Andrew Scheps - A Conversation About the Future of Audio and The Skills You Need to Be a Mix Engineer - Progressions: Success in the Music Industry

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Unless you're the artist, you are never making your own record, ever.
And you've got to stop trying to make your own record on other
people's music. That's Andrew Scheps, a man that needs little introduction
in the audio world. He's a Grammy winning engineer and mixer who's worked with icons
like Red Hot Chili Peppers, Adele, Metallica, and Jay Z.
Today we're talking about what it takes to have a career as an engineer in

(00:22):
today's recording industry. We hit it all from the technical skills you need. A
parallel compressor should be more compressed than you would ever
put on the insert directly because you only want to blend in
a little bit. But the way to think about parallel. Compression in general is
through the mindsets you must have. The guaranteed way
to not have a career is to sit at home being mad at the career

(00:44):
you don't have. How the current state of production has transformed
mixing. There's no more
tracking. It's all mixing constantly from the very
first track. And what that means is
that mixing is totally different than it was even
five years ago. Why? AI will probably never make lasting

(01:07):
art. But it can't step outside the training set. And if
you think about everything that, like, went, whoa,
what's that? And you got super excited about it. It was either the first time
anyone had done it or it's the only time anyone had done it. And to
always remember when we can. Work on a project with a musician or an artist
that is making the music we wish we could do. What's better than

(01:28):
that? Nothing.
So I look at your career, I see evolution and
diversification, where you've gone from like analog to all
digital. You were early on atmos. You made a plug in with
waves. You taught yourself to code so you could have a stem
printing solution. So you've added skills and you've kind of like, stayed on the

(01:51):
forefront of engineering. Right. So where do you
think engineering is heading from here? What do you think it looks like,
say, five years? It's a really good
question because there's so many things in play right now. Obviously there's
AI. Apparently Spike Stent has just announced there
will be an AI. Spike stent. I just saw that. Saw that? Yeah.

(02:12):
The website is not terribly informative. I have a feeling it was written
by AI, but, yeah, I mean, Spike's a
super smart dude and he's got his son working on it and stuff, so I'm
sure it's going to be pretty amazing, whatever it is. But that's been
the holy grail. Of the tech companies forever, right. Is to do that.
Because the size of the

(02:35):
market for professionals who make records is
nothing compared to the size of the market of people who
like messing around with music. So the
better their results can be, the happier
they're going to be. And that's going to be amazing. So that will
absolutely drive that kind of thing, just like with online mastering.

(02:56):
And, you know, it's already happened, but the mixing has been a tough nut to
crack. And I think it'll be really interesting to see what, what
Spike's thing is about. And I'm sure everyone else is working on it too. So
that's one thing. The other thing is
there's no more
tracking. It's all mixing constantly

(03:20):
from the very first track. And what
that means is that mixing is
totally different than it was even five years ago for me on most
projects because I am locked into stuff that everyone
loves, but it's not because it's tracked that way. So it's not
individual track processing, it's overall mix

(03:41):
processing. And that makes it cool
because some of it's done, but it also means it's
very difficult to back out of anything because other stuff starts to fall
apart because it's all processing that has everything in it
instead of just like, oh, they did something really weird with that one
drum mic, I can do something different with it. It's not that. It's like

(04:03):
that mic got weird because of the 40 tracks of
background vocals. Right. So that's just gonna keep going. And I. But
what's interesting about that is that moves mixing down
into the creators realm, which is
why something like an AI spike stent
is going to be exactly the right product for what's happening.

(04:25):
Interesting, because. Yeah, I agree. It's like mixing is becoming,
in my experience as well, parallels that less is more because so much has been
done before. Yeah, yeah. And just in an
inside out way. I mean, that mirrors mixing when you're working on eight track or
16 track because everything is tracked. I. So that you put the faders up,
you pan it, and that's the record, but it's

(04:47):
totally the inside out version of that. Yeah. Okay, well, okay, so you brought
up the spike thing. I know that came out yesterday. So that dates this interview
or was announced yesterday. I have an opinion about this and
you can fight me on this or agree, whatever. I think that AI and
mixing at first, it causes people
to get better. Like people that are good, that aren't charging a premium, have to

(05:09):
get great to separate themselves. But I think where the problem
lies is the people that are trying to do the
$100 mixes because they just graduated college, they're
not going to be mixing their peers work because their peers can
have spikes. AI or
whoever's AI, do it for them. So does that mean that

(05:31):
down the line, the craft suffers and there's a generation of people
that kind of miss out on learning it? Yeah, I mean,
that's certainly possible, but at the same time,
podcasts, online classes, stuff inside of paywalls,
outside of paywalls, I mean, that's exploded.
So the opportunity to learn is better than

(05:53):
it's ever been. I mean, I had to go to a four year college and
spend a ridiculous amount of my parents money to get a degree
to learn anything. Same.
Now, you don't have to do that. So it's
democratized on both sides of it. But where it's dangerous
is you have to have some foundational knowledge to

(06:14):
learn from anything that happens with the AI tools. I mean, you know,
ozone AI, it listens to stuff, and it fits
eq curves and things, and it's really good at it. But if you
don't go through the modules that it set up and, like, bypass them one
at a time and tweak some things, and then you're like, oh, okay,
that's what they're doing. And then maybe you'll do it yourself. Then you're

