I start by apologizing for the length of previous posts - I proceed to write the longest post in the series - I sneak in a link to a Pink Floyd song and no one suspects a thing - I raise awareness about canine tinnitus - I finally get to the point: telling you about an awesome new loudness monitoring tool
Hey! If you enjoy this article and want to learn more about loudness and how loud to make your songs, check out Episode 46 of the DIY Recording Guys podcast - which I host!
Honestly gang, I am sorry about the last two blog posts. This was supposed to be about bite-sized wisdom.
I’m going to get right into it.
Here are some of the ways I monitor peak and RMS levels throughout the mixing and mastering processes:
1. Watch the master bus meters right from the beginning
As I start building my golf ball mix (this the first part of the mixing process where I start unmuting the most important elements of the song - read this for more info) I’m immediately watching the master bus meters to make sure I have plenty of head room. I typically like my master bus to peak somewhere around -10 dBFS for my golf ball mix. Once I have all of my mix elements in place, I like to target -6 dBFS or -5 dbFS as a final value for the peaks in the track. If you do not do this from the beginning, you risk running out of headroom.
It's like when you're filling a glass of water for a toddler. Do you feel it to the tippy-top and then hand it to her slowly? No. Nor would I...not ever again. You fill it like halfway because it's going to go on a bit of a bumpy journey from here to over where the dog is chewing your shoe - which still had two years until it was planning to retire.
Right, so you risk pushing your entire mix close to 0 dBFS and possible clipping. In addition, even if your master fader is not clipping, you may be clipping somewhere in your master bus processing chain - like one of those awesome plug-ins that is supposed to make people's clothes fly off instantly.
If you find yourself at this point, you have to go back and start turning stuff down.
"Cool. No problem. I'll just turn down these here drums...oh...now they're too quiet...Ok I'll just turn down that there bass...oh...now the guitars are too loud....ok I'll just..."
You get what I'm saying. It's a pain. How do I know?
I've done it. Nope. Like way more than once. Look it doesn't matter HOW many times, ok?
The point is, once your master bus is too loud, you have to start turning stuff down. This is annoying at best or downright deranging at worst as you're forced to reevaluate your balances and readjusting everything in the mix. This is rework. Rework is a waste of time.
2. Use Trim plug-ins or Output Controls
As I work past the “static” part of my mix where I just adjust levels of tracks using fixed fader positions, I will start adding automation to faders and various plug-ins. Automation is basically programming that allows for faders, knobs, mutes, etc., to be adjusted automatically as you play a mix back. So, for example, you can use automation to make the drum bus +1dB louder during the choruses but then have it turn back down automatically for the verses.
This is a powerful mixing tool. The problem is that you’ve now given automation control of your volume fader. Let’s say you decide that the drum bus is actually too quiet for the entire song and you want the verses to be +1dB louder - while still having the choruses be +1dB louder than the verses. Well, you have a few options.
One option is to go into your automation program and turn all of the automation values up manually.
Hint: this can be super annoying when you’re dealing with something like a lead vocal which can have a 50 or more automation points. You can usually grab multiple points at one time and so on but I personally don't like messing with my meticulously (read: neurotically) crafted automation.
The better option is to pick the last plug-in in your processing chain and adjust the output control up +1dB. So if the last plug-in in your chain is an EQ, you can open the EQ and, if it has an output level control, you can turn that up +1dB which will turn the whole track up.
If all else fails, you can insert a “trim” plug-in (which is basically just another fader) upstream of the automated fader. You can use this to adjust the volume up or down going into the automated fader.
3. Use a reference track
A reference track is a professional track that is similar in style to the track you’re working on.
Think of it as a......reference.
A lot of people recommend pulling a reference track into your mix early in the mixing process and regularly toggling between your mix and it.
For me personally, this has never worked. I find it super distracting and disruptive to the creative process. I end up trying to get my kick drum to sound exactly like the kick drum in my reference track. It’s like a “hold my beer” reaction.
“Oh yeah? Your kick sounds like that? I bet I can match it”.
