Setting export levels before mastering and 32bit floating point
A common question we get is “how much headroom do you prefer on the premasters?” The answer is “anything under clipping is fine.” But to expand on that, clipping is fine too butonly if you send 32bit floating point files. The reason for this is because 32bit floating point is not actually going to clip until the bit rate is reduced.
If your DAW is limited to 24bit exporting (Logic) then it’s important that you keep any peaks below zero because we can’t reverse distortion in mastering.
This is useful info to the producers out there who work with processing on their master fader (Ozone for example) and remove it before exporting only to find that the level is now peaking over zero. Now you can save yourselves some time.
There’s been a small discussion on Twitter lately about the loudness wars, dynamic range and distortion in mastering. Because Twitter limits your thoughts to a few characters, I figured I’d do a short post here instead.
Rather than get into the loudness war debate which has raged on for a decade, I’d like to touch on a subject that you don’t hear about much…MP3s.
It’s safe to assume that music is being listened to on MP3 more than WAV or even vinyl these days. They save hard drive space, quicker download times and let’s face it, most people can’t hear the difference between 320kbps MP3 and WAV anyway.
The thing is however, most people aren’t listening at 320, they’re listening at lower bit rates. Soundcloud has become somewhat of the standard for searching out new music and they are converting to 128kbps.
If you take a master WAV that’s been mastered to peak at 0dbfs and convert it to 128kbps MP3, it’s now peaking at least 2db’s over zero! That’s audible distortion and people are now accustomed to hearing it daily.
Sometimes our transcoding system can create audio artifacts, as we transcode all tracks to 128 kbps mp3 for streaming playback. Uploading a lossless or high-quality lossy file will usually reduce these to a minimum, but unfortunately there’s not much we can do for the handful of individual tracks that are…
We use the Ogg Vorbis format for streaming. There are three quality ratings that we use:
q3 (~96 kbps) Spotify mobile “Low bandwidth” setting
q5 (~160 kbps) Spotify Desktop standard streaming quality
q9 (~320 kbps) Spotify Desktop “High bitrate” setting available with Spotify Premium
Ok, so you have to pay Spotify monthly to not hear distorted tracks, fair enough…
The only way you’re going to hear MP3’s that aren’t distorted is when the labels decide to account for it and either master quieter FOR MP3, or offer two versions.
To put things into perspective on how out of control things have gotten, think about your average bedroom producer who’s trying to match his levels to commercial masters and drives his limiter into distortion to get there, then uploads it to Soundcloud and gets more distortion on top of distortion…
The question I pose is, do people care? Is this the standard now? The only way anything will change regarding the loudness wars and internet playback distortion is when the consumers speak up and demand change. Labels will listen.
When push comes to shove, you’ve just gotta respect those that paved the way for music we have today. Kids truly do have it easy now in comparison…easily obtained software, running full studios on a laptop with a pair of speakers. Back in the day, we had to go into debt to pick up a few pieces of equipment and get started. There was no social media, no mp3s…the internet was barely a thing yet! People were selling vinyl out of their trunks and pleading with stores and distributors to pick up their release so it could be heard.
A lot of music that was made during this time is still revered as some of the best ever made and the source of inspiration for a lot of your favorite new artists’ careers. It’s truly amazing to think of the things that were done in Drum & Bass studios in the 90’s with a simple MIDI setup, sampler and a desk. These artists were true audio visionaries pushing boundaries in music. Ask any of the top Dubstep artists what they listened to growing up and they will say Drum & Bass. It’s undisputedly the foundation for what “Bass Music” is today.
One individual who has been there since day one, through thick and thin, consistently releasing some of the most deep, dark and mind-bending Jungle/Drum & Bass is Dominic Angas. Building his foundation as a Moving Shadow and No U-Turn artist in the 90’s, Dom is as legit as they come.
Knowing that Dom is almost as much of an audio freak as me, I knew he’d be up for a little studio interview. So we present the first installment of a new series: Artist Spotlight featuring the one and only Dom & Roland.
You’ve been in the game for a long time, how old were you when you signed your first tune?
Wax Musical came out in 1994 on Saigon Recordings, which was an offshoot of No U-Turn. I was 19 at the time.
A lot has changed since those days, what are some of significant changes you’ve seen over the span of your career?
