[MLA] from the trenches...

Thu Jan 18 12:19:23 CET 2018

Hi Kevin, 

You’re right, I did miss the part where you specified SPL. 

My instinct is that a lot of extra tech work would be needed to implement this, and that since we can’t even get Apple to use the LUFS meta-data they already have to achieve better results than Sound Check, the effort involved in trying to get this up and running probably wouldn’t be worth the effort, especially since it would be so open to gaming. I could be wrong, though.

Ian

> On 16 Jan 2018, at 15:16, Kevin Gross <kevin.gross at avanw.com> wrote:
> 
> You may have missed the part where suggested playback level is specified in SPL. Clipping point would depend on the reproduction system so there would not be an obvious loud-as-possible setting. This might cause producers to actually think about how they really want the material presented and maybe even pull out an SPL meter when auditioning it.
> 
> I do agree that it would likely to be gamed. But, if we can transform the game from who can squash the audio the hardest to who has the balls to publish the most unreasonable suggested playback level, we will have improved things for listeners and possibly given producers the break from loudness normalization they're asking for.
> 
> Kevin Gross - AVA Networks
> 
> On Tue, Jan 16, 2018 at 3:31 AM, Ian Shepherd <ian at mastering-media.co.uk <mailto:ian at mastering-media.co.uk>> wrote:
> 
> Hi Kevin,
> 
> Interesting brainstorming, but I think the outcome is inevitable. As soon as people understand how the feature works, they’ll ask us to dial in the maximum level allowed without clipping. And as soon as one artist reaches this point, everyone will want to compete with that “benchmark", and once again we’re slamming things up against zero. Unless I’m missing something...
> 
> Ian
> 
> 
> 
> 
>> On 15 Jan 2018, at 17:27, Kevin Gross <kevin.gross at avanw.com <mailto:kevin.gross at avanw.com>> wrote:
>> 
>> Bob has presented a problem and I'm just doing some fantasy brainstorming. Try to go with it for a little bit. Sometimes this leads interesting places.
>> 
>> Yes, "suggested playback level" would counteract normalization - if enabled by the listener.
>> 
>> A player supporting this would have three normalization options:
>> Loudness normalize
>> Original level
>> Suggested level
>> Loudness normalize would be the default. Original level is simple 0 dB playback. Suggested level plays pack with a gain indicated by metadata provided by the content producer.
>> 
>> I agree that this would likely create a shitshow but it could also be an educational exercise for producers and artists. Bob could ask clients, "How much louder do you want it?" And then he simply dials in that number and, without crushing, can beat anything the mix engineer has done. If we can move the loudness war to metadata, we can save the audio.
>> 
>> Would any listener ever choose Suggested level playback? The feature is marketable as, "listening to music as the artist intended." This is a fundamental selling point for high resolution audio which does have some market traction. If, somehow, the feature did not result in a shitshow, it might provide a usable listener experience and, for instance, solve Bob Katz's "acoustic advantage" problem. If it turns into a shitshow, no one will use it - this will be the educational part of the exercise.
>> 
>> Is it possible to implement this feature? If listeners had a calibrated playback system, Suggested level could be specified in SPL. If a track specified a playback level higher than the equipment was able to safely produce, a warning could be issued and a more reasonable level used. Users may still want a volume override to nudge playback a specified amount above or below the Suggested level. Calibrated playback is already not difficult for (factory) earbuds and home theater systems (calibrated using an AVR with a supplied measurement mic). It also should be possible to get it done for other constrained listening scenarios - laptop speakers, bluetooth speakers, automotive.
>> 
>> 
>> Kevin Gross - AVA Networks
>> 
>> On Fri, Jan 12, 2018 at 1:55 PM, Grimm | Eelco <eelco at grimmaudio.com <mailto:eelco at grimmaudio.com>> wrote:
>> Hi Kevin,
>> 
>>> If they want actual loud, maybe we can work on adding "suggested playback level" metadata to recordings.
>> 
>> ? What do you mean exactly? The way I read it would be counteracting the normalization. But on the other hand everybody and their mother would turn that button to '11', so if it is turned fully up for all songs, we can just as well leave it out :-)
