I agree with the idea that you can never 'master' your own work, not in the truest sense. A master, I guess, must have involved the fresh ears and perspective of another person.

That said, many of us employ something of a 'mastering style' process as we are either forced to financially, or are control freaks, or are impatient and want to stick our new club banger on Soundcloud ASAP...whatever the reason, it happens.

What's a ballpark readout for a 'competitively' mixed club track these days and what's available in 2020 to get you there without introducing unwanted artefacts?

Thanks folks!

I think I may have used RMS erroneously? What I mean to say, how LOUD are you making your mixes (or perceived loudness) and what are you using to measure this/as a reference point?

In my research I found Howard Massey saying this about LUFS levels:

"Worse yet, pretty much every delivery service normalizes the audio files they stream – a process that can easily degrade sonic quality. What’s more, they all normalize to different levels. Spotify, Tidal, and YouTube, for example, set a ceiling of -14 LUFS. Apple’s Soundcheck, an option in iTunes that goes through your library of music and analyzes the average volume of all the songs, can actually tell the player to turn down by as much as -16 LUFS. And most commercial television broadcasts drop them down all the way to a whopping -23 or -24 LUFS, depending upon the country.

Mastering to one of these target levels is up to you, and arguably not necessary, but bear in mind that the music you’re working with will be adjusted to one of these levels (or others) at some point after you release it, whether you like it or not. A good compromise – and pretty much a consensus these days – is to use -16 LUFS (or -6 dBTP [True Peak] if your loudness meter offers that unit of measurement, as the WLM Plus meter does) as the target for integrated loudness."

so..are you guys 'self mastering' mastering with streaming services normalisation algorithm in mind?

Good thread here explaining why most of the speculation on the internet about mastering to lower levels to “optimize” for these streaming services isn’t really what most ME’s are doing and isn’t really a good idea. Targeting Mastering Loudness for Streaming (LUFS, Spotify, YouTube)-Why NOT to do it. You could take it further and say dont worry about RMS / LUFS in general and just use your ears but it can definitely be useful to check some reference tracks in the same genre and measure the RMS / LUFS of those to get a sense of what a reasonable target is. If your references are modern dance music I’ll bet it’s louder than -14 on LUFS or RMS

Thanks for your reply!

Little slow round here, is this sub forum very active?

I guess it makes sense to create a master that will sound best on the intended broadcast system , which for us I guess is...club speakers?

I'll have to grab a couple of recent 'club tracks' (whatever that means..) and check levels for comparison.

I guess its crackers to pander to some streaming services aggressive normalisation processes, I guess they are going to get their grubby mitts on it regardless..

For my style of house I usually aim for like -8 or -8.5 LUFs. Generally it seems like in dance music the slicker more commercial stuff is absurdly loud (like -6 LUFs and higher), and things get more dynamic the further underground you go.

So the loudness you aim for will definitely depend on the genre and kind of DJs you'd like to play your tracks.

-6! That seems absurd.

Why push things so loud when services/broadcast systems will normalize fairly aggressively and all mixers come with channel trims? Seems bonkers to me!?

Watching this thread as I know this is a weak area of mine. I cannot afford to get stuff mastered externally but am definitely interested in this subject

Play some stuff you really like and check what levels they come in at.

Bare in mind well mixed tricks sound full and loud when they aren’t always. I’ve also heard stuff that sounds amazing and the levels are off the chart loud.

Really depends on genre as mentioned above. I’d really not go down that rabbit hole and just find a level that sounds good until you break it for self mastering. When you release stuff or have the money then go to a good mastering engineer but be prepared to realise a load of it is in the mix itself and they should just sprinkle unicorn tears on it with a touch of magic ears.

Also ozone 8 or 9is really good for testing.

Use the assistant and choose the cd, highest setting and then study what it did to the eq levels and dynamics to get it that loud, then go back to your mix and adjust then try again and again till the assistant does minimal changes and it sounds good.

It will give you a good idea of areas of your mix you need to focus on.

You can also use the assistant and track match something you like and it will adjust your track to match (bit hit and miss) This will also give you some pointers on areas to focus on for the genre you like.

If I was going to self master, I’d probably mix differently. For example, I mix my kick level to -3 VU, which is roughly a peak of -12 in the DAW. Leaves plenty of headroom for the Mastering engineer.

I don’t have the level of gear of the ME, the room or the ability to hear the song fresh again. So shooting to get it louder, I’d probably over squeeze it without realizing it.

To take that out of the equation, I’d probably mix my kick to -6db and balance everything from there. Not much headroom in that case and less chance of me ruining the mix with my hack job mastering.

This is an interesting perspective as it relates to my process. I keep my peaks at never passing -18dbfs by force of habit and to allow for dynamic range. In the past my tracks have been mastered by a third party.

If I try and make up that much gain, having not heard the track fresh eared like the ME, you think I'll inevitably punish the track more than they do?

If you mix in the DAW, what's the difference between setting kick peak at -12, then bringing whole lot up to -6 before your pre master render over doing doing it 'in mix'?

