Mastering Essentials Part 3 - How loud should you master ?
Вставка
- Опубліковано 22 вер 2024
- How Loud Should You Master?
How loud is loud enough, and how loud is too loud ? In this third episode of his video series for Sound On Sound, mastering engineer Ian Shepherd describes his simple strategies to help you find the ideal balance of loudness and dynamics for your music, and why it's so important. Is louder always better ?
VU Meter plugin mentioned in this video available from klanghelm.com/...
ABOUT THE SERIES
Mastering Essentials: The SOS Techniques Guide
In this six-part series, produced in association with Steinberg, well-known mastering engineer Ian Shepherd takes us through his approach to mastering. Whether you’re polishing individual tracks or assembling EPs or albums, Ian will explore the tools, skills and decisions that can help you achieve great-sounding results.
1. What Is Mastering?
• Mastering Essentials P...
2. The Three Ms of Mastering
• Mastering Essentials P...
3. How Loud Should You Master?
4. Mastering EQ: Balance, Don’t Match
• Mastering Essentials P...
5. Compression & Limiting In Mastering
• Mastering Essentials P...
6. Final Delivery: Requirements & Specs
• Mastering Essentials P...
More parts coming soon, make sure you subscribe to the SOS channel to be notified.
For more detail on Ian’s approach to mastering, you can get his free downloadable Home Mastering Guide at www.HomeMasteri...
Subscribe!
sosm.ag/subscribe
Visit www.soundonsou...
Like Sound On Sound: / soundonsoundmag
Follow Sound On Sound: / soundonsoundmag
Follow Sound On Sound on Instagram: @soundonsoundmag
Learning so much from this series! Cracking the loudness "mystery" has been the toughest part of mastering for me. Just ordered the VU meter. Thanks for all your work on this series!
Pleasure, glad it's helpful !
This is such an excellent series, great thanks!
Absolute gold. Thanks for sharing
Realy valuable advices! Thanks.
Great series... thank you for doing this!
Another gem by Ian, thank you sir! =)
Great stuff, i am loving this mastering series 👌
I do a lot of Hip Hop and Dr. Dre is often lauded as having the best mixes in the business so I checked the loudness he has outputing on his CD's and I was shocked seeing his most recent release pegging at -3db RMS and just utterly squashed. So I mimicked it just to see if I liked it vs how I would have normally mastered and I *hated* the overly loud, limited, squashed sound. My -10 RMS master sounded way punchier, bassier, and clearer.
Right then and there I gave up on reference tracks and loudness worries. I just master as loud as I can get something with the least amount of limiting and gain reduction possible (typically 3db max GR on the final limiter). Sometimes that means -10 or -12, sometimes it is as loud as -5 at points. Depends on a variety of factors but at the end of the day as long as it sounds good and when you play it in between other contemporary masters it holds up then you did your job. Imho
@@patrickchampagne5883 I would agree with you that Compton isn't great. However, I also disagree that the mixes and masters on 2001 are good as well. Similar squashed, overly compressed sound that I just can't stand. I personally think The Chronic was the first and last time he mixed at a truely high caliber. Excluding some of his work with Snoop, Em, and 50 whose early materials pre 2005 when mixed by Dre sound much more similar to The Chronic than 2001.
Thanks for your opinion though! I appreciate hearing it
You've stumbled on the hardest thing to achieve, ultra loudness while still sounding great (potentially, I don't know which tracks/album you're talking about).
The question to ask is do you like the sound of the Dre records despite them being super loud? You might like the loudness and sound of the Dre records but are simply struggling to get your own tracks as loud while also maintaining that punch, etc.
If you can't get your own music to the same loudness while it still sounding good, you might want to work on getting your masters sounding as good *and* as loud, just as a personal goal/exercise.
Overall though I don't like ultra loud masters either and fully agree on mastering to the loudness of your own choosing, since you're the musician as well. Something to consider though is how your music sounds after it has been uploaded to spotify, etc. When played with other artists that spotify will autoplay after, does your music sound as 'good'?
