Mixing songs with dramatic dynamic differences

[mpedrummer]

Hello, Imaginary Internet Friends!

I'm nearing the mix stage for a project, and as I'm listening to roughs, I find myself cranking one song at the beginning, only to run for the dial when it gets to the end.

At points, the song is simply gentle acoustic and soft singing. At the end, there's full band, the singer is really belting, etc, etc.

So...who has tips and tricks for this situation? I don't want to send this off to mastering only to get it back with the end smashed into nothing to make the beginning loud enough. How do you balance the "parts" in this scenario?

Thanks in advance...

Tim

[Snarl 12/8]

It doesn't sound to me like a good ME would just set the dials so it sounds good at the beginning of the song and then walk out of the room. Can't they ride that shit in real time? Find someone that's mastered the same sort of shit and send them a sample and ask them what you should be aiming for at mix time. There might be some sort of key to doing it, like "don't let the bass get out of control at the end" or something. But those dudes have multi-band compressors and $1000 volume knobs and shit.

[Dakota]

When I master stuff like that, I might do the soft and loud sections as separate runs with different settings, then stitch the final master together from cutting between both.

There is also the parallel to series compression approach which works well for that kind of thing.

[MoreSpaceEcho]

yeah this is easily dealt with in mastering. but you could also just adjust the levels of the tracks at the beginning of the song so that you don't have such a dramatic volume change at the end.

[mpedrummer]

Lemme rephrase somewhat...

I have an excellent mastering guy, Garrett Haines (of TapeOp infamy), lined up for this project. I have the utmost confidence in him and his abilities.

I just want to maximize my chances.

The real issue, at the moment, is that everyone I ask to listen to the tune comes back with "it's too quiet" and most of them then fail to listen past that to provide me with something I DON'T know Turds.

Anyway, thanks guys.

Tim

[Ryan Silva]

When I master stuff like that, I might do the soft and loud sections as separate runs with different settings, then stitch the final master together from cutting between both.

There is also the parallel to series compression approach which works well for that kind of thing.

I a have always wondered if you do not request a "double processed and then cross faded in the middle" type of thing will the ME most likely recognize the vast dynamic qualities of the song, and apply that technique. Or should we be saying "Hey, could you process the two halves of this track separately" or will it be done as a stander approach?

Thanks

[Nick Sevilla]

Hello, Imaginary Internet Friends!

I'm nearing the mix stage for a project, and as I'm listening to roughs, I find myself cranking one song at the beginning, only to run for the dial when it gets to the end.

At points, the song is simply gentle acoustic and soft singing. At the end, there's full band, the singer is really belting, etc, etc.

So...who has tips and tricks for this situation? I don't want to send this off to mastering only to get it back with the end smashed into nothing to make the beginning loud enough. How do you balance the "parts" in this scenario?

Thanks in advance...

Tim

May I suggest that you take a little time out from listening to it.

Then, decide the following :

1.- So do I leave the mix the way it is, approved by the artist and myself?

2.- Do I change the levels of the beginning part, making it louder, so the level transition between beginning and full level is closer?

3.- Do I change the Full part level downwards, to better blend the beginning to the end?

Usually when I mix a song that has this, I tend to get a consistent vocal level throughout, and mix the instruments around it, even with a transition from one acoustic instrument to a full band.

As long as you have the vocal somewhat consistent, then the change can be dealt with a little better.

Of course, in the end, you must at least try to respect the artists' vision. If they want it really soft at the beginning, and really loud at the end, then by all means do it like that.

Cheers

[MoreSpaceEcho]

I a have always wondered if you do not request a "double processed and then cross faded in the middle" type of thing will the ME most likely recognize the vast dynamic qualities of the song, and apply that technique. Or should we be saying "Hey, could you process the two halves of this track separately" or will it be done as a stander approach?

i think it's pretty safe to assume that it's a relatively standard approach. personally, i process different parts of songs differently all the time. ok, maybe not all the time, but often enough. whenever i feel like it calls for it, client request or not.

[Ryan Silva]

I a have always wondered if you do not request a "double processed and then cross faded in the middle" type of thing will the ME most likely recognize the vast dynamic qualities of the song, and apply that technique. Or should we be saying "Hey, could you process the two halves of this track separately" or will it be done as a stander approach?

i think it's pretty safe to assume that it's a relatively standard approach. personally, i process different parts of songs differently all the time. ok, maybe not all the time, but often enough. whenever i feel like it calls for it, client request or not.

Good to know, I have a mix ready to go out that is full of such tracks, I don't like attending mastering sessions as a rule of thumb, but I may for this one.

Thanks

[Z-Plane]

Call me crazy but what about mixing the quiet bits louder ? Automation, bussing or lots of fingers should get you close enough to decide if it works. And yes, many MEs have mutliple parallel options to treat different sections of a song, as well as volume knobs. Seriously, you can and should take care of all this yourself, if its already a cause for concern I wouldn't leave that until the mastering stage to see if it gets resolved.

[mpedrummer]

I guess the issue I have with that is that I still want it to sound dynamic, except without actual dynamics

I was sorta looking for tricks like "duplicate the vocal track, make compression settings for the quiet and for the loud, and then switch between the two" or something like that.

Also, just to comply, you're crazy.



Tim

[Z-Plane]

If you use a DAW, I would first of all put a master mix on a channel and draw a little volume automation on just to see how that works. Perhaps bounce a separate vocal and instrumental, then try a little side chain with the vocals triggering a comp/limiter on the rest of the mix, this sometimes helps alter the perception of dynamics more favourably than other methods. You could always send it off to fight in the loudness wars, again just to see how your mix reacts to a firm squashing. With all of these things, the general vibe from the mix should still feel like instruments being played quiet/loud rather than being mixed as such.

[MoreSpaceEcho]

I guess the issue I have with that is that I still want it to sound dynamic, except without actual dynamics

what noeq said about keeping a consistent vocal level is a good place to start. just balance everything around that and you can still have a sense of dynamics via the arrangement. i.e. when the band kicks in it seems louder but really it's just fuller.

[Dakota]

+1 agreed w/noeqplease and MoreSpaceEcho - the key thing to get right (or wrong) is getting the lead vocal in a good and proportionate relationship for both dynamic levels of the song.

In any case, it seems to me that most listeners just can't deal with too much *actual* volume dynamic. If a listener has to ride their volume control just to listen through the whole song, annoyance sets in fast.

What I work on in these kind of circumstances is *psychologically perceived* volume rather than actual volume. One can make a section "feel" quiet or hushed by timbre choices rather than turning it down.

[losthighway]

Totally. And this can be achieved using eq settings as well. Also by changing the way the instruments are mixed to 'sit in' the speakers. Sometimes with drums for example you can have a quieter sound that leans hard on a mid heavy room mic kind of sound with the close mics pulled lower. Then when it's time to rock you can let the listener rediscover a kick drum with more attack etc.

That's just one example. I like a recording with some liberal dynamics, but to split the difference you can reduce 'volume change' and use more 'perceived intensity'. A little distortion anywhere can potentially add to the 'perceived intensity'.