[MLA] Loudness normalization with MQA

[Kevin Gross]

That is not in line with the published ReplayGain standard and would
introduce non-factual metadata. I assume you'd propose using the album
value in both track and album metadata. To make this fly, you'd have to get
consensus that track-based normalization is not useful. I beleive Bob Katz
introduced a dance-mix use case where he asserted that track-based
normalization was useful.

The better solution is to make players work properly. Make album
normalization the default mode. Most players handle missing loudness
metadata by applying a fixed, sometimes configurable, attenuation. Players
in album normalization mode can use the track normalization metadata if
album data is missing. Players should also have strategies for handling
playback of material with missing peak level metadata.

Perhaps such advice could be included in the ReplayGain 2.0 documentation.
I haven't made progress on finishing that. I have been working on paid
work, AES67 and live video networking technology and so my time has been
scarce lately.

On Mon, Feb 6, 2017 at 3:59 PM, Grimm | Eelco <eelco at grimmaudio.com> wrote:

> > Hey guys.... you probably know my prejudices:  Use album loudness data
> all the time. If you listened to the demo that I produced you would have no
> question about that.
>
> The question is: replaygain has two fields of loudness metadata: one for
> track loudness and one for album loudness. Unfortunately many services just
> read the track loudness metadata. We _could_ advice Tidal to write the
> album loudness values in the track loudness fields. What's your opinion on
> that?
>
> All the best,
> Eelco
>
>
> _______________________________________________
> MLA mailing list
> MLA at grimmaudio.nl
> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://grimmaudio.nl/pipermail/mla_grimmaudio.nl/attachments/20170207/4969fd5e/attachment.html>