Just some stuff to consider.
There is a lot that goes into mastering a track that loud without destroying all the excitement. It's possible but the techniques are pretty advanced. You'll have to use a fair bit of multiband compression, probably multiple stages of glue compression, dynamic EQs, and likely a couple of limiters. Even going above -10LUFS without sounding squashed and removing all the excitement is somewhat of a feat.
fantastic teacher!
I have downloaded the trial version of Wavelab Elements , i have been trying to see if you can do the automation you do on Another Day Calling in this video? Can Wavelab Elements do this? Would be nice to know and see how you actually did it, like how you opened the automation part in the first place etc? Thanks!! These are great videos, and i have learned something major from them..........That being that i have been mastering all my previous work WRONG!!!!!!!!!! You have put me on the right path to better masters, so a big thank you for that!!!!!!!!! Best Regards!
This is gold! Thank you Ian.
just awesome! Thanks!
My pleasure :-)
Love it!
Thanks for the information 🙏💪
This is so good.
Are LUFS like -11 still the target? Spotify wants -14 integrated LUFS and Apple Music, -16 LUFS. Don't know if that's changed since 3 years ago, but it's been these for a couple of years now. I use YouLean to measure, and I shoot for -14. Seems to fit when I add my music to popular playlists.
Brilliant! I think this can also be applied when i digitise my record collection. when i record straight from turntable to digital interface> Pro Tools the recording is always at a low db so i use a limiter to boost the overall volume before exporting as a master wave file.
That's a little different and bad practice if the goal is to digitise without altering the original master/record. What you're essentially doing is a quick and dirty 'master' on top of an already mastered record.
The main problem you have their is gain staging. If you haven't already got a dedicated turntable preamp (with a proper RIAA curve and boosting circuit) I would highly recommend getting one (Schiit Modi is good from what I've read). Then, you need to set the gain on your interface appropriately to get a decent level (bumping around -12dB is plenty) so you're getting the most out of your interfaces' A/D converters (analog to digital). Boosting too far will obviously add preamp noise, but if you're boosting the recording you'll be boosting noise anyway.
THX!
🙌
I think the klanhhelm has gain knob built in?!
I'm on elements and thinking is it worth purchasing pro? This series is pushing me towards getting pro💯
Hey Ian - Really interesting to see how you used the VU for mastering! I have one question though - What is your opinion on using filter curves with the VU to get closer to the audio actually perceived? And which Weighing do you prefer / or maybe why you prefer no weighing at all? Thanks ;)
it's not VU, it's LU or LUFS, so Loudness Units, it actually takes care of all the filters you mention in a standardised way. you may want to check out "EBU", SL and ML (short term loudness, and momentary loudness) in the MVmeter2 plugin which is free, have a nice day !
@@moliver_xxii I have both Hawkeye and MetricAB, so i know what we're talking about sure! I just wanted to know if he uses weighing and if yes why ;) -
@@joost3783 ok !
I need to know what the song at 16:12 is. Please, can someone tell me the name and artist?
Got a question here , i used many limiters to stay at -1dbfs .
While these limiters seem to do the job , i saw with a " true peak viewer" that my songs went higher than that value .
Sometimes , it shows that i'm hitting at -0.5dbfs even if i set up my limiters to not go beyond -1dbfs . It's crazy .
Could you explain here please?
It's a bit of a complex topic, but in a nutshell dBFS tells you the *sample* peak, but in some cases the reconstructed analogue waveform (or data-encoded stream) can overshoot higher than this, which is what you're seeing. That's why I suggest using a True-Peak aware limiter and meter, to avoid this. I show how this works more in the final video in the series.
Hi Ian, First of all, thank you for this video series, I'm a producer and mixer and now I finally found the best beginner tutorials for mastering. I have a little question If my song recorded and mixed at 96Khz what is the best way to dithering to 44.1Khz or I don't need to dithering?