Not everyone had access back then to make and release a record because of the initial outlay on buying equipment or renting studio time. I got a loan from the bank to buy the equipment I needed…£5000 i think it was. I bought a Roland sampler, a CD-Rom drive, an Atari 1040 with Midex running Cubase, a Yamaha SPX-990 effects unit and a portable Sony DAT to record my masters with. To eq or add effects to something, I had to run the sound out of the onboard sampler output eq and resample it back in through an input using a cable. It sounds dated now, but it worked. There was a lot more variety in the tone of the sound as everyone had different equipment.
Also back then we used to listen to music only on vinyl and cassettes. If you owned a track then most likely you had actually paid for it. The internet was not widely used.
I think young peoples attitudes have changed too. Back then it was cool to like something no one else had heard of. I was excited to hang out down the record shop and spend my money on something not everyone else had. It made me really feel part of the scene. Especially if the guy in the record shop slipped you a white label promo from under the counter.
Nowadays, without stating the obvious differences in format, things have come full circle. Apparently, according to a researched article i read, it’s much cooler to like the same artists your friends like. Being avant-garde in your musical taste gets you labelled as weird or eccentric. Which just sounds dumb to me, but then i’m 38 now, which in young person speak is f***ing ancient.
Was it hard to say goodbye to Roland? It definitely had a signature sound to it.
Not really…I never said goodbye. He still sits in my rack looking content among all the newer equipment surrounding him. I actually turned him on the other day by twisting his little power knob. I used him on my last album on the track “Terminate” which i wrote only using my desk, hardware effects and outboard samplers.
Your tracks always sound larger than life, truly epic with cinematic soundscapes. What inspires you? Movies…Music…Life?
Movies, life, other peoples music, the usual things. I grew up playing classical music, and with both my parents being opera singers, i suppose I inherited a love of all things epic…I can’t sit through an opera without getting antsy or falling asleep though. I was always a sci-fi aficianado too which helps elongate the mind.
Do you ever make music other than Drum & Bass?
I have dabbled, but never released anything myself. That may change soon. I have mixed Rock tunes and other electronica for people though.
You seem to stay on the cutting edge of production, always trying out new plugins and gear. What are some game changers that you’ve found recently?
Valhalla Ubermod sounds good on just about anything….It’s almost as good as my hardware Eventide. The flanged delays are excellent. A lot of the Airwindows plugins are also good. I like their Drive and Desk plugins a lot. My DAW, Reaper, is a game changer. The Slate Digital stuff is good too, although his bullshit way of marketing it is almost enough to make me not want to use it.
And what are some of the mainstays that you’ve stuck with for a long time?
I’m a long time DMG Audio supporter, I’ve known Dave since he worked at Focusrite when i beta tested the Liquid Channel. I’ve been involved with ideas and testing since he started the company. All his plug ins are excellent and coded in the most modern and refined way possible. I’ve been messing with Pitchfunk a lot recently, a current favorite. I’ve been a Soundtoys user and tester for a long time too, Echoboy is pretty much on every track i make somewhere. The clipping on Filterfreak gets a lot of use too. Acustica Nebula and it’s sampled eqs is also a very underated plugin that i’ve used for a long time.
You were on the beta testing team for DMG Audio’s Equilibrium. What were some of the ideas that you contributed to the development of that eq?
Band linking and grouping, The Flat-top shape, right-clicking the A/B/H button to flick through snapshots. Theres quite a few! There’s a group of us, we all bounce ideas of each other and come up with the best way of implementing things for all the plugins.
Your productions seem to rely a lot on harmonics and distortion, what are some techniques you use to get that dirty Dom & Roland sound?
I basically just run things really hot into plugins or hardware until it sounds too shitty, then back off a bit…simples. I may put an eq first as its a good way of controlling what frequencies saturate first. Hardware almost always sounds better especially when played loud due to the large amounts of aliasing present when you use anything digital to distort.
How did you find the transition from analog to digital?
Easy…I did it quite early on. So i had a good understanding on what each was good for. I learned how to use Logic when i had my studio at Moving Shadow HQ, it was handy having Rob Playford there to bug with my questions. I think i was on Logic 3 back then. Audio instruments didn’t even exist back then, Logic was like a glorified tape machine.
Analog and digital both have their perks, would you agree?
Yes, anything with distortion involved including compression, analog generally sounds better. But rules are there to be broken.
You were a Logic user for ages and recently turned to another DAW. It’s difficult leaving something that’s so comfortable. What provoked that change?
