Mixing Question: Problems getting healthy dynamics range, re: vocals

Tesgin

Member
I'm having difficulty getting a healthy dynamic range in my mix. We're dealing with a rock song, three guitars, drums, bass, vox.

I am reading up on what are reasonable targets for dynamic ranges for different styles of song. Basically, I've read numerous articles that say the target is a Loudness Range (EBU terms here, LUFS, etc.) of no less than 9. Found a Pro Tools article that's a littler more realistic, I think. It says to shoot basically for between 5 and 7 LU, depending on style, and lists dozens of popular songs along with their Loudness Range.

My problem is that I can't get my song's dynamic range above 3.8 or so. I've pinpointed the problem: it's my vocals. If I remove the vocals, all my measurements are great and it sounds great. -12 to -14 LUFS Integrated, PLR-S around -8 or 9. PLR Short Term around -9. Loudness Range (LRA) around 6 or 7.

When I add my vocals in I just can't keep the LRA within reason. It's always less than 4. I've tried using less compression, adding volume envelopes (which helps the numbers, but doesn't sound right -- sounds not natural).

I'm thinking I just need to re-do the vocals altogether, which is my next step.

Honestly, I like the sound of the song, but I recognize I don't have enough sonic variety in the dynamics.

Is it okay if the "band" has sufficient dynamic range even if the vox don't?

Thanks in advance. I've worked hard on this, just can't seem to get it. Any thoughts or tips would be most appreciated.

Tesgin
 
It says to shoot basically for between 5 and 7 LU, depending on style, and lists dozens of popular songs along with their Loudness Range.

Not sure where "it" says that...I would think more along the lines of -16 or -14 (Amazon, Spotify) is what you will need for streaming/broadcast, if go with 5 -7 they will crush your mixes down to mush.

If you're just making CDs to sell...do what you want, just don't rip audio off the CDs to then send to streaming/broadcast.

Right now I'm working on a pretty loud mix, lots of stuff going on, but with a healthy dynamic range...and I'm hitting at -22 right now, so I have room to work with.

Maybe you can post your mix...kinda hard to give advice without hearing the actual mix. :)
You need 10 posts to provide links...but we can work around that.
 
Thanks, miroslav.

So, to clarify, I'm not talking LUFS Integrated; there my song is around -15 or so. I'm talking LRA. So, as Steinberg describes it, it's an analysis of the average difference between the loudest and softest parts of your track. measures the average difference between the loudest and softest parts of your track and "ignores" statistically insignificant variances in loudness (e.g., fade-outs).

The goal there is LRA of 5-8 or so.

Hm,m. Try this:



---------- Update ----------

Hey, it worked.

Looking forward to your feedback.

Thanks.

TB
 
Last edited by a moderator:
Thanks, miroslav.

So, to clarify, I'm not talking LUFS Integrated; there my song is around -15 or so. I'm talking LRA.

OK...I thought you meant that *without* vocals, you were hitting around -15.
AFA the "LRA" thing...TBH, I'm not sure what that is, and I don't use any Steinberg stuff...so I can't reference their LRA numbers.
Like for instance, I have a plug that shows a dynamic range meter with optimum being -8 to -6...which means nothing out of context for anyone who doesn't have that plug.

So I took a quick listen on my laptop (I'll listen again later on my studio monitors)...and I'm hearing a rather peaky/bright high-end on your vocals in a lot of spots. I don't hear a lot of low end energy that could cause problems with loudness, something I deal with on my own vocal...but that peaky high end could also be the issue too.
Again...I'll listen on my studio monitors in a bit...since laptop speakers don't translate much of the lows to get a better picture of what you have.
Other than that, your mix doesn't seem to be nuked...there's a decent dynamic range there. Vocal compression might help, but I do think there's a some EQ issues there.

I'll post back in a bit...
 
Yeah, the vocals are out front a bit.

I tried to edit your link so others can hear it easily, but I got a strange 'Non Secure' warning. Yet it opens for me.

Let me know if it working now.

If not post the link again for me. I'll fix it.
 
All that matters is how it sounds. There's no good reason a decent song can't just barrrel along and be about the same average loudness all the way through.

BTW - None of the streaming platforms that I know of will actually "crush" anything. That doesn't even make sense. A mix isn't dynamic enough so you compress it even more? No. They will literally only ever turn it down so that the integrated loudness matches their target. This would mean a less dynamic mix will peak lower, but the DR - the difference between average and peak - will remain the same. None of them will turn up a "too quiet" mix, though. That would require compression/limiting, and nobody does that.
 
BTW - None of the streaming platforms that I know of will actually "crush" anything. That doesn't even make sense.

If your mix is very hot and compressed to "straight line" norms, and it's coming in at say...-6 LUFS...and you upload to a streaming service that normalizes it and reduces it down to -16...that's the equivalent of "crushed". So what was overly loud mush, will now be cold mush.
It will sound wimpy and actually perceived much softer than the original intention of it sounding very loud.

Not sure what part of that doesn't make sense to you? :)
 
If your mix is very hot and compressed to "straight line" norms, and it's coming in at say...-6 LUFS...and you upload to a streaming service that normalizes it and reduces it down to -16...that's the equivalent of "crushed". So what was overly loud mush, will now be cold mush.
It will sound wimpy and actually perceived much softer than the original intention of it sounding very loud.

Not sure what part of that doesn't make sense to you? :)

Ironically - maybe, this is not a metal guy. That is usually an issue there.

For this I feel it is more a mix issue that can be resolved without worry. Stop looking at numbers an get the mix right. Listen on multiple speakers and find the best for the song. Forget output volume if the mix is not there yet. Which I would say is not...
 
OK...I thought you meant that *without* vocals, you were hitting around -15.
AFA the "LRA" thing...TBH, I'm not sure what that is, and I don't use any Steinberg stuff...so I can't reference their LRA numbers.