Dither is only used when reducing the bit depth, don’t use it for reducing sample rate. Always reduce sample rate before reducing bit depth. And reduce the sample rate as the last thing you do (minus bit depth reduction)
Just to add a little more detail to what @@frankuswalk said - I agree that it's best to convert sample rates before you reduce the bit depth, and provided you store the 44.1 kHz version in a floating-point format, you don't need to dither.
But if you use a standalone application to do the SRC and then save it out to work in a different application at 24-bit, for example, then you DO need to dither. The SRC processing is done in floating point, so unless it's saved the same way, you need to dither.
Basically the simple rule is - always dither after any processing, whenever saving to fixed point (24 or 16 bit). If you're staying in floating point, you don't need to dither.
As others have said, whenever you are exporting at 24bit or 16bit (for CD) you'll need to dither. Most of the dithering algorithms are the same or similar. You'll want to use your final limiter at the end of the chain to do the dithering. It doesn't matter which limiter you're using; any good mastering limiter will have dither built in whether it's Fabfilter Pro L or Waves L1. I would recommend not to use the dithering in your DAW since you really want to listen to the dithering and choose the one that sounds the best for the mix. Most of the time in Pro L, I find that the Optimized dithering sounds the best. If you're exporting a mix before mastering, you should export as 32bit floating point and then you will not need dithering. But any time you export at 24bit or lower, always use dithering.
I'm a bit confused after I've watched this. Ian, you're saying we need to start by just pushing the gain of the song up to match the reference track? But aren't we supposed to get there with the use of all the different plugins and stages, not merely turning the volume fader up?
Well... yes and no :-)
I *will* be using them in the later videos, but their purpose isn't really to make the songs loud, it's to make them sound as good as possible *when* they're loud. The loudness is achieved by adding gain - but balanced EQ and dynamics mean it doesn't clip or slam into a limiter too hard.
The main reason I suggest doing it first is efficiency. If we did EQ first and then turned it up, we would need to immediately tweak the EQ again because the change in level will change the way we hear it (because of the equal loudness curve). You could crank up the monitor gain instead, do all the EQ & compression first and then turn it up later (at which point you'd need to turn the monitor gain back down) but then the limiter would start working and potentially change the sound you just decided on. You'll end up with the same result eventually, but it's more steps and more confusing, in my opinion.
This way everything is done in knowledge of the final sound of the limiter - which is only working very lightly, remember. In fact when I work I'm applying and adjusting EQ, compression and limiting all at the same time. I don't recommend you work that way when you're starting out, but it gets easier with more experience.
This approach also encourages minimalism - if you turn something up and it sounds fantastic with nothing but light limiting, you're done. Maybe it needs an EQ tweak, but if it then sounds fantastic, you're done - and so on.
Does that help ?
@@ProductionAdvice yes! Thank you for clarifying it. I'm looking forward to the next video!
@@filipjonathan Great :-)
homemastering
If you don't see the sign-up box, please temporarily disable and ad-blocking software or browser settings.
not work for me!
Sorry you're having problems, please can you try a different browser ? If it still doesn't work, please email ian (at) productionadvice.co.uk and I'll try to help you get it sorted
Thank you Ian. Question about monitor levels; Is there a sweet spot for full range near fields that is different depending on room size? I struggle with harsh mixes and wonder if it’s partly because I am not listening loud enough so I am not hearing the highs and lows as flat as I should. Does room size also dictate how loud we can listen? My room is about the same size as yours, can you say what spl reading you get with pink noise at your listening position? Again thank you. 🙏
If the room isn't well enough treated I can imagine you might get extra early reflections which can make things sound thicker than they should do, yes - and you might make things too bright to compensate. Or bass build-up at important low frequencies, which again will encourage you to pull those frequencies out, and result in harsh-sounding results - I talk about this in Episode 2.
