[MLA] More good news (Spotify)

[Ian Shepherd]

Hi Thomas,

Thanks for the feedback.

> recently we’ve seen a host of insanely loud tracks that probably saturates the SC measurement; result being that they end up way too loud. Like 12 dB or more

That’s a scary prospect !

> Playing around with HF tones can sometimes buy one or two dB extra with BS.1770, but at the risk of angry customers, burned tweeters and defunct lossy codecs. Another technicality we shouldn’t take too seriously for all the confusion and ripple effects it would cause. Should it ever become a problem, filter or reject at ingest

I think we need to take this issue seriously. I’ve seen a lot of people accepting the idea that normalisation is a win-win, but I also get a lot of kick-back from artists and engineers about loudness management - these are people who have invested in the idea that “loud is good”, and don’t want to believe the alternative, regardless of objective facts - rather like people with entrenched political beliefs. 

Arguing with people like this is ineffective, but they make a lot of noise and we need the results of our alternative message to be as plausible and effective as possible to help win “hearts and minds”.

So for example if someone is “on the fence” about normalisation, and experiments with more dynamic masters on TIDAL (say) they will quickly become discouraged if they find other examples that “beat” them, even by a few dB. Bob L’s experience with the Avenged Sevenfold album is a classic example. Luckily in this case they are just going to ask for a slightly louder re-master, but we only need a few mainstream artists going online to rant about problems they’ve had with normalisation to cause a lot of damage. In the case of SC and ReplayGain we can say they’re not very effective algorithms, but since BS.1770 is the benchmark, we need it to work as well as possible, in practise.

It’s not realistic to think that services like TIDAL, Spotify etc will “reject at invest”, and users will rebel at any suggestion of filtering, so I do think some workable method to deal with this technique is worth seriously considering.

Ian