>> 
>> Cheers,
>> Eelco
>> 
>>> I don't expect this would actually work and it would likely get ridiculous quickly but it might also be educational and it might relieve some pressure to crush.
>>> 
>>> Kevin Gross - AVA Networks
>>> 
>>> On Tue, Jan 9, 2018 at 7:32 PM, Bob Ludwig <gatewaybob at mac.com <mailto:gatewaybob at mac.com>> wrote:
>>> Thanks for the encouraging words, Eelco!
>>> 
>>> Best, 
>>> Bob L
>>> 
>>> On Jan 9, 2018, at 7:27 PM, Grimm | Eelco <eelco at grimmaudio.com <mailto:eelco at grimmaudio.com>> wrote:
>>> 
>>> Hi Bob,
>>> 
>>> I very much pity you. But also your artist.
>>> 
>>> From another viewpoint: at least he is aware that Spotify applies normalization. He says it sucks but he cannot do anything about it. I always said that normalization will not stop people from crushing their masters. The main goal is to offer a plain playing field so people who don't crush their masters will sound equallty loud (or louder). The artists who really love the crushed sound with 'the better brightness' can just keep on demolishing their art - and that's fine with me. 
>>> 
>>> But to be their mastering engineer must make you seriously ill, I am sure.
>>> 
>>> The good news is that in the indie world many people are picking it up already. At my school all the current students want to learn about making dynamic mixes and masters at the moment, and a good friend of mine (a talented mix & mastering engineer) never has to master above -13 LUFS for about a year now because of Spotify...
>>> 
>>> Take care,
>>> Eelco
>>> 
>>> 
>>> On 9 jan. 2018, at 22:27, Robert Ludwig <gatewaybob at mac.com <mailto:gatewaybob at mac.com>> wrote:
>>> 
>>>> Hi everyone,
>>>> 
>>>> As is so often the case now, I’m mastering an album where the mix engineer has premastered it for their client’s “listening references”  SO loudly, and the client expects me to at least match it.
>>>> 
>>>> By far, for me the loudness wars are the worst they have ever been and now it isn’t so much the artists or A&R, it’s the mixers themselves.
>>>> 
>>>> I’m doing this new album, the engineer sent me his crushed “listening refs” and his “less compressed mixes”.  
>>>> In addition to matching his level (with hopefully a more clear sound), I also did one with some dynamics, about -1dB less compressed.  These were labeled “even louder” and “loud” !!  
>>>> 
>>>> Here is a comment I just got from the artist:
>>>> 
>>>> "Between ‘skip rate data’ and being ‘loudness normalized’, Spotify really needs to give everyone a break. Jeez louise, soooo laaaaaaame!!!!!!!
>>>>  
>>>> I don’t actually know what to suggest. Especially now I know that whatever loudness I want my music to be, Spotify are just going to turn my shit down. That sucks. I still have Bob’s “louder and even louder” mixes and I’ve always gone straight to “Even louder”, feels obvious, no? This is probably part of the reason why I felt what he sent most recently was that bit less impactful. To cut to the chase, I don’t want this music to exist at anything other than Bob’s “even louder” level. That gives it the better brightness too, for me.”
>>>> 
>>>> This makes me ill…
>>>> Bob L.
>>>> 
>>>> PS. please don’t quote this any place, it will get back to the artist eventually.
>>>>  
>>>> _______________________________________________
>>>> MLA mailing list
>>>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>>>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
>>> 
>>> _______________________________________________
>>> MLA mailing list
>>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
>>> 
>>> _______________________________________________
>>> MLA mailing list
>>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
>>> 
>>> 
>>> _______________________________________________
>>> MLA mailing list
>>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
>> 
>> 
>> _______________________________________________
>> MLA mailing list
>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
>> 
>> 
>> _______________________________________________
>> MLA mailing list
>> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
>> http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl <http://grimmaudio.nl/mailman/listinfo/mla_grimmaudio.nl>
> 
> 
> _______________________________________________
> MLA mailing list
> MLA at grimmaudio.nl <mailto:MLA at grimmaudio.nl>
> http://grimmaudio.nl/mailman/listinfo/mla_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/20180118/a145d852/attachment-0001.html>