Monitoring level is very personal so I'd be cautious about following any numbers, but for what it's worth the level in my room is around 80 dB SPL (C-weighted) with loud mastered music, peaking up to a few dB at the loudest moments. For stereo pink noise @ -14 LUFS, the measured SPL is 74 dB (again, C-weighted)
But as I say in Episode 2 where I talk about monitoring, it's more about finding what's comfortable for you. Loud enough to feel satisfying and exciting, but not so loud it fatigues you.
Hope that helps ?
@@ProductionAdvice Ian, thank you that does make good sense. It could very well be the room despite it already being treated for early reflections - I think the carpet isn’t helping. It may be making it too dull sounding. Although we shouldn’t rely entirely on the numbers I appreciate you sharing yours. I have heard 85 spl c-weighted pink noise is the target but that is way too loud in my room. So again thank you for your help and great advice.
This is great information!! But this maybe a dumb question but where do you get all the wav files of these commercial tracks? I have tracks I would love to have but not sure where to get? Artist website? HD tracks? Love to know. .... oh wait ......just found Itunes downloads. Somedays its a brain freeze
😂 Yeah for WAVs you'll need to buy the songs, but it's a worthwhile investment. For quick references I use lossless streaming services like TIDAL or Qobuz
@@ProductionAdvice But aren't most bought songs online some of the compressed file formats?
Thanks Ian , great tutorial. Do you still master a CD at -1db tp the same as streaming? I understand -1db tp for streaming allows for whatever conversions may take place to guard against overs, but is that also valid for CD?
Great question - and yes, I do.
Even on CD, it's possible to get peaks above zero at the DAC, if you're peaking right up to zero. In theory an inter-sample aware limiter will prevent that, but I prefer to play it safe.
And perhaps even more importantly, CDs get ripped, at which point the risk of clipping after a lossy data-encoding stage is just as much of an issue. If you follow my "no louder than -10 LUFS" suggestion it's fairly unlikely, but following these guidelines has served me well over and I'm sticking to them :-)
@@ProductionAdvice Thank you Ian, all great points! One level to rule them all! I like it! 😎👍
@@Beatledave7 Pleasure :-)
For CD I always go -0.1dbtp and run Pro L at 8x oversampling and master at 96k. I can be pretty certain that I won't go over 0 doing this. People still want their CDs loud and you don't want to sound weak next to a loud CD.
Lovvvvve it 🍒😃
Ok. So am I right thinking that you're not sure you wanna hit the limiter plugin with "a too hot signal" and at the very same time you're ok to set your VU ref.calibration at -11dB which is already quite high, especially if you use it, for instance, with some analog emulation plugins (and there's no guarantee you wont use it in your mastering pipeline) which usually take between -20 to - 14dB, and some of them can't be even calibrated to anything else but to "classic" -18dB..?
Yes and no 🙂
Firstly bear in mind that in the later videos I'll show how to use compression and EQ to keep the limiting at a minimum. And secondly, for analogue modelling plugins or actual gear I'd certainly suggest calibrating to -18 LUFS. With hindsight I guess I could have been clearer about that in the video 😕
Hi Ian, thanks for a really great series!! I got a question here. One thing I don’t understand is the need for using a trim tool or clip gain to increase the volume before any eq moves. What’s the difference of just using the fader in this case, the percieved volume and feeling will be the same, or??
Hi Hector, There's no difference in sound, but there are a few other factors.
First, in many DAWs the fader is post-insert, so adjusting the gain there won't affect the level that hits the dynamics processing (next episode !). That's not the end of the world, but I prefer to tweak the gain going in rather than having to keep messing with thresholds and make-up gain.
Second, it minimises the level-shift when bypassing the processing, which can have an effect on the way we perceive any processing, so I prefer to not adjust gain inside an EQ or dynamics processor.
And finally, it encourages minimalism. If we always get in the habit of adjusting the gain first, we hear the most reliable impression of the final sound, and are less likely to get sucked into unnecessary EQ and dynamics processing. If we can adjust only the gain and nothing else - great !
