I try to normalise the files again, all the Audio Values stay exactly the same, so apparently nothing happens with the new setting.
How can I solve this? I see that in my mAirList 7 folder, there is a database file ‘peakdata.db’. Should I delete that file in order to get the new values?
Thanks for your help.
The only value that would have to change would be “Gain (dB)” under “Playback” for the respective songs.
The level values of the songs remain the same.
Is that exactly the case for you?
Hi Liam,
No, the Playback Gain does not change either.
Thanks for your help.
Please excuse my standard questions, but it’s better if I know exactly how you go about it.
In mAirListDB you select all the songs and then go to bulk edit, right?
From there you click on “Normalize” and then on “Apply” → “OK”?
Yes, exactly. That’s what I do.
Or I try it with an individual file with “Analise” on the “Audio File” tab of a song.
However, the values there should always remain the same, regardless of whether we are working with -14 or -10 LUFS.
The only thing that would have to change would be the value in “Gain (dB)” under “Playback”.
I restarted the database app and now the “Gain (DB)” value changes. Thanks.
Depending on the track itself, sometimes it can happen that the amplification value rermains unchanged. Especially when you decide for such a high target value.
How is this possible?
The R 128 looks at two parameters at the same time:
LUFS)dBTP)Normally, the True Peak should not be higher than -1 dBTP. However, if a high target value would calculate that the amplification is so high that the True Peak would come above -1 dBTP, it stops before reaching the expected playout loudness.
Some tracks may fulfil both, others not and so you still have loudness variations in your playout.
Okay, but why do some tracks do not change their amplification after changing the target valu and mass edit?
I think, the calculation went wrong before and does it still now.
Let’s calculate:
+ amplification = target loudness.As long as this matches, everything is fine. But when the calculation looks like this:
+ amplification ≄ target loudness…we have to look for the reason of this inequality.
Let’s say, the analyzed loudness is -10,1 LUFS, True Peak -1,3 dBTP, target value -10 LUFS: It works.
-10,1 LUFS loudness + 0,1 dB amplification = -10,0 LUFS playout at a True Peak of -1,2 dBTP.
It is more likely that the True Peak is above -1 dBTP, eventually it is positive calculated. Let’s say, +4,8 dBTP, and the analyzed loudness is -12,1 LUFS (Jason Donovan, “Sealed With A Kiss”).
Target value -10 LUFS:
-12,1 LUFS loudness + 2,1 dB amplification = -10,0 LUFS playout,
but the True Peak would be +4,8 + 2,1 = +6,9 dBTP, which means 7,9 dB too high.
Correct:
-12,1 LUFS loudness + -5,8 dB amplification = -17,9 LUFS playout.
The target value isn’t reached and the real playout loudness remains below it, due to the high True Peak.
Does this change when the new target value is -14 LUFS?
Unfortunately, it doesn’t. See above.
I think exactly this happened what you were wondering about.
As long as the analyzed loudness added with the amplification is below the target value, you should take the True Peak into account.
If you export your database as a *.csv file and do the calculation like explained in a separate column and you have too much results below your target value, the target seems to be to high.
At least, it depends on the tracks you want to play. If this is music which is Top 40 of the last 20 years, the mastering should allow such high target values.
If your music is more dynamic, this could cause problems. In my private database targets of -14 or -16 LUFS would be to high for a standardized average playout loudness.
Good luck!
Thanks for the clarification, Uli. This makes sense.
All the files are normalised again and it sounds OK.
Have a nice day.
tldr: mAirList will analyse the integral loudness value once and assign it to the file. To meet the target value simply a (positive or negative) amplification (“gain”) will be assigned to the file. If the target value is altered afterwards, no further analyzation is needed (the result would not change anyway because the file remains the same) but only the playback gain value will differ.
Thank you, Tondose. It’s clear to me now.