(06:36):
never going to become a mixer. It's just not going to happen because anyone can
slap an ozone on there, and I'll do it for a hell of a lot
less than $100 because I'll mix 500 songs a day.
If all I got to do is slap an ozone on there, we're done.
So it's not about that. And I think this is where the
whole AI thing always trips up a little bit,

(06:58):
is that AI can actually be creative,
but it will only work within its training set. And the training set is
hundreds of thousands, if not millions of things. It's easy to think, like, oh, it
listened to ten songs, and, you know, these training sets are gigantic,
but it can't step outside the training set. And if you think
about everything that, like, went, whoa, what's that? And

(07:21):
you got super excited about it. It was either the first time anyone had done
it, or it's the only time anyone had done it. And it could be anything.
It's the big guitars on creep. It's the distortion on Beck
Odelet. That shit had never happened, but
that's what grabs people and really makes them want to listen. So there'll be this
amazing, huge pool of what you would

(07:43):
then consider middle of the road stuff with middle of the road mixes that might
sound really good when you just put them on, but that's not lasting
art, which is really what we're trying to do. Yeah,
well, I think it drives. It drives people to, like, embrace their
preferences even further as opposed to trying to match what the radio sounds like, because
that's what the AI is going to do. So, you know, like, somebody like Chad

(08:05):
Blake is perfect example. Like, you know what you're going to get? You're going to
get Chad Blake's sound. I think that's going to drive more people to make
artistic choices, and that's probably going to be better. I hope so.
Yeah. Yeah. Okay. So kind of along these lines of I've heard you say
that you believe there's no rules, and I agree with that. I think, like,
you've said what comes out of the speakers. If it sounds good, it's good.

(08:26):
But I do think that when you're young and you're coming up, you're coming out
of college, you're searching for rules because you don't know what to do.
How do you think somebody balances that beginner's
mindset with whatever
the rule is? It's hard
because it used to be you would work in a studio and you'd see

(08:48):
lots of different people do stuff, and so you would learn
the rules while watching people do different
things. So, you know,
tutorials are great if they're good.
And I think that it's hard because it's not
free, but stuff behind a paywall is generally better because otherwise they'd be

(09:10):
out of business. So mix of the masters and pure mix and those
companies, especially pure mix for, like, pure educational videos.
It's an amazing resource. Absolutely amazing. So
the thing, like, there are no rules is true, but there are
fundamental laws of physics and math and electronics
and all that kind of thing. And that's never going to change that stuff.

(09:32):
So that's your toolbox. So you
want to know that really well because then you can be super creative. Otherwise,
you're at the point where you're putting ozone on and you have no idea what
it's doing. You'll never learn anything like that.
So you need to be the person who takes the toaster
apart to see how it works, you know, and hope you don't

(09:53):
die when you plug it in. But it is that, like, you
can learn it from a million different places instead of having to get a gig
at a studio, which is great, but you've got to
pursue the knowledge of it. There is no shortcut to being
good. There just isn't. No, no, because it's
the musicality that makes it good. That's what makes people want to listen again, not

(10:15):
because the snare sound is better. Do you think you discover that
musicality as an engineer over time, or do you think some people just kind of
inherently find it quick? Both.
I mean, in the same with the actual engineering. I mean, there were people I
went to college with, and everything they ever did sounded like a record. Like
Neil Avron. Holy shit. I mean, his stuff. First year of

(10:37):
college was insane, and I feel like I'm still not
the level he was sonically when he started. And this
is. It's 40 years ago. It's 40 years
ago this year. Ah, I'm old.
But the creative thing, I mean, that
just. I mean, think about a musician. You know, Coltrane didn't

(10:58):
pop out of the womb with sheets of sound. I mean, there's a progression,
and again, there's the learning your instrument. There's like, he studied scales. For him not
to. We don't have to make Coltrane the center of this, but Eddie Van
Halen practiced a lot, too. You get
more creative as you go. There's some people who are super creative right
away, and for them, it's just about learning the tools to pull it out of

(11:20):
their head. But I have definitely learned how to listen
and learn to find the things that I really want to bring
out because I think emotionally, that's going to make people connect and stuff
like that. When you're busy, you have an opportunity to practice things while you're
making records. You can toss up an extra microphone, do something weird to it. Nobody
ever has to hear it except for you. Are there ways that you think engineers

(11:42):
can practice these days and kind of hone their skills outside of just actually
working? Yeah, because if they have a laptop and
interface and a couple of microphones, they can go make a record.
So you were saying earlier it's unlikely that
the younger ones are going to be mixing stuff by their peers, but they will
if they go to a club and say, I like your band, let's go make

(12:03):
a record in your living room. Yeah. Nobody's
expecting anything at that point, which means
you can experiment and you can be a creative partnership
with the band, trying to find sounds or something. That's cool. And
I think that that's a great way to do it. And I couldn't do it.
I mean, you think that maybe it was better before,

(12:24):
but it wasn't. I mean, if you couldn't get into a studio. I mean,
I bought gear very early on so that I could
do stuff without having to get into a studio because nothing I wanted to work
on had any money. Yeah. Access to technology, I think,
is pretty powerful. And like you said, the access to
knowledge is amazing. Okay, I feel like I know how you would answer this question,

