Old 11-22-2017, 10:10 PM   #1
RDBOIS
Banned
 
Join Date: Feb 2016
Location: It changes
Posts: 1,425
Default Compressing a crescendo?

As per the title - How do I compress a song with lots of dynamics, especially, a song that goes from quiet to loud.

Sometimes, there are song that, while the sound level is climbing in a loudness (i.e. crescendo), have some peaks that ought to be tamed all along the crescendo.



Many of my songs look like the image above, or have two or three climbing sections. I know that sticking a compressor on the Master Track is probably not a good idea, as it would require way too much compression of the loud section to get to the soft section at the beginning to the song. I don't want a sausage looking waveform.

Does that make sense? I'm I explaining the problem correctly?

Same goes for a limiter; doesn't seem like a good idea.

How does one tame the peaks that occur in various places along a song that increases in loudness? Or any really dynamic song for that matter.

Should I ride/automate the threshold value of the compressor, so that I'm always seeing the same gain reduction (e.g. -3 db)?

Is there a way to tell the compressor, 'do what you need to do' in order to always have the same gain reduction ---- assuming that a steady gain reduction over the entire song would actually tame most of the peaks. Is this strategy nonsensical?

Is there a way to sidechain a signal that represents the perceived loudness, that would tell the compressor to adjust the threshold accordingly?

Just do everything manually in the volume envelop; don't use a compressor?

How do you do it?
RDBOIS is offline   Reply With Quote
Old 11-22-2017, 10:33 PM   #2
Fergler
Human being with feelings
 
Fergler's Avatar
 
Join Date: Jan 2014
Posts: 5,207
Default

You got it, automate the threshold. Some compressors do this automatically by setting the threshold as compared to the RMS level of input. But this is a case where automation is a much easier and more accurate method.
Fergler is offline   Reply With Quote
Old 11-23-2017, 11:35 AM   #3
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

A few years back there was this thing that everybody got all crazy about called "two stage compression". It was actually more like parallel compression but with a comp on both signals. One compressor is set to smooth down the low part, and the other with a higher threshold to handle the louder part. This actually makes the lower part louder and closer to the the loud part but also more consistent. I've never tried it, and haven't heard much about it in a while. Was an interesting fad.

That actually does look like a really wide swing. Looks like either the beginning will be inaudible in many typical listening environments or the end will be absurdly loud (and probably distorted) or your listener will turn the volume knob in the middle somewhere and won't get the full experience anyway.

But you need to be careful about compressing for looks. Especially when your talking about the peaks in the lower section. They look pretty proportional all the way through to me. The material gets more dense - more instruments playing more often probably - but nothing looks too far out of whack. You need to listen to it, though. Does any of it actually sound like a problem?

But all I've got is what it looks like, so...

It looks like you could get away with some fairly broad volume Automation just to sort of "undo the slope" and then maybe some compression over the top of that.

I would go back to the individual tracks if possible, though. Control their internal dynamics and re-evaluate the overall arrangement and mix a little better and it should just fall into place and a little bit of mix bus compression after that will be icing on the cake.

But again, it doesn't much matter what it looks like.
ashcat_lt is offline   Reply With Quote
Old 11-23-2017, 11:53 AM   #4
Judders
Human being with feelings
 
Join Date: Aug 2014
Posts: 11,044
Default

ashcat_It nailed twitch the ol' "use your ears", and Fergler is right on about riding threshold if you want consistent gain reduction throughout...

But you have to think about the effect you are producing. Generally, if I am doing something that crescendos like that, I'll leave the compressor to only work on the loud parts. This is because our ears compress loud noises, and quiet sources in the real world sound more dynamic to us. It also allows for more expression and dynamics in the quiet sections. When it all kicks off at the peak of the crescendo, the compressor kicking in can make it sound like it is even louder than it is, if done alongside filling more frequencies in with arrangement.