Hope that helps ?
Thanks for your reply Ian, yes.. that makes sense, specially getting into the habit of always gain staging, I forgett that most of the time after the initial gain staging.
thanks for the great series SOS and Ian,
i have a question about the initial level you set (from the clip vol and gain plugin). why push it to the limit ,isnt it better to have the songs "matched" levels ,in around -18 vu and then use the threshold on the limiter/maximizer when exporting ? especially if decide to involve analog gear at some point ? or its because you know you gonna go only ITB mastering ?
Either approach will work, and yes - if you're sending to external analogue gear (or analogue modelling plugins) then -18 is a better choice. Even in this case I would increase the gain of the returning signal to be in the right ball-park as I listen, though.
The reason is that if you decide to balance to a lower "interim" reference level like -18 first, you'll need to start with a higher monitor gain initially, and then later increase the final gain and adjust it back. You'll also then be adding limiting on top of the other processing you've chosen, which could change the way that it sounds and add extra work as you adjust.
Whereas I've found the approach to be most efficient. It does risk hearing too much limiting to start with - but that's actually useful info to have - "if I want to be this loud, I need heavy limiting or other dynamic processing". Otoh if you increase the level and it sounds great just with very light limiting and without extra processing, that's a nice minimal way to achieve a great result. And since the levels I'm suggesting are fairly conservative, hopefully nothing will need pushing too hard anyway.
Does that make sense ?
@@ProductionAdvice thanks for the detailed answer' yep it does makes sense
If I have to work on loudness first.! Why all mastering engineers asking for a mix within -8 to max -3 db .?
How loud should you master? -14 LUFS, with peaks below -1dBTP. Mastering any hotter will just reduce perceived loudness because of downward normalisation. The loudness war is dead and dynamic range is back.
Some people would agree with you about -14, but not me. Should an acoustic ballad and a heavy rock track both be mastered at the same loudness ? That doesn't make musical sense. Also some genres sound better with more intensity.
I do agree with you about dynamics though, it's why I started Dynamic Range Day :-)
There are a lot times I master hotter than -14LUFS. In fact, most of the time, I shoot for around -9LUFS. I master a lot of rock and metal and people are used to hearing it that way. With normalization, it means we don't have to care about the level anymore which is a really great thing but people are used to the sound of the music hitting into the limiter and if you want to master commercial music, the limiter is part of that sound. Otherwise, the song will sound cheap and unfinished. Human brains are a funny thing.
So how loud do I master? As loud as the song calls for. I bring the level up until I get the sound that I want. Who cares if spotify gains it down?
The only thing is, I never master lower than -14LUFS because I don't know how streaming platforms are going to turn me up and I want to make sure I'm not being put into a limiter. So classical music will get limited to JUST -14LUFS but no louder.
+1 on the -1dBTP. If you know the track is only for streaming, the streaming providers really don't want peak values higher than -1. For CD I still bring it to -0.1dBTP and that's only to catch the few accidental peaks that may go over because of sample rate.
2 questions...is the podcast coming back? I see a lot of folks using the TC Electronic Clarity for physical, always-on metering...any thoughts on this or similar units? Are they a good buy? Does the work flow beat having individual plugins of the same sort? That sort of thing.
Hi ! Yes, there will be more podcast episodes coming out soon :-) And yes, the Clarity M is an excellent loudness meter, in fact I hope to get one to play with soon.
Update - new podcast episode is out, and quite relevant to this video - "Are you having a LUF ?" 😛 themasteringshow.com/episode-73/
Great series thanks for doing this SOS and Ian. One question what does calibrating the VU meter to -11db actually do? does it make the scale of the meter -11 instead of -20? or does it make the 0db -11? could someone please clarify im feeling dumb!