(12:45):
but would you tell somebody right now that they should go to an engineering school
and pay that money in 2024?
It's really hard to say, to be honest. Like, when I went to school, a
four year college degree was a big deal. Like, you had to have one. There
was just no way you weren't going to have that now. So unless you're
in a specific field, sometimes it's hard to argue that a four year degree is

(13:07):
worth it. Like, if you want to be in tech, tech companies hate it when
you've got a four year degree because they want you to learn the way they
want you to do it. Right. Whereas you would think that you would
need an advanced degree in computer science before you could even
start. So that's part of it. And I think
that there are people who get stuff out of schools and there are people

(13:27):
who don't. So I think it's actually
really personal. There are people who can learn really well. And for me,
it was the best thing I could ever have done because I want to know
how stuff works so that I don't even have to think about it.
But when anything happens, I know why it's happening, and I can react to it
and make it do what I want. Some people, that doesn't work,

(13:49):
that's not the thing. So it's impossible for
me to say one way or the other. I agree with. That's kind of how
I think you have to understand what kind of learner you are. Like, if you
are a self starter, more independent, just want to get your head in there
and do it, then I don't think you need to do it. But some people,
they like the structure. They need the classroom. They need that. And the other thing

(14:10):
is, I met a ton of people, a ton of people who do what
I do, and a lot of them are still in the business. I mean, I
went to college with John Paterno and Joe Barisi,
and I just call them up when I got a question, and it's fantastic.
So meeting your peers like you were talking about.
Yeah, that's invaluable, really. But you don't have to

(14:32):
pay a huge amount of money and go to a multi year school for that.
Yeah. Yeah. Obviously, a lot of my closest friends that I trade gigs with, we
all went to school and moved out here in the same few years, and
that's how it is. I want to shift a little bit to more of a
career thing. A lot of the questions that come up on the show are
things I'm curious about or things I'm struggling with, and I don't always

(14:53):
admit that, but that's, you know, you just did. Exactly.
Now everybody knows, but you get to sit down with cool people and ask questions.
Everybody in this industry kind of believes that there's, like, one credit on the horizon
that's going to, like, change their life and, like, make them
an engineer star. Right. And so I've been lucky to work on a few
records that went number one, and every time I have that feeling

(15:15):
where I'm like, this is it. It's going to launch. This is going to change
everything. And, like, as you probably know, not much changes.
Right. So do you have advice to producers or
engineers that think that that's about to happen to them? How. How they should actually
think about it, how they could maybe take advantage of it, build momentum? Like, what
are your experiences throughout your career? Well,

(15:36):
I mean, look, I've been ridiculously lucky, but if you
look at my discography over the last
35 years, because I did go to school for four years, so not 40
years. It's a slow burn. I mean, I had a lot of
assisting and vocal editing and vocal tuning, like, through the boy
band era, I was tuning vocals. That's how I paid the mortgage,

(15:58):
literally. It's terrible. And there were a
couple of things that were a big deal to a
few people, so it definitely jumped me up a little bit. Like 99
problems. That was a little bit of a jump. And then
the Adele record opened up doors to some stuff, but I don't really work in
that genre that often, so it's not like, oh, now I'm

(16:20):
set. But look, that was a
mixing credit on album of the year of,
like, the biggest selling album three years straight or something like
that. So there's one of
those, you know, it's nothing something.
I think what's more important is that there were projects when I

(16:42):
was working on them, and I thought musically, like, this is
my favorite thing I have ever worked on, and if this
doesn't come out, I'm done, like the music business and
whatever, and twice that I can think of just off the top of my head,
they didn't come out. Or if they did, absolutely nothing happened. The promo was
terrible, and there are these amazing records that nobody's

(17:03):
heard of. And so instead of
saying, like, well, I'm going to get out of the business, you
realize how incredible it was to have
two opportunities to work on something you love that
much. Yeah, it sucks when it's not going well. And you
really do think that. And I've seen there's so many people

(17:25):
who are, like, mad about a
credit they feel they deserve but didn't get. Like, you know,
I engineered some overdubs. I should get an engineering credit. And,
I mean, that's the whole, that's a whole different topic about, you know, making sure
you take care of your credits before the project starts because you will never get
a credit that you think you should get if you haven't already negotiated it. So

(17:46):
you can take that question off the list. It's a brutal,
brutal business, and
I am just ridiculously lucky that I've, you know, made it or whatever,
but I'm still scrambling a lot of the time. Like, I'll go without
work and then I'll have some work that I take because I didn't
have work, but it's low paying, but it's taking up a huge amount of time.

(18:08):
And then all of a sudden a gig comes in that actually would pay for
some stuff, but I can't take it because, you know, there's always a juggling
act. I just happen to be juggling the really good balls most of the time.
Yeah. But right after the Adele win, right after the Grammys, I didn't work for
three months. I had absolutely no work. January, February,
March of whatever year that was none. Not

(18:30):
a single paying gig. How'd you feel like that? I
curled up in a ball and I just go and do something
else. One time when that happened, I went and started teaching a class at
UCLA. Another time, I helped a buddy out who had a low
voltage installation thing. He did audio, but also stuff that
made the curtains open and close and tvs come up out of people's beds.