It all depends on what you want to achieve though, that's just how I've approached this kind of thing most often.
Judders is offline   Reply With Quote
Old 11-23-2017, 01:27 PM   #5
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Quote:
Originally Posted by Judders View Post
ashcat_It...
It's actually LT as in Lorenzo's Tractor. Happens regularly, and doesn't really bother me, but sometimes I like to spam a little.
ashcat_lt is offline   Reply With Quote
Old 11-23-2017, 01:31 PM   #6
Judders
Human being with feelings
 
Join Date: Aug 2014
Posts: 11,044
Default

Quote:
Originally Posted by ashcat_lt View Post
It's actually LT as in Lorenzo's Tractor. Happens regularly, and doesn't really bother me, but sometimes I like to spam a little.
Well, I never knew! I'm always too busy making sure autocorrect doesn't change it to ashcan!
Judders is offline   Reply With Quote
Old 11-23-2017, 11:33 PM   #7
RDBOIS
Banned
 
Join Date: Feb 2016
Location: It changes
Posts: 1,425
Default

Got it on the automate the threshold.

Got it about using my ears and fixing crazy peaks on individual tracks.

Seems reasonable to not being bothered too much with the variations in the silent parts, but focus a bit more on the loudest section at the end, to make sure that section doesn't clip, for example.

My mixes don't normally sound too problematic. That waveform image is not one my songs, it's Stairway to Heaven.

I'm rather new at mixing, and because I've heard so many 'experts' say == "compress this, compress that, limit this, limit that, leave -3db of head space, master to -0.1 db", etc., LUFS is this, RMS of that, I often feel confused about how it relates to my songs, especially those with a steep crescendo?

I know I've already ruined one my songs by over compressing and raising the volume with a limiter. The song I had composed was meant to be very quite and get much louder in time, but ended up being loud all the way --- which sounded strange especially given that the beginning of the song had only one classical guitar while the end had two electric, one acoustic, a synth, a bass, and drums...

A part of me really doesn't like compressing the sound of instruments. But,
I recognize that I'm not always playing guitar, bass, etc., with equal strength/volume throughout the song. I know that compression really helps the vocals. But still...

What I really don't like it slapping a Limiter at the end of the mixing process. Perhaps it's because I'm a novice who doesn't know what he is doing, but I really don't understand why I would add a plugin to change the dynamics of the song I work so hard mixing? I mean, once I'm done mixing the song sounds perfect to me (that was the goal of mixing, right?). But, then I add a Limiter, set the limit to -0.1 db (for example) and crank the nob to make everything louder - with a resulting song I don't enjoy (e.g. the quite guitar in the back is too loud, some drum hits have become disproportionate, the vocals have become harsh, etc.).

Strange to hear so many artists and producers complain that they hate the final cut after mastering. That they liked their mix better... I even heard Daniel Lanois talk about this issue regarding his latest album with Neil Young. You would think that if someone like Daniel Lanois is happy with the sound of his mix, that it would be cut that way?! But he suffers like most...

And here I am, doing it to myself. Because, why?

The waveform I showed in the image example is exactly what I want; the song putting to listener in a calm peaceful state, almost daydreaming, to the heartbeat getting faster with no room left to daydream, the mind lost in an emotional frenzy. If the waveform becomes a 'sausage' then all is lost...

At this point in my amateurish home studio producer/mixer growing phase, I feel I need to break free from most of what I've learned about compression/limiting. Especially once I'm done with the mix. Heck, it's not like some big-exec will complain that it's not LOUD enough, that listeners will boycott because they need to crank the volume nob on their I-pod docking station, or that I care if my songs don't sound like other songs on the radio.

I don't like sausage-like music.

Just looking for a way to smooth out a crescendo.

Thanks the tips
RDBOIS is offline   Reply With Quote
Old 11-23-2017, 11:56 PM   #8
ashcat_lt
Human being with feelings
 
Join Date: Dec 2012
Posts: 7,272
Default

Do that then.

It's worth noting that the song you've showed us has been proven to sound pretty damn good even after all of the multiband manglation that a big money classic rock station can put on it to make it louder than their competitors when you're seeking on your car stereo. That takes something right there. It takes a great mix of a great recording of a great performance of a great arrangement of a great piece of music. So...get on that...
ashcat_lt is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 11:56 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.