[MLA] Normalizing on Tidal

Wed Mar 2 17:37:45 CET 2016

Eelco I think I understand what you're saying and it would be great if
there were more systems where normalization adjusted the final volume for
the setting (regardless of whether that's an analog or digital control). An
alternative, as Bob points out, is to use floating point or other high
dynamic range audio path. To my knowledge none of this exists for portable
devices. Using loudness normalization on a portable device produces
distortion for higher DR masters and/or artificially limits maximum
playback level for the lower DR masters.

On Wed, Mar 2, 2016 at 8:05 AM, Bob Katz <bobkatz at digido.com> wrote:

> Well, Apple's OSX is floating point all the way through so it doesn't
> matter one bit (pun intended) whether the player's volume control does the
> manipulation and the system volume does a little manipulation afterward,
> nothing is lost. That is, it would be nice if Apple would then dither the
> signal to 24 bits at the last stage on the way to the fixed point output or
> the internal DAC.
>
> Bob
>
>
>
>
>
> On 3/1/16 5:21 PM, Kevin Gross wrote:
>
> Unfortunately all the ReplayGain implementations I've seen don't actually
> adjust the system volume; They just attenuate or amplify the player's
> output in accordance with the loudness measurement. Sound Check on portable
> devices works the same. There is no happiness for anyone here. With this
> approach, modern loud recordings are attenuated and play at a lower maximum
> volume when normalization is enabled. Amplification on older dynamic
> recordings introduces clipping or limiting at any playback level.
>
> On Tue, Mar 1, 2016 at 2:09 PM, Grimm | Eelco <eelco at grimmaudio.com>
> wrote:
>
>> Hi Kevin,
>>
>> > I agree with what I think Eelco is suggesting; Streaming services
>> should transmit unnormalized audio but include loudness metadata so that it
>> can be normalized by the player.
>>
>> I was referring to stationary media players. They all read replaygain
>> metadata and merge that with their volume control. So in a way they all
>> have implemented norm-l already. Now Tidal offers the loudness data in
>> replaygain format, the Tidal stream can be treated exactly the same as
>> local audio files. If the mobile devices apps would adopt this feature I
>> would certainly be very happy. In fact it could be very easy: there's
>> already a media player under development that plays under iOS, Android,
>> Linux and Mac OS and it can merge volume control data with replaygain data.
>> It plays local files, network files and streams from Spotify, Tidal, Deezer
>> and Qobuz. The only thing it needs for norm-l to work is the original audio
>> and the loudness metadata.
>>
>> But this may be just for the happy few. We need to convince the big
>> players too...
>>
>> Cheers,
>> Eelco
>>
>> >
>> > On Sat, Feb 27, 2016 at 4:06 AM, Grimm | Eelco < <eelco at grimmaudio.com>
>> eelco at grimmaudio.com> wrote:
>> > Hi Kevin,
>> >
>> > > Things do get complicated when multiple loudness targets are in play.
>> The easiest thing to do is to produce to the highest target. Our whitepaper
>> recommends improving players. We should stay on that message and resist
>> efforts to put even more targets in play.
>> >
>> > Yes, so -16 (to -20) LUFS is our target. The fact that Tidal offers the
>> metadata to 3rd party integrators is however something different. In all
>> media players it is standard that the replaygain level is merged with the
>> volume control level and forwarded to for instance ALSA sound control in a
>> Linux system. The fact that Tidal now offers the replaygain values
>> alongside their audio means that streaming audio can be treated exactly the
>> same like locally stored audio. How logical can it be?
>> >
>> > Cheers,
>> > Eelco
>> >
>> > > On Feb 26, 2016 9:16 AM, "Thomas Lund" <thomas at lund.one>
>> <thomas at lund.one> wrote:
>> > > Hi Eelco,
>> > >
>> > > The catch is to get metadata safely out of the system alongside
>> audio. It’s not as attractive to use an upstream NormL gain control and
>> then (analog) interfacing to the amp and speaker.
>> > >
>> > > I agree with your 2+3. In case streamers use a fixed target, we have
>> to convince them that one-size-fits-all doesn’t cut it. There needs to be a
>> target setting at -16 LUFS or lower. To me, that’s more important than if
>> we get the whole NormL enchilada from day one.
>> > >
>> > > cheers,
>> > > Thomas
>> > >
>> > >
>> > > > On 26/02/2016, at 09.34, Grimm | Eelco <eelco at grimmaudio.com>
>> wrote:
>> > > >
>> > > > Hi Thomas,
>> > > >
>> > > >> BTW, I’m promoting the NormL principle everywhere, but the ideal
>> is of course to have access to a late gain control in a *system*,
>> especially if interfacing from the streaming device is analog. Hopefully
>> Apple has listened...
>> > > >
>> > > > Well, Tidal did so for sure. If a company like Roon Labs has the
>> replaygain data, any Tidal integrator has and they are free to implement
>> Norm-l.
>> > > >
>> > > > In the end that is perhaps the largest step that Tidal set:
>> offering the raw loudness data to app builders that integrate Tidal
>> support, in stead of rendering the loudness data directly in their API (at
>> their own default target level).
>> > > >
>> > > > I will check with someone who is deep into this streamer code stuff
>> how the loudness metadata rendering in Spotify, Deezer, Qobuz etc is done.
>> > > >
>> > > > So it all boils down to three questions for streamers:
>> > > > 1. Do you offer the raw loudness metadata + raw audio data to
>> downstream services via your API?
>> > > > 2. In your own apps: what is the internal target level?
>> > > > 3. If the loudness level of a track is lower than that target
>> level, what happens?
>> > > >
>> > > > I will try to find the answer to question 1. Questions 2 and 3 will
>> be answered soon when we finished our research project for the TV show. I
>> will keep you updated.
>> > > >
>> > > > All the best,
>> > > > Eelco
>> > > >
>> > > > PS Florian: will there be a loudness panel again at AES Paris? I
>> would be happy to present the results of the research project...
>> > > > _______________________________________________
>> > > > MLA mailing list
>> > > > MLA at grimmaudio.nl
>> > > > http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>> > >
>> > >
>> > > _______________________________________________
>> > > MLA mailing list
>> > > MLA at grimmaudio.nl
>> > > http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>> > > _______________________________________________
>> > > MLA mailing list
>> > > MLA at grimmaudio.nl
>> > > http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>> >
>> >
>> > _______________________________________________
>> > MLA mailing list
>> > MLA at grimmaudio.nl
>> > http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>> >
>> > _______________________________________________
>> > MLA mailing list
>> > MLA at grimmaudio.nl
>> > http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>>
>>
>> _______________________________________________
>> MLA mailing list
>> MLA at grimmaudio.nl
>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>>
>
>
>
> _______________________________________________
> MLA mailing listMLA at grimmaudio.nlhttp://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl
>
>
> --
>
>
> If you want good sound on your album, come to
> Bob Katz 407-831-0233 DIGITAL DOMAIN MASTERING STUDIO
> Author: *Mastering Audio              *Digital Domain Website <http://www.digido.com/>
>
> No trees were killed in the sending of this message. However a large number
> of electrons were terribly inconvenienced.*No more Plaxo, Linked-In, or any of the other time-suckers.
> Please contact me by regular email. Yes, we have a facebook page and a You-Tube site!*
>
>
> _______________________________________________
> 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/20160302/ff221729/attachment-0001.html>