What target loudness are you talking about? If you are referring to the target values that streaming platforms are using, this is achieved automatically. Your song will end up at their target loudness regardless of whatever loudness you set at the mixing or mastering stage.
The only thing you need when mixing is to avoid unintentional clipping and too much dynamic range reduction that would make the mastering engineer’s job impossible or very hard. But if your gain staging is correct in the first place that shouldn’t be a problem. Unless you mess up your mix with a badly set master limiter of course.
The one I set. If you’re mixing a song that is part of an album, each track needs a similar loudness, otherwise it will be a poor experience for the listener.
Correct, but not everything is loudness normalised. CDs for instance.
This is a home recording forum; very few members send their mixes to mastering engineers. They either master the recording themselves, or ‘pretend’ master on the fly, in which case you need a limiter to attenuate loudness.
Personally I think ‘pretend mastering’ is an unnecessarily derogatory term. Mastering on the fly is a skill in itself, and the one huge bonus is that whatever you hear coming out of the monitors is the actual finished product - no nasty surprises after the separate stage of mastering.
Ah, ok I’ve gone back up through the thread. I think I hadn’t appreciated that the conversation had meandered into a question of loudness-war style sonic destruction, and that you were talking about something other than the original topic of the thread, so I was replying to you as if you were. But having seen your post quoted here, I better understand what’s happening - my apologies.
I don’t believe each track in an album should have a similar loudness, quite the opposite. Unless your album only has songs that are all very similar to one another, but that would be weird and very unusual. Soft/slow songs shouldn’t sound as loud as upbeat or more agressive songs.
Fortunately for us, Spotify has an album normalisation feature that allows the albums to be streamed with these loudness differences as the artist/producer intended. Every mainstream pop/rock album I can think of has loudness changes between songs. I suppose there are music genres where it is more common to have loudness consistency across the album, like maybe the heavier subgenres of metal or EDM, but I don’t listen to these so I wouldn’t know.
There are some people who listen to CDs, vinyl or cassette tapes for sure. But they are such a small proportion of the listeners… it doesn’t make sense to target a tiny fraction of your audience. Besides, to experience a loudness difference between 2 CDs/cassettes/vinyls you have to switch them first, which takes some time anyway. So if you have to adjust the volume after having ejected and inserted the next album, it really isn’t a big deal.
Last I checked, stats were showing that more than 85% of the music that is listened to is normalised.
However, even though the original topic isn’t explicitly about loudness war, it still is about maximizing loudness in a mix, which isn’t radically different…
Anyway, I enjoyed learning about Andrew’s process as it is something I had absolutely no experience with and I love learning new stuff. The technical side of it is interesting to me, it is just the ultimate goal that doesn’t really resonate with me.
We’ll have to agree to disagree. I say it’s the norm. You’re not going to find an operatic aria on a death metal band’s album or vice versa. Even if the tracks are of differing styles, they still need a similar loudness, otherwise the listener has to keep turning the volume up and down. It’s exactly the reason the CALM Act of 2012 was introduced.
Depends on your definition of ‘tiny’. CDs and vinyl combined still have 20% of the market. After all, you can’t sell a download or a stream on a merch table - and you certainly can’t autograph one.
In any case, I don’t specifically target only non-loudness normalised environments, I try to cover all eventualities so that my masters will sound good on any platform.
If you have to turn the volume up and down between tracks while listening to a CD, do you think that is a big deal?
I rarely listen to music for recreational purposes, but my band’s sales are 90% physical copies, mostly CDs.
Now some questions for you:
Do you think it’s OK to mix 10 album tracks without paying any attention to their individual loudness?
Do you think it’s good practice to leave loudness to chance in the hope that streaming services will normalise your work for you?
If you are working for a client, what do you think the artist is going to say when they receive their recordings and play them back on Windows Media Player or any other non-loudness normalised playback environment, and they are all separate levels of loudness?
Well I’ll have to insist
I have been studying loudness and normalisation for the past 2 years, analyzing hundreds of songs played from Spotify and Tidal in different genres. It is easy to check for yourself: just do it if you don’t believe me.