(18:54):
You know, you gotta stay busy. Yeah, look, the last thing I'm gonna do is
complain and whine. I'm lucky.
Yeah. Well, I don't know. I don't think.
I hate to use the word luck. It's more like it
works, but it's building trust with people. It's following through,
it's being prepared when the opportunity comes. And the more you have those

(19:15):
things stacked up in your favor, the more it seems like you're lucky,
because the right people call, you. Follow through, they call again.
Cycle continues. Yeah. The guaranteed way to
not have a career is to sit at home being mad at the career you
don't have. You know, if you want to be in the right place at the
right time, you got to be everywhere all the time. That's the thing. So you've

(19:37):
got to be out there seeing bands, meeting people, joining
stuff. Join Naras, join AE's. Maybe
if you're a little more technical. There are groups in the UK, there's the
MpG. Like, there are groups everywhere where you get to just go hang
out and talk about plugins. Do it. One of those people will offer you a
gig at one point. Yeah. Or just bring you up to somebody. It's like, top

(19:59):
of mind. It's like, oh, you have to just be in the
space. Okay. So this kind of leads to another question I was gonna
throw in somewhere. Is there a limiting belief that you think holds people back
in this industry? I mean, for me, it's imposter syndrome.
I've got a gigantic, colossal, world sized
case of that. I don't know that it holds me back,

(20:21):
because fortunately, while I'm actually digging in
on something, unless it's just a disaster, I don't have it
until I send a mix. But I have it going
into sessions. If I'm tracking, I mean, I've pulled over to the side of the
road and thought about, well, I could call in sick and just go
home, like crazy stuff. And then you get there and you get going, and

(20:44):
it's okay. So I think it's that, for
me, I will do stuff almost like I'm in a trance. It's like, okay,
I'm doing it. Like there isn't a choice anymore. I'm doing it, and then
it starts happening, and then it's okay. It's like, you remember,
this is what I do. This is my thing. So there's that, and then the

(21:04):
opposite of it is what we were just talking about, which is
90% of the gig, everyone assumes you know everything
about what to do, to do your job. So 90% of it
is, don't be a dick, which is my other phrase. It's whatever
comes out of the speakers. And don't be a dick, because you need to be
able to be somebody people want to have in the room, and they

(21:27):
will gladly take someone who's a little slower, who isn't a
dick. That is the truth. And I think
that's something that you learn in the big studios because you
see it. You see when you get kicked off of a session and when an
engineer producer doesn't want you as the assistant again. But I think when you're
independent, you're working in your backyard now and you're just working with your friends. I

(21:48):
don't think people understand that the way that they would if they came up through
the system. Yeah, well, it's also
the sort of studio hierarchy, like in a bigger studio. You learn very,
very quickly that the assistant, if they have an idea, tells the engineer, the engineer
tells the producer, the producer tells the artist. Nobody except the
producer talks to the artist unless it's your 12th record with them

(22:10):
and it's one big happy family. But when it's. Yeah,
when it's you in your backyard, you think that you're always making your own record.
Okay, so that's the lesson to learn, though, is unless you're the
artist, you are never making your own record, ever.
And you've got to stop trying to make your own record on
other people's music. You're making their record. And if you're not present

(22:33):
understanding what it is they're going for, even if you don't like it, you
think it would be so much better. That's not the
job. Yeah, there you go. There's another one. That's good.
They're falling out today. There's a clip. We've got a clip, people.
I can't not talk a little bit of technical stuff with you, although I would
love to just do career all day. I'm looking at my notes. Let's talk a

(22:55):
little bit about adding skills to your repertoire.
Taught yourself to script and created bounce factory. Did
you learn to script to create something like bounce factory, or did you realize the
power that you had when you started automating things and, like, learning
how you could make the computer work for you? It's a little
bit of both where, like, I've tried to teach

(23:18):
myself programming because I'm a total geek and thought, oh, that'd be really cool. Yeah.
And several times in my life, and it never went well because I didn't have
a project. Then I discovered sound
flow and immediately built a couple of things, like tiny
things. I've got my function. My f 19 key
swaps the input state of the bottom track in my session. Doesn't matter

(23:40):
what session's open because that's how I monitor things. The rough or
my previous mix is on an audio track at the bottom. If I switch to
input, it's the live mix. I used to have to scroll to the bottom of
the screen and then click the input button. Now I don't. That took me
about ten minutes. And that's when I knew, like, okay, so for a while, it
was just automating, like, little things, session prep or setting

(24:01):
preferences or stuff like that. And then I started
writing the bounce script that most people who use
soundflow will eventually write, and it'll be very, very specific to
themselves, which is fine, that's great, because that's
all you need. But then I started generalizing
it when I realized how many different types of things I bounce.

(24:24):
And. Yeah, so I learned because I had
to learn because I had to write this app. But there were a couple of
apps before. There's the offset counter, because you
can't offset the minutes and seconds counter in pro tools. So you open up
this counter, put the cursor anywhere you want, hit a button. Now
that's zero, and you can locate to the numbers you just got in the email

(24:45):
with all your mix notes. Exactly. It's genius. So
it was that it started off as that, and then it just turned into bounce
factory, which was two full years of just coding because
Covid. I mean, that's. I discovered bounce factory during COVID So. That
was one of your Covid adventures? Like, I started a podcast. You made bounce
factory? Yeah, yeah, those two things. That's what I did. What

(25:07):
are some of your favorite automations? Music or non music that aren't like the
ones that people know about, like bounce Factory or the
counter? I think all of mine are music
just because I use it when I'm sitting in the chair and I don't have
hue lights and, you know, but I know a lot of people controlling
everything with it. Prep a session

(25:28):
in less than ten minutes now. So
it imports tracks from my template, but it knows which format it
is because I told it. Oh, this is stereo. This is atmos with all the
crap. This is atmos from stems. It imports the right tracks, it sets the
preferences the right way. I can color code and add to groups
in seconds now. Like, it's

(25:49):
basically every single time I do anything that's
bothering me. I could do it again in four or 5
seconds, but I will stop for 2 hours and write a script, and then I'll
never have to do it again. Yeah, I have
an extremely long list of things that I want to build, but I
don't take the 2 hours to build them. But it's a long list.