I haven’t personally used that specific plug-in, but based on some experience with other plug-ins, I believe it sets the “zero point” of the meter to -11, just as you said. Not a dumb question at all - this stuff can be confusing!
Right - it just adjusts the scale of the meter.
By default it's calibrated to -18, so pink noise @ -18 RMS or -18 LUFS will read zero, and so will your music if it has balanced EQ and reads around zero. That's a great place to be for mixing, or looping out to external analogue gear, but for mastering you almost certainly want to be louder.
Adjusting to -11 on the VUMT means that pink noise @ -11 LUFS or -11 RMS will read zero on the VU, and again if your music has balanced EQ and you adjust it so it's hovering around zero VU, it will also be close to those values.
That's still fairly conservative as a final loudness these days, but I think it's right in the sweet spot of being loud enough to sound great, and work well online (or on any other format)
@@ProductionAdvice Ian, thanks so much for the reply, and the fantastic videos and website. I think there may be a typo in your second paragraph above - when you say "Adjusting to -18 on the VUMT means that pink noise @ -11 LUFS or -11 RMS will read zero on the VU", I think you meant "Adjusting to -11 on the VUMT". Is that correct? Just wanted to clear that up to avoid accidental confusion. Thanks again for everything!
@@ProductionAdvice ah I see as it’s not really measuring the peaks?
Thanks for the great replies guys!
@@liquidsolids9415 Oops yes, sorry ! I've edited my reply to fix it
so -10 LUFS is the level to shoot for?
That's the loudest short-term value I suggest, yes, but only at the loudest moments. Depending on the material that could give you a wide range of overall (integrated) values, though. I wrote about the idea in more detail here: www.productionadvice.co.uk/how-loud
@@ProductionAdvice Hello , isn't it the momentary loudness term please? Because the short term measures the loudness of the past 3 Seconds , while the momentary measures the loudness of the past 400 Milliseconds . For example , i have a mixed song only (no master) which has momentary term (-17) and short term (-23). Yhe highest value will always be for the average of the 400 ms (momentary), i think . Am i right?
@@mago97615 Momentary is almost always higher, yes, but I find it reacts too fast for me to use for making loudness decisions . I find it most helpful to judge based on the short-term loudness, even though it works over the longer window as you say.
Isn’t there Christmas bells or triangle on that wave in the middle
No :-) But if you mean something "tinkly" in the top end and you're listening on mobile, you might be hearing low-bandwidth lossy data-encoding artefacts...?
How is the room sounding so dead of reflections with so little visible acoustic treatment? It sounds like it was recorded in a highly treated acoustic space/booth.
A few things. There is actually more treatment, but it's in front of my seating position so you don't see it from this angle - check out Episode 2 @ 4'16". It calms things down at the first reflection points, reducing the reflections that get back up to the far end of the room and into the live side of the mic. Next I'm fairly close to the mic - probably only 6 inches away - although it is a condenser so it's pretty sensitive to the ambience. And finally I'm not talking very loudly :-)
My fire wall rejects you free down load
Sorry you're having problems, please email me ian (at) productionadvice.co.uk and I'll try to get it sorted for you
Hey Ian - Really interesting to see how you used the VU for mastering! I have one question though - What is your opinion on using filter curves with the VU to get closer to the audio actually perceived? And which Weighing do you prefer / or maybe why you prefer no weighing at all? Thanks ;)
I prefer no weighting, because I've been using real VU meters for 25+ years, so I'm really used to the way they react. That's just me though, if you're just starting out using one and find there's a different weighting that makes more sense to you or feels better, go for it !
There's actually some benefit to be had from the authentic weighting - if things sound loud but look low on the VU, you probably have too much mid-range, for example. Or, if it sounds a bit quiet but the VU is very high, there's probably too much bass.
No meter is perfect, they all need you to learn how they read for different situations and adjust, so the there's plenty of room to tweak.
@@ProductionAdvice wow dude thanks a lot for the thorough answer!
@@joost3783 Pleasure :-)