what you choose as a MASTER REFERENCE may make or break you!

[andygabrys]

Hi all,

Something occurred to me recently that I think is pretty important.

We often choose a commercial production as a reference for our productions, something in a similar style, maybe an a la for a listing.

Have you ever listened closely to the same song on both YouTube and via CD or the iTunes or Amazon music stores?

a good example would be Katy Perry's "Unconditionally"

here on her Vevo official channel https://www.youtube.com/watch?v=XjwZAa2EjKA

and on Spotify http://open.spotify.com/track/4fwbGKNExPtPHbor1TBSY4

and here on the iTunes store https://itunes.apple.com/us/album/uncon ... =690928457

What do you notice?

It might take a minute, but listen to 20 seconds of one, and then 20 seconds of the other (choose the chorus please)

if you have meters on your audio system / interface check out where each production is peaking. I have an Apogee interface that has a meter on it and its easy to see whether or not a particular production is peaking just below digital 0.0 dB versus much lower?

All productions on YouTube, Spotify (free version here), and other "free" streaming services are artificially pulled down in volume. So while its the same production, they are peaking maybe 3 dB lower in volume. I don't think these services "limit" the volume of the music by adding additional processing (as with most contemporary pop productions that are already way LOUD would end up distorted and unlistenable if they did that) but rather when the tunes are placed on these services, they just pull down the volume several dB.

Contrast with the actual CD, or the version on the iTunes store. They sound more lively, a bit brighter (likely just apparent brightness due to more volume) and of course, louder.

Makes sense cause all these free services need to retain some incentive for you to buy the actual music right?

HERE"S the possible IMPACT on your productions and your pitches though:

If you reference against the YouTube version, you might get close on the mix. But you will be fooling yourself when you go to master. You production won't be anywhere close to the same impact and volume as the original song.

This may not make an impact if the style of music you are producing is softer in nature, or if you believe that limiting your productions to the same apparent volume as a commercial track doesn't get you any better forward results.

But if you produce Hip-Hop, EDM, trailer music etc. it would be worthwhile to check out the full volume version of the song either on CD or via the iTunes store (yes, buy the ref song, you can write it off in most cases as a business expense - however consult your accountant on that - I am not a tax professional).

Its also really worthwhile to go the extra mile if the publisher you are writing for wants your production to have the same emotional impact as the ref track.

HTH

[jasonrumley]

I would have never even thought of this but great point. With everything cranked up so loud these days, a few db is crucial. Thank you.

[Kolstad]

I don't really use reference tracks for loudness mastering anymore, I aim to comply with the new loudness standards, to be future safe. The CALM Act has been passed in the US, which demands US broadcasters to comply with the ATSC A/85 standard from now on. So,

iTunes radio use it.
US tv stations use it.
The latest Daft Punk and U2 albums comply with it.
Italy, France, Germany, Switzerland, Austria, Norway and Spain comply with the European equivalent R128 (1db lower)

Bob Katz has declared that the loudness wars are over, and I believe it. In a year or two, everyone will be doing it. No doubt in my mind (anymore).

So, I'm done with loudness matching from random sources for my music, I'm complying with the new standards, and will do so for everything I send out.

Check out Ian Shepards blog http://productionadvice.co.uk/u2-21st-c ... #more-7263

[andygabrys]

good point Magne.

an interesting thing to note is that Ian mentioned DR9 as the standard employed - which is still at least a db more than most people tend to limit in a non-commercial setting. Crest of 10 dB RMS is pretty normal as far as loudness, and is usually achievable without resorting to special mastering techniques (excessive saturation etc.)

Questions to you:

1. Since you have been employing metering to your above referenced standard - what do most of most of your finished masters tend to exhibit as far as peak and RMS volume?
2. Since the standards specify short, medium, and long form program loudness, does a song equate to a short term program?
3. What are the specs that you adhere to for that length of program?
4. What are the corresponding Peak and RMS levels for those who haven't invested in meters that allow time weighted averaging?
5. What type of meters are you using to establish your levels? Are they resident in your DAW or are available as external plugins?

Just wanting more information - thanks!

[Russell Landwehr]

Makes sense cause all these free services need to retain some incentive for you to buy the actual music right?

I never took you for a conspiracy theorist, Andy.