As i now understand a great deal about how audio is implemented…I just couldn’t keep using Logic. A lot of its code and the way it uses it is not up to modern standards. There are several things i couldn’t deal with any longer. Here’s a few…
Can’t import or bounce anything at the resolution your DAW is working at (32bit float).
Put something on a bar loop and it’s actually a bar loop plus a little bit extra time, whichever cycle setting you use. Useless for fine tuning a groove.
You have to press stop and start to hear any fade changes you’ve made. They removed some very useful things from the Audiosuite menu like sample rate convert.
I moved to Reaper as i could keep all my Logic key commands and most importantly the way I zoom in and out using the marquee tool…it’s very straightforward and easy to learn and set up. I keep coming across new things i can do. It sounds great too, much cleaner and more open than Logic. In terms of sonics, I’d say Pro Tools, Cubase, Reaper and Studio One sound the best, then Logic, then Ableton Live at the bottom. Along with you and other people we did very boring blind tests to come up with this result.
Right now using Reaper, my up and down keys pitch my audio regions up and down in cents using realtime sample rate conversion (like a sampler) so for most simple tasks i don’t even need to use a sampler. The waveform updates too which makes it easier to see whats going on.
I was amazed that none of these essential things were addressed when Logic X came out. I’m very happy now i made the switch a couple of years ago.
Tell me about Reaper macros.
One i set up recently for beats does this: find all transients, split at transients, quantize to grid setting, force legato, apply crossfades one millisecond before transients…like Pro Tools’ Beat Detective.
You have an editor where you can string long lists of commands together and then assign the whole lot to one key. It’s most excellent.
Any other tips or tricks that you’d like to divulge?
When eq’ing things make sure you level match the eq’d and bypassed version…your brain loves playing tricks on you. Louder for some reason always sounds better.
As far as I know, you’re not a big compression user for the most part, is this true?
I like compression for some things like bass, I’d just rather not use it unless i need to. I like to distort or edit to limit my dynamic range manually, it’s more precise.
You put a lot of focus on moving notes back and forth finding the perfect groove. You also have been known to take a groove from a breakbeat and apply it to another, un-quantized. A lot of producers making music for DJ’s can’t fathom the thought of being off the grid.
It’s all in the groove as they say. People often laugh at me for pulling out the sample delay plugin and delaying a hihat by 3 samples, but i have proved enough times that i can hear and feel it. Even drummers that play to a click often play their snares late or rush their kicks…it’s the difference between an average drummer and a great one. It’s no less important for electronic music. Remember we are talking about miniscule amounts here. But it’s enough to feel the difference especially in the context of a track.
You’re responsible for creating one of the most sampled breaks in Drum & Bass history, the “Tramen.” It involved layering several 70’s breaks together to create a new break. How do you feel about layering breaks these days?
I made the Tramen from a few well known old breakbeats, Three of them layered together cut to the same groove. They were all originally off vinyl, I then eq’d them all to do a different task within the beat…break 1: hats bite, break 2: ride groove and sheen, break 3: weight and roll. I still layer breaks if i need to, but would much rather find one that has all the elements i need in it, that way i don’t have to have my brain done in by working out the phase relationship between them as I layer three kicks or snares on top of each other. The three breaks were then squashed/distorted into one by driving the input gains on my analog desk (very hard to get right in digital) then resampled and chopped up again.
So Trace came round to use me as an engineer for a few tracks and convinced me he should use it (laughs). We ended up doing a few tracks together that all turned out to be classics. Sonar and Mutant Jazz Revisited were the first. One of the tracks I left the break clean on its own for a bar, thats where everyone else nicked it from. Everyone from Ray Keith to Bad Company and Optical have based whole tunes around it since. It used to piss me off, but now i suppose i feel honored.
After all this time, you’ve now done your first record for Metalheadz. How was it working on the Jah Remix for Goldie?
After all this time….(laughs). Because my name wasn’t on Ed Rush’s Skylab EP, no one knew i mixed and made it with him back in the 90’s.
Jah was a hard one to do. It took about a year on and off to get it right. I wanted to keep the vibe of the original but update it. It didn’t help that i had none of the samples and had to make most of them from scratch. It was such a good tune i wanted to get it right. Goldie knew i was doing a remix for him but i wouldn’t tell him which one it was. He was blown away once I finally played it to him.
