Article on Mastering for Streaming

[KevWind]

I found this kinda interesting

https://www.pro-tools-expert.com/pro...%20-%208290291

[Bob Womack]

Hi, Kev!

It was an interesting, well-researched article! There are three considerations that he doesn't really deal with in depth:

Platform processing: What is always lurking in the background when we are considering mastering standards is, "What is the platform going to do to my file once I deliver it?" Having gotten my start in broadcasting, my thought process was always "throughput." That is, "How can I best cause the end product delivered to the consumer to sound closest to my my own mix?" We have always mixed with the downstream processing in mind. In broadcast, that meant that if our product was too dynamic and dropped too low, the "Volumax" or similar broadcast audio processing would bring up the quiet passages, and with them the noise. If we were too loud, the Volumax or similar system would crush the loudest passages. Today, the whole streaming market is forced to think in those terms. In essence, we have to out-think the processing to keep the product from being manipulated to sound a way we didn't intend.

Adjacent selection level: If it is important to you that the listener not perceive that your song is quieter and as a result, falsely assume it is possibly lower quality, then you've got to master to a density point that will make it show up as loud as the surrounding material.

The middle man: I routinely master for multiple clients and multiple platforms but one thing is constant: my clients have a hard time delivering multiple to multiple platforms. They manage to screw up which goes to whom despite careful labeling. As a result, over time I've narrowed my output formats down to three categories of output: Broadcast, Streaming, CD. Because of the CALM act, the FCC has broadcasters align their processing to deliver -24lufs. If I deliver -24lufs and carefully manage peaks, their gear doesn't alter my work. In streaming, we've pragmatically discovered that when we master to -14lufs with managed peaks, YouTube doesn't muck with our files and they match up to adjacent files well. The other streaming platforms seem to play well with -14lufs as well. And then, when mastering to CD, and I still do, I typically master to -12lufs so that the end user can slap up one of my CDs after a Don Henley disc and not have to dive for the volume control. So, my product goes out with broadcast and stems in -24lufs, streaming goes out at -14lufs, and CD goes out at -12lufs.

And the clients still call and ask to be reminded which goes where.

Oh, and I use iZotope Insight as my reference loudness monitor.

Bob

[KevWind]

Makes sense Bob thanks ..My self I am basically mastering for YouTube videos
So shooting for -14 lufs.

Another thing I read (somewhere) was (especially if shooting in 4K video ) is to always upload but select to " Schedule" the publishing (the default is 12:00 AM the following day) The article said this gives Youtube more time to process and you end up with the best processing codec vp 0.9 ...

I have Nugen Audio 'Master Check" as my meter

[jim1960]

Relatedly, last September the Audio Engineering Society released TD1008, an updated set of guidelines for streaming loudness.

This table is from the document and shows the recommendations made:


These aren't recommendations to mixing or mastering engineers, but rather to streaming services and online radio stations. Whether those platforms adopt the recommendations is ultimately up to them but Apple, who was doing its own thing to a greater extent than most other services, has, earlier this year, adopted the recommended guidelines (although it only applies to the newest OS and iOS).

Apple Switches to LUFS, Enables Sound Check by Default
In a nutshell, Apple was the last major streaming service not using normalization by default, and not using LUFS. But that has changed recently, meaning that the loudest songs will now be reduced in level by default on YouTube, Spotify, TIDAL, Amazon, and Apple Music, on new devices.

And perhaps more importantly, the results between these services will be much more consistent, because they are all using LUFS. So rather than the situation only a few years ago where there were multiple Distribution Loudness levels and methods of measuring loudness in use, things today are much simpler. All these services now use LUFS, and they’re all normalizing to -14 LUFS, apart from Apple, who are following the AES TD1008 guidelines and normalizing music to -16 LUFS.

So we're not quite at uniformity but we're a lot closer than we were a few years back when the range was as wide as -12 to -18 (if I'm remembering correctly).

The article below might be of some interest, also. It's an interview with Bob Katz and Rob Byers. Katz is a well-known mastering engineer. I'm not familiar with Rob Byers but it appears his expertise is on the broadcast end of the spectrum. Both are listed among the authors of AES TD1008.

Loudness Standards For Music Discussed By Experts

[FrankHudson]

I too thought it a good article as far as it goes. I appreciated the info on the LUFS standards for various places.

Like Bob Womack I wished it'd gone into a cookbook suggestion for different services, though the info was there to help you "season to taste"

I deliver my streaming files through a distributor. I've noticed (particularly with acoustic guitar pieces) that Apple seemed to sound like it had more compression when I listen to my files there. Since I can't pick and choose (my distributor sends out one file to all) I've started aiming for 15 LUFS or maybe even 15.5 on solo acoustic guitar music, instead of 14.

[DupleMeter]

It’s frustrating that we don’t have a single streaming standard for music. It’s would be nice if we could get a single standard to adhere to. Heck, Spotify can’t even make up it’s own mind. They have 2 standards; a “normal” & “loud” standard.

I will warn that blindly trying to hit a loudness spec is not always the best way forward. Overdoing the limiting can really make a great recording sound claustrophobic.

IME having a well tuned & calibrated mixing environment will yield much more consistent mixes that are consistently hitting the loudness spec without having to go through gymnastics in the mastering phase.

[shufflebeat]

Ian Shepherd's podcast is quite listenable and, even though there have bee recent changes which he updates for in later episodes, his explanations of the background tech are well informed and useful to normal folks.

https://themasteringshow.com/

#24 is "mastering for online streaming".

#36 might also be an eye-opener for some.

[KevWind]

Originally Posted by shufflebeat

Ian Shepherd's podcast is quite listenable and, even though there have bee recent changes which he updates for in later episodes, his explanations of the background tech are well informed and useful to normal folks.

https://themasteringshow.com/

#24 is "mastering for online streaming".

#36 might also be an eye-opener for some.

humm a Bit dated. And don't know if he corrected in later podcast but for example Youtube now recommends 24 bit not 16