(26:11):
But you know what? When you do it. It's so
awesome. It is. You're happy because you're
not doing the thing. It isn't just. It's not going to make me sad. It
actually makes you happy. And you're super creative. As soon as it's done running
that script and you can get back to work. Yeah. Like I told you
when we chatted, like a week or so ago, I found bounce factory right

(26:34):
after I had my daughter. And it was all of a sudden,
like, what am I going to do? Like, how am I going to do this?
And to be able to automate something to that level and know that I can
print an entire album of atmos stems, because God knows,
like, somebody's like, we're gonna mix atmos and you're like, I don't want to print
those stamps. I don't have that time. It's just like, I don't know anybody

(26:55):
that fights against automation just makes me a little bit crazy. There's one of my
friends that's gonna listen to this and he's gonna know who he is, but
I just. I wish everybody would just get on board a little bit, you know?
Well, but, you know, if they're happy doing it that way,
then cool. I guess if you're happy, right? Yeah,
yeah. I mean, if it starts impacting your life because he can never get out

(27:15):
of the house because he's printing stems or she printing stems 24 hours
a day, then, you know. But see, the difference for me, first
of all, it's that it's exactly what you're saying. You can do other stuff. I
can go watch something with my wife. I can go on a walk. I can
sleep, that kind of thing. But it's also
for atmos. I would have been printing ten or twelve stems per

(27:37):
song because it's such a pain in the ass. And it's a twelve song record.
That's still 144 stems, and that's a
lot. Yeah. I'm printing 40 or 50 every. I
basically print every single thing separate so the audio is
baked in. But if I want to split up the background vocals
into a sphere of background vocals, I have them separate. I

(27:59):
can do it. Yeah. Without even thinking about it. Yeah. And it's not. It's
not a heavier lift for you because it's just you've set it up. Because I
agree. If I wasn't automating my bounces, I would
be trying to do atmos with far fewer stems for my stereo mix
because just out of time. And now you have the freedom.
There's been this thing over the last, I don't know, ten years, where

(28:21):
labels want you to deliver a multitrack. It's like, well, get that from the producer.
That doesn't come from the mixer. But then, you know, none of the sounds are
there. There have been edits done since the producer, like, they sent you a new
vocal and all that stuff. So now I actually have a
multi track where you can just put the faders up and you're really, really close
to the mix. Yeah, that's a big deal for the band.

(28:43):
You know, I get calls from Hozier's
musical director. Front of house. Actually, I think he's front of
house. I don't remember. But he's the guy who emails me and says, hey, I
need some stuff for this song. I would have to open
it up and print specific things, but I just send him the folder, and
he loves it. He just pulls up whatever he needs. He can submix on his

(29:05):
own. And there you go. You eliminate back and forth and extra questions of,
like, can you split these? I already did. Yeah, it's all done.
The multitrack delivery, that sparks a question. I feel like there's a debate on the
Internet. I don't know. Maybe this is just my algorithm. You never
know about the multitrack mix session
being the mixer's intellectual property and

(29:27):
not sending it out. Yeah, I will never give anyone a
session. Absolutely not. But you would print
everything, commit it, and send it. Yeah, and it's not
because I feel like everyone should have it. And they can take apart
my thing and see, but they can't see what I did, and they don't know
what the source material was. So whatever, they could hear

(29:49):
a cool delay effect or something, but
it's because the artist is going to need it. They're gonna need it.
And you can't trust labels to hold onto anything. I mean, you know, witness
Sony setting half their catalog on fire, and
you realize they only have one copy of everything, like crazy. So if a
major's gonna do that, a small label, you know, it's gonna

(30:11):
flood. Like something's gonna happen. And this way,
you have the music, which is the whole point. So
it used to bug me when it was a huge amount of work, when it
took forever. No way in hell was I sending that to you. But
now I'm doing it anyway. Now, if you don't want atmos on your record, I'm
not doing that. So there isn't a multitrack to send, but if I'm

(30:33):
doing the atmos then that's what I want. And so once it
exists, I'll send it. If I mix it in stereo and they're getting
someone else to do the atmos for whatever reason, they don't get 50
stems. I split it up reasonably, but they're getting ten or
twelve, you know, because that's the deal. Enough for live, enough
for sync, no crazy overlaps. I mean, that's, that's

(30:55):
rational. So for you it's not, it's not about an IP thing.
It's about you don't want the session because five years from now it's not going
to work. So take these stems. These are better. Yeah, the IP thing would be
the session and that's because, and this
is another thing where we were talking about how production is now mixing. So this
gets very, very murky. But let's say they've

(31:17):
hired somebody to mix the record and it's not going well, but
they like some of the stuff that person has done and then they hire me
to mix the record. I should absolutely not
get what they've done, I should get what they were given. I agree with
that, but at the. Same time I absolutely have
to get whatever created this ridiculously processed rough

(31:39):
mix that sounds amazing. You cannot send me raw
tracks and then a mix that's got
7000 plugins on it. And I would spend my entire life trying
to recreate what was already done so that I could tweak it.
So I have to get something that adds up to the rough. Yeah,
yeah. This makes me feel so much better about my life. The fact that

(32:01):
you're echoing these things because the battle of the rough
mix sounding nothing like the files is just like, it defeats the
purpose because by the time I match your rough mix, a, it's not gonna be
the same, it's going to be my sonic preferences. And b, I'm tired.
Like, I'm no longer creative. And all we did was get back to zero. That's
a huge, huge job matching it. And I'm not good at matching mixes.