[shorttonpro]

Definitely a good point about referencing. Not just due to level,s but the codecs used, especially in YouTube really degrade the sound. I'm okay with using Spotify or Deezer premium accounts for the purpose of referencing production aspects, but for an actual track to track comparison for mastering, I do prefer to have the actual Wav, Flac, or HQ MP3 file so I can check things out in my meters and spectrum analyzers.

I generally use Voxengo SPAN for spectrum analyzing (FREE!) http://www.voxengo.com/product/span/
and brainworx bx_meter for levels, but there are plenty of options. I like to see peak levels, average RMS, and dynamic range. I'm not necessarily looking at numbers over a set period of time, just what the meters say after a momentary glance during the loudest sections.

I base loudness decisions off the genre of music, intended use, reference, and in most cases - client preference.

For me, I'm okay with getting a high level for Pop or Rock music if that's what the reference calls for, because it will be compared to this. We're talking -6dB RMS, 4dB of range, and -0.1 max peak. Other genres are more dynamic, and hip-hop's low-end benefits from more (-8dB RMS) or so.

I'll note that when creating in these louder genres, I try to make contrast between sections in other forms, since dynamic contrast is only a smaller part of the equation (think width, instrumentation, frequencies etc...)

If it is for licensing only, then I try to keep things around -10dB RMS which is a pretty good compromise of loudness and dynamics.

These of course aren't standards but just my personal tendencies, though things are starting to move in a more dynamic direction.

[TimWalter]

Wow wow wow. I learned a lot from this thread. I will study to it more later. Thanks Andy for the OP and others for adding info. Tim

[Kolstad]

good point Magne.

an interesting thing to note is that Ian mentioned DR9 as the standard employed - which is still at least a db more than most people tend to limit in a non-commercial setting. Crest of 10 dB RMS is pretty normal as far as loudness, and is usually achievable without resorting to special mastering techniques (excessive saturation etc.)

Questions to you:

1. Since you have been employing metering to your above referenced standard - what do most of most of your finished masters tend to exhibit as far as peak and RMS volume?
2. Since the standards specify short, medium, and long form program loudness, does a song equate to a short term program?
3. What are the specs that you adhere to for that length of program?
4. What are the corresponding Peak and RMS levels for those who haven't invested in meters that allow time weighted averaging?
5. What type of meters are you using to establish your levels? Are they resident in your DAW or are available as external plugins?

Just wanting more information - thanks!

You really need to read more of Ian Shepards blog posts and videos on mastering, Andy. He's the expert. This is the "how" I try to follow http://productionadvice.co.uk/lufs-dbfs-rms/ According to this the RMS levels should be around -11,5, but mine is up around 14 for the test sine (on the Toneboosters plug-in), so I haven't pinned down the basics just yet (remember I'm a songwriter).

These meters can be expensive, but I use the Toneboosters plug-in (very reasonable), http://www.toneboosters.com/tb-ebuloudness/
Ozone 6 advanced also has the meters, I think (but I only have the std version). Ian has developed his own plug-in, which I haven't tried, but may want to this year. There are others, but I'd think more plug-ins will come for this.

I'm still working on getting this right in my workflow, but I want to get to a point where I can say that my music easily can adhere to the std. Not all there yet, but working on it. I tend to mix my material too loud, and my masters was getting out of hand. What will happen when Radio and TV broadcast the music is that they will turn it down according to the standards, which means it makes more sense not to push the mix/master so much and keep more dynamic range. Otherwise your music may sound flat and lifeless when compared to more dynamic tracks at these broadcast levels.

I just like the idea of referencing a standard, and be done with it. That may not work for all, and at all times, but as a starting point it can simplify much work, I believe.

You may still be right about choosing the right references when it comes to delivery time with specific clients, and even when following these standards, choosing the right reference sources may mean the world for getting the mix right. Good stuff, Andy.

[Len911]

I'm not sure about pulling down the volume. I think mostly that stuff is in the hands of the uploader. If the uploader listens after they've uploaded and it's distorted, I think they would delete it and start over??

Then there's the Sonnox Pro-Codec, it must count for something??

I believe all streaming uses compressed formats, like mp3, mp4, flash, not wav nor aiff.

[playagibson]

I started to use tracks that were mastered for iTunes, by bands that I know get the best mastering engineers.
I a/b to that and it's made a huge difference.
I wish I did that years ago.
It's a never ending learning curve.
Rich