Sorry but no, it isn’t. The CALM act was specifically directed at TV broadcasters and targets the loudness differences between commercials and other programmes. It doesn’t apply to audio streaming services and has nothing to do with loudness differences between the songs in an album.
There certainly is a big vinyl comeback indeed. But I very much doubt it is ever going to catch up with streaming. RIAA says physical sales are 7% of the market in 2020 (source).
On that we can definitely agree. One master to rule them all!
Why would you? CDs are not normalised so each track is played back at its intended loudness.
Absolutely not. Have I said something that would make you think otherwise? All I am saying is that not all songs in an album have the same structure, dynamic content, emotional content, and therefore ideal loudness. Again, don’t take my word for granted: do verify this yourself.
Good or bad, it’s what you have to make do with. We have no influence on the loudness targets that they use. However, we do have an influence on the relative levels between several tracks in an album. That is why Spotify has a specific feature that turns the normalisation off when you are listening to an album, so that the original loudness differences from one track to the next are respected as the artist intended. (source)
Well I do this all day long so I can safely say that provided you are setting the loudness according to your client’s vision and your professional input, all is fine.
I do master for CD every once in a while. I did one album 2 weeks ago. There are other things you need to pay attention to as well, for which we have lost the habit, like setting the right amount of silence between two tracks for instance.
By the way, Bob Katz was talking about this very feature yesterday night. Here’s an extract from the livestream:
Is it bad that I actually prefer the new version better and I liked the old one too.
Nice work Andrew. Well polished and clean without too much shredding on the track!
Just to clarify for a teetering old soul like me, do you use the Omni Channel during the tracking phase to control it on the way in, or is it a step in the mixing process? I guess you could render the track with the effect and bring it back in so you could then tweak things other than the transients with less CPU usage.
Funny coincidence: this showed up in my YouTube feed this morning:
He doesn’t go into as much detail as Andrew though. And you could argue that his comparison is biased because he is not using a great limiter, you could probably get close to the clipping result with a good limiter that lets you control the transients (or with a transient shaper in addition to the limiter). But anyway, he makes the same point in the end, and it is pretty self explanatory.
And by the way, he features @bozmillar (Little Clipper) too.
Clipping drums was a trick that took me way too long to figure out. I didn’t watch the whole video, but I’ve found that when done right, clipping drums makes compressors downstream sound way better, and keeps everything sounding more even.
By the way @AJ113, about loudness consistency across an album (or lack thereof), here is an example of a classic rock album that I have recorded and measured. I have others too, but for this one I have a picture ready so here it is:
I’ve ended up the defacto mastering chap a few times on album projects and yes, consistency’s the big thing. It’s something I’ve had to address a couple of times with professional mastering engineers on projects I’ve produced - where they’ve prioritised getting the quiet acoustic track or the sparse ballad up to its loudness potential, but lost sight of what that then takes away from the dense heavy song that follows it which sounds quiet in comparison.
What do you do as a starting point? I tend to use short term lufs and check the loudest part of each track, then adjust from there - so if I decide to aim for -10 lufs (for example) on the densest parts of a track, the one that never takes its foot off the gas might hit -11lufs but a more dynamic one might be -14 integrated. But after that it’s just listening and adjusting like always…
Exactly, that’s the thing: just like you need contrast within a mix, you need contrast within an album to make it interesting, in most cases.
I usually start with the loudest song and adjust the others so that there is some kind of consistency in the lead vocals. That doesn’t necessarily mean that the lead vocals loudness is constant throughout the album but I find that the vocals being the main focal point, usually when the loudness change from one song to the next make sense in the lead vocal, the rest follows through. I pay special attention to the transitions too, and will sometimes use volume automation on intros or outros to make sure that the transitions are pleasing to the ear.
I don’t use a loudness target when mastering, but I do measure the loudness. I am a dynamics lover so I usually end up somewhere between -14 and -11. Occasionally higher on some projects when the clients is after a compressed sound rather than just loud for the sake of begin loud.
and I liked the old one too.