Other than the psychological effect of 'no fresh ears' , in digital, headroom is headroom, right?

I agree that Ozone's matching functions can be useful. I have used Ozone in a more primitive sense in the past by finding a mastering preset that, when applied, best matches a reference track, and flicking it on/off periodically during mixing to A/B the two and find a happy middle.

Well i mix through a compressor and saturator so they would react a lot differently. If it was me id relevel the mix track by track. You are asking questions that you can figure out just by doing the work and seeing how it turns out for you. No real way for me to answer.

Im the same, i mix via with a tegler creme always on main buss and that passes into the 2 channels on my Audient asp800 with hmx and iron on lightly adding some saturation and back into an apogee sympthony |(on the hunt for something really good to add saturation on the 2 buss, thinking overstayer possibly). I tend to have everything at no more than -12-10 in my daw then make it up to -6 pushing the creme output.

Then I use Ozone as above to get a feel for how it sound mastered, smash it to bits, pull it back etc etc then go back to the mix and amend where required. all my latest stuff has been mastered properly but now and again for a quick fire remix or something short for other purposes I have self mastered but try and stay away from that. I do however send a self master to mastering engineer for critique and advice as love that side of things too.

sorry, I understand that if you have processors on your mixbuss (I never do) they would react differently to different gain staging.

I guess I was wanting to explore if there is any difference between pulling individual tracks down to pulling the mixbuss fader down when approaching a pre master render. Never mind, don't stress.

OK cool, so when you self master - what's your target for RMS and peak loudness? What style of dance music are you making?

Electronic music, sometimes a bit 4/4 housey but more often more broken down and chilled but always with quite deep lows.

From roughly -8.5 -7 LUFS maybe peaking into the 6 realm on a spike.

2 recent examples just for giggles

Self Mastered


Mastered by Mike Marsh


Bare in mind that's on Scloud though but you can see the levels.

hi

thanks for these, i only just got round to listening.

so -7.5 pushing -6 RMS seems to be where people are ending up. Ill aim for that and then A/B/ with commercial releases for comparison

I'm into melodic techno and I really appreciate there's a better "resistance" to crazy loudness in that genre, compared to many other electronic dance music genres. Melodic techno tracks are allowed to breathe more, as it's obviously part of the genre sound.

The average loudness of a collection of 46 techno tracks I've measured, mostly melodic and released within the last 5 years, is -9,7 LUFS. Removing 5 more intense Drumcode tracks expectedly lowered the average volume slightly to -10,0 LUFS.

I myself usually target around -11 LUFS. It gives the music room to breathe and I lose less of my dynamics work when streaming services turn the volume down less (compared to mixing louder than -11) - and when listeners subsequently compensate by raising the volume control. But psst, don't tell the guys in that other loudness thread, they'll give me hell...

I decided to not selfmaster anymore, and no force in this world is going to change that. I would rather spend 70 EUR per track and ask an engineer to do it in a few hours if it is that urgent. But this headache and struggle I cannot stand anymore. I have a friend who is a ME which helps ofcourse.

You do all the fine mixing only to destroy it in the mastering process? No way.

By the way I mix to hit 0db on my masterbus without using any 2bus processing. I could bring the raw mix into the club (if they would exist at the moment).

You mix to hit 0db? And your mastering engineer friend is happy with this? That doesn't leave him much room to play with!

Still, if its working for you - more power to ya!

I’m not sure why anyone would mix to hit 0 and then not get why mastering is miserable for them..lol.

I’d say my entire tracks peak at -6 or so before I get them mastered. Every Mastering engineer I have used compliments the mixes and thanks me for the headroom - and I have used some snobs.

He was indeed very sceptical until he have got the material to work with. Technicaly there is no difference between a premaster that peaks -12db or a premaster that peaks -0.3 as long as there are no overpeaks, where I put significant production focus into avoiding this.

I use to achieve this without much compression, so the ME has a lot of dynamic range to work with.

I have found out that my mixes became to a major degree better because it forces me to do almost everything better to achieve a balanced 0db mix: gainstaging, eq, recordings, arrangement, seperation of frequencies, reducing dynamic without compression, filling into the holes. You name it.

Fair enough.

I mix with peaks hitting -18 to -12 - I find that 'breathing room' (even if its only perceived) gets my mixes breathing more and allows for finer control of saturation/perceived loudness of 'the loud bits'.

I guess at the end of the process i could drag my master fader up and have my mixes pushing 0db like yours - just find I wrestle things less if I dont.

Indeed man this is a dance on the rope to mix this kind of pushed to the limits, and I would also say it is a very bad advice to mix into 0db, because it will be misunderstood in 9/10 cases.

But it makes my gainstaging decisions so well. It forces you to work "correctly". You have to pull down the faders on most channels, because it will clip with even the slightest overpeak would show up otherwise. When mixing with a lot more headroom you have more room to stack identical frequency content. But the result of this will be that the comp / limiter has to work a lot more on the mastering-stage, which will end it in undesired / unintentional directions. I would say at least try it out. A well balanced mix can sound very similar to a finished master, at least very close. without any mixbus processing.