(32:23):
I'm terrible at it. But here's the other thing, is that,
you know, we're saying how there's so much mixing as part of the production. Now
on my phone at the gym, it's just cycling music. Now, I've listened to
enough records that Apple suggests things. It does some really
weird stuff sometimes. Like it'll play the same song twice in a row, like
really. But every once in a while stuff comes on

(32:45):
because like, I've listened to something I've done ten years ago, whatever, and
a song will come on and it's like, wow.
And sometimes I think, like, man, that sounds really good.
And it's because I got raw tracks, but there wasn't a rough mix that was
anything other than those raw tracks. Yes. So
things are recorded with intent. The sounds mean something and they're what

(33:08):
they're supposed to be. But I could do anything, and some of those
mixes are better than anything I've done on projects
where it comes to me sounding great. So you'd think, like, oh, well, that's an
easy gig. It already sounds great. Just fix a couple of things. But
that can take as long as mixing from scratch.
But you can't mix from scratch, which means you also can't, like, you're

(33:31):
having to be creative inside someone else's brain at that point,
which is really, really difficult for me. Okay, Dell's
advocate, do you think that those mixes are your favorite mixes because they are all
your sonic taste, or do you think they're
arguably better mixes? Technically? No, I think they're better. Okay. I
really do. I think that there are.

(33:54):
I mean, part of it, obviously, is because it's my taste, so I like
them. Right? Yeah. I mean, it
has to be that, because what's the other filter? I'm listening in. But I
also do think that's why they're better, because I
had. I had control over everything. So it isn't that,
like, I couldn't make the vocal sound better on a really mixed

(34:16):
mix. Like, sure, that's my job and that's what I'm going
to do. But I can't shape everything around the vocal.
I can't get into the mid range the same way, so I can't make
things have separation or space if they're already smashed
together, because then things start to fall apart. So,
yes, it is because it's my filter, but I'm

(34:38):
right. I'm definitely not right.
Oh, that's good. That's good. Okay, so let's do
atmos for a second. We keep touching on it, right. My
perspective of atmos at the moment is that it's
a label artist game, and it's a little bit more about pleasing apple than it
is making art. But you are doing significantly more atmos than I

(35:00):
am. Do you see artists and producers getting excited?
And I should say that I love it and I think we should do more
of it. Do you see artists getting into it? I guess is the
question. When artists hear it on
speakers, their minds are blown. If it's a good mix.
There are so many bad atmos mixes. I mean, there are more

(35:22):
ways you can screw up a song in atmos than in stereo.
Absolutely. You can pull stuff apart, and now the groove is gone, the song is
gone. Nobody wants to hear it, so that's a problem. But
when it's good, and by good, I don't mean it sounds good. I couldn't
care less if stuff sounds good, but it feels like their song.
You cannot do that in stereo. You can't. It's

(35:45):
impossible. And there is something emotionally
that just can destroy you if you're listening to
a good atmos mix that you know you'll enjoy
it in stereo, but you cannot get across what happens. And there are
actually people doing doctorates on why that's the case. Interesting.
This isn't just like, hey, I'm in a room, I like atmos. Like, this is.

(36:07):
It's clinically proven stuff,
but it is a format for speakers, so
that's the thing. And 99% of the consumption is
on headphones. So that requires the ability
of the listener to externalize, to hear something as if it's coming
from elsewhere. And I can't do it

(36:30):
like four or 5 seconds a week, things will explode
out while I'm working on the headphone version of a mix. And then it
collapses again, and then I'm really sad. But
it's not just one service. I mean, it used to be that
people said, oh, well, Apple will bury your album if you don't deliver atmos.
Well, no, they have a lot of playlists that are

(36:52):
immersive only. That's true. Obviously, you
can't be on that playlist, and they're pushing the format. It's not that
they're dumping your record down to the bottom of the bin, it's that your record
can't go on all the stuff they're putting on the front because that's what they're
trying to promote, and now they pay more for it. So there's a little bit
of incentive. I mean, I have no idea what the numbers are, but.

(37:13):
But the format itself, like, talk to anybody who has mixed
atmos. And they just love it.
Absolutely love it. The problem is that people that don't like it have never heard
it in a good room, and they've never got to do it on speakers. And
I just. I feel like. I wish there was more focus on making the binaural
experience better, because then you would bring more people on board.