No, I meant that I'm getting a decent LUFS Integrated (around -14 or so), decent Short-term LUFS (around -9 in the loudest parts of the song), and a decent PLR (Peak/Loudness Ratio) Short Term (9 or so). But my LRA is only 3 or so. Frustrating that I can't widen that.

Maybe, as ashcat_It said, as long as it sounds okay, don't worry about it. But I do want to learn how to do this right. I think there's something for me to learn here.

I think I'll take a crack at re-recording the vocals altogether. Maybe THAT'S where the dynamic range problem is. I just didn't vary the vox enough.
 
...So I took a quick listen on my laptop (I'll listen again later on my studio monitors)...and I'm hearing a rather peaky/bright high-end on your vocals in a lot of spots. I don't hear a lot of low end energy that could cause problems with loudness, something I deal with on my own vocal...but that peaky high end could also be the issue too.... Vocal compression might help, but I do think there's a some EQ issues there...

It's a bit harsh (piercing) in my phones. It's competing a bit with the center guitar.
 
There's no good reason a decent song can't just barrrel along and be about the same average loudness all the way through.

Thanks, ashcat_It, that's helpful. I guess a guy can get overly neurotic about numbers. :)

BTW - None of the streaming platforms that I know of will actually "crush" anything. That doesn't even make sense. A mix isn't dynamic enough so you compress it even more? No. They will literally only ever turn it down so that the integrated loudness matches their target. This would mean a less dynamic mix will peak lower, but the DR - the difference between average and peak - will remain the same. None of them will turn up a "too quiet" mix, though. That would require compression/limiting, and nobody does that.

Yeah, I totally get that. My overall LUFS Integrated is in the range of -14 to -16. The 3 I was talking about (with the rx'd target of at least 5-8) is the LRA, the Loudness Range. My overall level of compression is good enough that it won't get crunched by a streaming service, but the range within the song from the loudest LUFS to softest is too constant. Am I making any sense? I hope I'm explaining that properly. Just trying to figure this out myself.

Tesgin
 
I tried to edit your link so others can hear it easily, but I got a strange 'Non Secure' warning. Yet it opens for me.

Let me know if it working now.

That worked. Thanks.

(Not sure what the nonsecure warning is. it's just my dropbox :) )

Tesgin
 
Hi Tesgin...

Its a good thing Miroslav asked you to post that mix. I was about to do the same until I saw you had already linked to it.

First of all, we need to completely forget about any articles you've read on the topic of LUFS or dynamic range for a little bit. None of it applies to this.

So what's going on here is there are some major dynamics management issues with how the layers of the mix are structured. This is NOT an issue that has anything do with dynamic range. It has everything to do with dynamics processing, and the way the EQ was managed. The dynamic range problems are a symptom, the two other issues are the cause. Basically, your vocal is having a hard time finding adequate dynamic range because the foundation you've given it to sit on is really unstable. This is creating abrupt drops in vocals in some places, where the vocal is buried then surges way too far forward. I would go back and re-assess the relationship between the kick, snare, bass and whatever cymbal the drummer is riding in certain parts of the track. I would also strip the vocal of any effects that are on it until you can figure out a way to rebuild the signal chain. It sounds like you have some kind of room or doubler effect on the vocal and its really hurting it. You're going to have to use some very aggressive compression on the vocal here, but that's ok and perfectly normal in this genre.

Its really hard to say how to go about doing this without having the session infront of me, but there's 2 ways you can do this. You could scrap the entire mix and rebuild it from the foundation (the kick, snare, cymbals, bass, and only one guitar). And perhaps some of us could talk you through it from there. Or you could start by modifying the vocal chain so it places you in a better position to decide weather to re-shoot the vocal THEN rebuild the mix. There is no one-off solution or formula any of us can give you to make this work, because the tone and dynamics management problems prevalent through many different parts of that mix.

If you want to go after the vocal right off the cuff, I would start by autotuning it, because the missed notes are distracting. I would immediately saturate, then anchor it with an aggressive Optical compressor running as high as 10 to 15 db of GR with a fast attack and fast release. I'd run a transient designer after it to make up for some of the lost detail, then I'd probably place a second FET style compressor behind this with medium attack and slightly slower release. By that point you should be able to hear it clearly and keep it from disappearing on you.
 
After listening to it on my studio monitors...mmm...yeah, there are some out of balance things happening.

I get the feeling you focused the most on the guitars. :D
The Bass and the drums are kind of dull drone...and then the vocals seem to be in some isolated sonic space, peaky and they don't follow the changing levels/dynamics of the backing tracks. Like where everything drops out or goes down in volume, the vocals are way out there...then when everything kicks back in, they seem to get sucked up by the backing tracks.

If the vocals were recorded like that, and not something you did with post EQ...yeah, you might want to record them over.
What was you vocal recording chain?

I would also have the drums up more, but not just in level, they're pretty dull sounding, they need a bit of punch and crack, otherwise, the bass guitar is smothering them.
The rhythm guitars are a tad mushy...they could use a touch of articulation, otherwise they can work if the rest is sorted out.
The lead guitar should also come up and be a bit more ballsy, IMO.

Oh...and I agree about the tuning thing on the vocals...if you re-track them, you need to hit those higher notes a bit better...or use some pitch correction.
Sometimes, you can also mask it a bit with background vocals, but that depends if the arrangement really calls for them.
 
I don't have much to say other than agree with the posts you have received. There is a disconnect in the mix that needs to start with the beginning...

I have a week off. Maybe send me the wav files and I can see what I can do.

I can't believe I just saw a statement from Miro to tune vocals. lol! I'm gonna go throw away my blanket now...:)
 
Back
Top