[MLA] Interesting info about Spotify normalization

[Kevin Gross]

Yes, Eelco is right. Lift the soft parts. This does not require crushing
the loud parts.

We're not going to come up with an optimal solution for challenging
environments. We're going to come up with something that improves listener
experience and maintains artistic integrity and is transparent in a
reasonable listening environment. We may fail. I think it may be time to
try though.

The other thing that may be related, or not, is that we know many streaming
services are doing some sort of limiting. It might be good to put together
some technical recommendations to avoid nasty limiter implementation and
tuning. If it's gonna happen, and it's happening, it might as well be done
right.

Kevin Gross - AVA Networks

On Sun, Mar 18, 2018 at 3:11 PM, Grimm | Eelco <eelco at grimmaudio.com> wrote:

> Hi Bob,
>
> Like Kevin suggested between the lines: it may be about time to define the
> ‘noisy environment’ processing. What I read in your message is that the
> micro dynamics should not be changed. That means we still need a target of
> -14 LUFS or lower (with 3 dB extra headroom): the processing is on the
> softer part of the sound.
>
> I think it is time for MLA to again show the way to the future and write a
> white paper about this. Perhaps we better wait with publication until -23
> LUFS targets on smart phones are possible. But then we should have a
> solution ready to help people control the huge dynamic range they got.
> Would you be willing to take the lead in writing this white paper?
>
> Cheers,
> Eelco
>
> Dynamics processing in streaming on mastered material is just another
> backwards step to the ugly Orban, ugly-sounding and ugly radio land. No
> ONE, I mean no ONE knows what's optimal for every venue. NO ENGINEER has
> preknowlege of what the user's venue or user's conditions are going to be.
> And giving the user the choice rarely results in anything good.
>
> Furthermore, presets called "Loud" and "Normal" are just ways of making
> the sound "squashed" and "more squashed". Applying downward compression,
> limiting and Orban style processing presets on top of music which can vary
> from the most unprocessed to the most smashed is just going to further and
> seriously distort the smashed recordings, and ruin the depth and clarity of
> any high quality material. Thus returning to the distortions that we
> successfully got rid of by inventing R-128 in the first place! This is WHY
> the EBU and ATSC do NOT advocate ANY loudness processing, simply loudness
> normalization. Do you think that Spotify knows better than experienced
> broadcast engineers?
>
> There are two proper solutions to the noisy venue problem. The eventual
> and proper one is to move to is a noise-based processor that incorporates
> the microphone in the user device and raises the gain of low level material
> without any level-based time-constants.
> As we wait for a microphone-noise-based processor to be perfected, I would
> propose interim presets labeled "noisy environment" "medium noisey
> environment' and "quiet environment". These would use parallel compression
> which is designed to be invisible at high levels and only effective
> meaningful at low levels. NOT the kind of processing that I wager Spotify
> is using. I wager they're using ugly, undiscriminated, downward compression
> and limiting.
>
> Otherwise vendors and streamers should keep their noses out of the
> processing business. It failed and sucked with radio and they're just bound
> to repeat the mistakes of FM radio :-(
>
>
>
> Let's call a spade a spade and do things right.
>
>
> Bob
>
>
>
> On 3/16/18 12:07 PM, Kevin Gross wrote:
>
> On Fri, Mar 16, 2018 at 3:22 AM, Grimm | Eelco <eelco at grimmaudio.com>
> wrote:
>
>>
>> Let’s hope we can nudge Apple to implement the “extra headroom with
>> normalization on” option soon. Then we just switch to a -23 LUFS target and
>> have a beer :-)
>>
>> Cheers,
>> Eelco
>>
>> We have talked in this group in the past about the potential need for
> different processing for different listening scenarios. Spotify is touching
> on this with their "Loud", "Normal", "Quiet" options. Based on the fact
> that loudness-normalized material is played off a different server, it
> appears they're doing some dynamics processing. We've said that we'd look
> at this once we had solid recommendations for unprocessed loudness
> normalization. Maybe it is time to get started.
>
> Kevin
>
>
> _______________________________________________
> 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 <(407)%20831-0233> 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.
>
> _______________________________________________
> 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
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://grimmaudio.nl/pipermail/mla_grimmaudio.nl/attachments/20180318/5336f47c/attachment-0001.html>