What’s next for DRP?
A single from us is next (Dom & Hive - DMT). An album of DRP back catalogue remixes from my favorite artists in the scene, a few singles from new DRP artists Armando and Xanadu with albums to follow and a few singles from me.
I have a lot of new solo music made too, i just haven’t had the time to release it, i may release some of it on other labels.
When it comes to mixing dance music, the digital eq is the master chef’s blade. The sharper and more precise it is, the quicker you’ll get your job done.
Every DAW has its own onboard EQ but they’re not all created the same (see the‘Use Plugins That Work With You, Not Against You’ blog entry). Once you establish that a digital EQ is error-free and in good working order, then it really just comes down to ergonomics and preference. Beyond that however, some EQ’s go way deeper than others, providing you with onboard tools like M/S and wide arrays of filter slopes, curves and even phase adjustment!
If you’re running an ancient computer, maybe you’re better off with onboard DAW EQ. Nothing kills workflow like a processing error every 10 seconds. But if not, give these a look:
Over the past few years, Pro-Q has become a staple within the dance music production community. It’s super efficient on processing, has a great built in analyzer, nice big interface and sounds great.
In addition, it also has the ability to create unlimited (?) bands simply by double clicking the interface and any of them can be put into M/S mode for treating the mono and stereo information separately. Other EQ’s offer M/S but I’ve found them to be more limiting usually.
It’s no wonder why this one is so popular, ergonomically sound and at no cost to sonic quality.
The follow up to Equality, brought to you by the brains behind Sonalksis plugins, Dave Gamble. Equilibrium is hands down, the next level digital EQ. Nothing like this has ever hit the market til now. Not even close.
The amount of features this plugin has is ridiculous. For people who like to get under the hood to tweak and hack their software, this one is for you. And for people who don’t, this plugin is also for you because it’s optional.
I feel like i’ve barely scratched the surface of what this EQ can do and every time I use it, I find new features. It really deserves an entire article of its own but I’ll try and break down a few of the coolest bits.
First off, let’s talk about digital EQ’s for a second. When push comes to shove, what makes digital EQ’s differ from one another sonically? There have been studies done that prove every digital EQ (without a saturation circuit) can sound identical by copying curves. What sets them apart, potentially, is the phase response.
Gamble has ingeniously integrated an adjustable phase response curve in this EQ. Why? Because when you take saturation out of the picture, phase response is a big part of what gives analog eq its sound and vibe. He’s given us the option to manually alter the phase but imo, that’s far too advanced for most users. What’s great is that he’s created presets that emulate classic analog eq responses (Pultec, SSL, etc).
He’s also allowed the switching from IIR to FIR mode as well as 'impulse length’, both of which directly affect quality. The speed of your computer is important when it comes to raising this parameter. My i7 Macbook Pro can’t run a single instance at it’s highest setting.
Anyway, as you can see, it goes DEEP. Beyond all the geeky details, this thing sounds incredible when set right. And when not set right, it sounds as good as Pro Q. You also get a ton of filter shapes and curves, M/S, every type of analyzer imaginable (spectrograph!), and endlessly customizable GUI options.
This is THE pro digital EQ for 2013 and beyond, highly recommended.
Developed by the creator of the parametric eq, George Massenburg, the MDW EQ emulates the curves of his legendary GML 8200. I personally love this one. It’s simple but sounds incredible (do all digital eq’s sound the same?). It comes in 3 and 5 band instances and the filters go above and below the frequencies most EQ’s offer (10hz and 41khz).
There’s just something about the 80-90hz area on this eq that just knocks. Can’t explain it, coding voodoo.
Super high quality transparent sound but unfortunately available as RTAS/TDM only. If you’re a Pro Tools user, highly recommended.
Hi I had a question about mixdowns after reading your VOL:4 article. -- 1. If you are mixing with a limiter on the master, is your stereo still set to -6db? (But your limiter boosts gain by let's say ~4db)? -- 2. compression/limiters on mixbusses - is it fine to send a premaster with these on the mixbusses? -- 3. If I have nothing on my master out after I finish a mix, master out set to -6db, does it matter where it peaks at? Or do I not want it to peak at a certain volume? (limit or compress)
Volume 2 was a video called ‘How To Set Levels’. Despite the fact that I turned the limiter off on the master, it directly deals with your question and can be applied the same way.