That’s not productive. I want my kicks (and everything else) to sound the way in which they best serve the track. Anyway, I personally don’t like to mix with reference tracks but it’s more a function of my own neuroses.
Where I DO love to use a reference track all day every day is at the end of a mix when I’m just tweaking the mix bus or when I’m doing my own mastering.
I like to listen for things like:
a. How much low end does my mix have compared to the reference track?
b. How loud is the vocal track compared to my vocal track?
c. How “sparkly” is my upper end? Is it too “trebly” or not “trebly” enough?
d. How does my mono mix sound compared to the reference track mono mix?
e. And, of course, how loud is my mix compared to the reference mix?
4. Use an RMS meter
In Part 1 we talked about how the “average level” of a waveform is what we actually perceive as “loudness”. Most of the meters in our DAW, however, are not showing us the RMS value. They’re showing us the peak value. We saw in Part 2 how the peaks of two tracks can be the same while the perceived loudness is different. Some plug-ins have built-in, pseudo-RMS meters which measure on a scale called dbVU (decibels Volume Units). This is kind of a “dumb” value that doesn’t tell you much in and of itself. It’s usually used to make sure you’re not over-loading a plug-in or piece of hardware. However, it’s very useful on a RELATIVE basis. In other words, you can run your reference track through a plug-in with a dbVU meter to get a “baseline” and then run your track through that same plug-in to see how the RMS level compares.
Other plug-ins will actually show you the dbFS RMS value of your mix. Again, this is the perceived loudness we actually care about. These are super useful.
I use the output meter in the Izotope Ozone plug-in suite - which is what I use for mastering.
5. Use this awesome website
Right, this is why I even started writing these posts. You know those songs that have a really long intro?
You put the song on because you thought of the chorus in the shower and you think, "ohhh right. the intro. I'll go fix a snack." and then you come back and the intro still going and at some point you forget what song you even put on? And then it's too late to play loud music because the neighbors' dog has a rare kind of canine tinnitus that really acts up after 10PM?
Anyway, this blog is kind of like that.
Where was I?
Oh, so Ian Shepherd has worked with Ian Kerr for a while now making some great monitoring tools for various parts of the music creation process. You can check them out at: www.meterplugs.com
One of the latest things they've developed is this freakin' awesome website: www.loudnesspenalty.com
Remember in Part 2 when we talked about how the loudness wars were ending because streaming services had started turning down tracks that were too loud? Well this website let's you quantify that effect for the different services.
It works like this:
You drag your track (I've tried both MP3s and .WAV files) into it.
It analyzes the track and tells you how much each popular streaming service will turn it down.
What a relief!
I always found that, in order for my tracks to sound competitively loud to my reference tracks, I had to push them louder than I actually wanted to. Almost always, I would reach a point where the punchy transients started to get a little smeary and I still had a little ways to go in terms of loudness. And this would make me sad ☹
Now I see that all of the streaming services would turn my track down anyway!
It’s an awesome tool and it’s free to use.
Here are my results from a recent mix:
The first master I did was using a pro reference track and a qualitative, close my eyes and listen, method for matching loudness. Here are the results
Nice! I was kind of pumped. This meant that EVERY service was going to turn my track down.
So I went back and adjusted the limiter settings to try and get closer to what these systems were normalizing to. Here is my second attempt:
In my DAW, this one actually seemed quieter than I was comfortable with and yet it would still be turned down a hair.
Finally, since this is a track I'm using in my portfolio for Calm Frog Recording, I had the track mastered by a top-notch mastering engineer. I ran his version through this test as well. Here are the results:
Whoa. Interesting! So the mastering engineer's version was almost as loud as my original version.
My takeaway is that the loudness war maybe isn't quite as "over" as I implied in the last post.
Mastering engineers are still going to use judgment developed over many years of experience; judgment that has gotten them this far. It will probably take some time before these types of tools and the impact of streaming services turning songs down takes full effect. Still, I definitely think that the momentum has shifted away from loudness; and this is a good thing for all of us - producers and listeners alike.
Cheers and happy recording
Vadim
תגובות