(37:34):
Well, but, like, one of the things is, if you don't, we can go
geeky for a second. If you don't have a personalized HRTF,
it doesn't matter what they're doing on the encoding side, it's not going to work
well for us it might because you're using a
generic HRTF which is based on something and it's based on
white northern western hemisphere. It's not

(37:56):
demisphere men. Okay, so we have
a better chance of it working for us, but asian women
have absolutely no chance of it working without having a
personalized one. So finally, with phone true depth
cameras and stuff like that, you can get a reasonable HRTF and that will
help. But the good thing is, is that consumer

(38:18):
speakers are getting better sound bars that will hook up to a couple speakers
behind you and a subwoofer. Now all of a sudden you've got a five one
system with up firing tweeters and you can start to get the feel of
it because it's not like, oh, the speakers have to be in the right place.
You just need to feel like you're in the middle of whatever the hell's
going on, instead of it just being in front of you like a television. Yeah,

(38:41):
yeah, I. That's the thing that I'd never gotten into binaural is I'm one of
those people that I don't hear anything behind. I think that's a thing for
some people, right? Some people can. They get the rap and I don't, I just
get the room directly behind. You is actually almost impossible.
Okay. Once it's off to the sides, you should be able to localize a little
bit. And I can't, but it's, I mean, back to the seventies,

(39:02):
there was a, you know, binaural sound effects record or something like that.
And it's because everything is completely
balanced. Delay, frequency, all of that. Just like if
it's right in front of you. Right. So it's almost impossible to hear
that something is behind you. But as it's moving, that's the thing where
people who are lucky get to hear it moving, and I don't, it sounds

(39:24):
weird. Have you. I've heard, I haven't heard it. The Sony
360 tech, I've heard that. That is actually pretty amazing in headphones. Have you
heard that? I've done a lot of Sony 360 mixes. It's a different
headphone algorithm. Again, I never
externalize, so I hear the worst case version of it.
I'm hearing the encoding. I'm not hearing a binaural

(39:46):
mix. All right. Yeah. Anyway, I don't
know. I like atmos. I hope more people get on board, but
everything you just mentioned is, I feel like, the barrier for why
people. Hate on it and for people like us
who can make it to the west coast in January. Go to Namm,
because PMC has a booth. Neumann has a booth. There are probably

(40:09):
others where you can just sit there and listen to people's atmos mixes all
day. And if you don't walk out of there at least wanting to try it,
I'd be surprised if you like music and listening to
music, it's like, I am so excited to play some of my stuff
at NAmm because my room is tiny. Like, it is
literally the smallest room you can put an atmos system in. A

(40:32):
friend of mine who I went to college with works for Dolby, and he was
visiting, he's like, let me measure your room. And he's like, yeah, you've got a
couple inches to spare. Yeah. So I can't
hear the separation in the speakers at all. Right. Right.
But I get to Namm, put it up in one of those booths, and
it's just awesome. That's amazing. And the

(40:53):
PMC booth is just stacked with speakers. It's like, more speakers and chairs in
there. Yeah, yeah. It's amazing. It's crazy.
It's amazing. Okay, I wanted to ask you. We'll step away
from atmos parallel compression. I feel like a lot
of people, when they're starting out,
they're using parallel compression because they hear engineers like yourself talk

(41:16):
about it, and they just take an 1176, they hit all buttons in,
they pin the meter, they push it up until something sounds bigger, but they never
think about that compressed signal. So can we talk about the actual
compressed signal? How are you approaching that compression
and attack and release? And what are you thinking about? Well, you're not going to
like my answer that much because it doesn't matter what it sounds

(41:38):
like on its own at all. Like, not even a little bit. Okay.
Because, like, one of the drum parallel crush things
in my template, and it's been there since the very first time I
was doing anything in the box, which is before I had a console. It's
Fairchild plugin. Right? Right. You put that on drums? Sure you
do. So it's Fairchild. It's set up kind of like it would be on

(42:00):
drums. If you listen to it on its own, it's pretending to be a
fairchild, which means it's pretending to have low end and be
slow and whatever. When you blend it in as a parallel thing, stuff gets
brighter. Yeah. Okay. So I don't know
why. So there you go. Okay, here's the
thing. A parallel compressor should be more compressed than

(42:22):
you would ever put on the insert directly because
you only want to blend in a little bit. But the way to think about
parallel compression in general is if you put an
insert compressor on something, it reacts to the loud stuff right below
the threshold. It leaves everything alone above the threshold. It turns it down
a lot. When? Well, if you're compressing a lot

(42:44):
at a high ratio, when you've got something
absolutely crushed and you're blending it in,
you still have your uncompressed signal. So you've got full range with
all of your transients. All the stuff above what the threshold would have
been completely untouched. And then you start
slamming this slab of the same sound, but

(43:06):
doesn't have the dynamic range, and you're blending it in.
It's not exactly this, but technically you're making the quiet
stuff louder instead of the loud stuff quieter. And the
character of that is what I like. It has nothing to do
with. It's bigger if you do it one way or whatever. I
love the sound of an uncompressed vocal. I love the sound of

(43:29):
uncompressed drums. The transients on uncompressed
kicks and snares are like, they surprise you. They're so
awesome. You start compressing, the drum kit will sound cool, you'll get some great
noise going on. But the kick and snare don't make you laugh when they're way
too loud. Like that's a thing. There's some emotional thing to
really getting pummeled with the transients. Yeah. So you use parallel

(43:51):
compression to get the drums so that you can slam a bunch of guitars over
them, but still hear them. But you still have your transients,
so that's why. But to be fair,
over the last year, year and a half, I've gotten to the point
where I am using almost no parallel compression anymore.
Wow. Almost none. The rear bus hasn't made

(44:13):
an appearance in a mix in probably a year.
How are you getting that same quiet stuff
up? I don't know.
I mean, part of it is if I'm mixing something that has already
been mixed, well, then I don't need it. There's no room for it

(44:33):
anyway. That's also true. But even on the stuff that isn't like that,
it's not about technically what I get at the other end.
It's that I no longer like the sound of it as much. Right. So
mixes are taking me longer because I'm not just like ooh, cool.
That is my starting point, because I really do have to do everything
individually, which I'm sure part of the Internet will be very happy, because

(44:56):
apparently putting anything on your mix bus is cheating. I didn't realize that,
but everything I do is bespoke. But you're
silly if you don't have a template of some sort, because it's, like automating
stuff. You're going to be building the same stuff over and over and over.
But, yeah, I just don't like the sound of it as much. And my
mixes are a little quieter because of it, which is crazy for me because I'm

(45:18):
the loud guy. Yeah, but parallel compression,
don't spend a huge amount of time listening to what the parallel thing is. Spend
a huge amount of time listening to it as you blend it in as a
parallel thing, which I think. Is the best
advice all around for all mixing people are like, what's your mixing tip? I'm like,
when you're turning the guitars up, listen to the vocal. Like, listen to

(45:40):
everything else. Like, yeah, you're right. Solo, we all. We all know solo
doesn't always help. Yeah, I mean, it's good to find out, like, what it is.
Yeah, but then it doesn't matter what you do until it's back in context.
So. Listening to everything you said, I translate as faster
attack, though, because you. Do you want to kill those transients in those parallel tracks
or you let those live? No, because it'll actually double them up. I mean,

(46:02):
compressors in general. I'm slow attack, fast release. Okay.
Okay. But, yeah.
Before we get into my closing questions, I wanted to ask you. Well, first I
want to say thank you for always sharing so much. You're on so many podcasts,
you share so much with the engineering community, so thank you. But is there a
question that you always wish somebody would ask you?

(46:27):
No. Perfect. I mean, I could make a
joke about, you know, ask me what I want for dinner.
I don't know. I don't know. I mean, I'm lucky that I do get to
talk about this stuff all the time, and I do mix with the
masters seminars. Like, that's one thing the Adele record got me. They got in
touch and wanted me to do a mix of the masters seminar, so I've done

(46:48):
13 of them now, 14, something like that, and I just did one a
few weeks ago, and I'm basically
talking for a week, so I get to
say everything. But what's great about it is you put your process
into words, and some people, like oh, don't inspect the process. But I
like to inspect the process because not only do

(47:10):
you have to articulate, you don't have to say why you do it,
but you say what it is that you're actually doing, but also because I do
them every year I see the stuff that's changed and
obviously changing from the console to the laptop was a big deal,
but there's so much that changes every single
year about how I think about things and what I do that it's

(47:33):
great, but it means that I've managed to say everything
at least a thousand times. I mean, is there anything you feel
like no one's asked me? I mean, no, that's one of. When I was
like, oh, I got to reach out to Andrew. He's on my dream list. And
I was like, like, what the fuck am I going to ask him? Well,
see, okay, so here's what we'll do. We'll take up that time with me

(47:53):
saying that everybody has to go listen to low roar, the artist low roar.
Which I started listening to this morning, sounds amazing. So there you
go. I believe that right before this podcast,
I may have recorded the last thing on the last
album. So cool. Amazing. Very depressing
in a way. If you don't know what I'm talking about, you guys can look

(48:15):
at the Internet and see what's up with loroar. It's very sad, but
this album is now almost done. That's great.
I just thought of a question that no one's asking you, probably because this question
doesn't exist. I meant to ask it earlier, so I'm cheating a little bit, trying
to figure out how to get this question in. Back to AI mixing. What do
you think about the idea of this is basically what Spike presumably has

(48:36):
done from what we've readdeveloped. Training an AI on your
own work to then operate as an assistant for yourself.
Yeah, I think that that's
great. But I think again, and I'm really
curious to know what they did with Spike's stuff.
To get a good model, you need

(48:59):
hundreds of thousands of things that it learns
from 10,000 things. I mean, Spike
has mixed a metric fuck ton of records, but I
don't know that he's mixed hundreds of thousands of songs. So is it
enough? Maybe it's been supplemented with
stuff he likes. You know, I don't know. I don't know what they're

(49:21):
doing at all. But, like, for me, what the
AI is amazing for is separation of things. Like the
low roar record. Most of it was tracked with Ryan at a piano,
singing. But I needed to do stuff with the vocals. And
I actually managed to split the bleed out of both the
piano mics and the vocal mics so that I could treat them

(49:43):
completely separately. And if I wanted the vocal bleed
from the piano mics because it was part of the vocal sound, I have
it separate. But it's not in the piano mics.
Yeah, that's huge. And that's AI. But that's been
trained on probably a couple of million recordings of vocal
and piano to be able to do that. Now, did you do that? Was that

(50:05):
with RX or was it with something else? No, that was with
Steinberg spectralayers, which at the moment, if you're not a
total geek and running stuff in these Google collabs
online, by far the best piece of software I've used for
this stuff. It can be a little clunky, but when it works, it's
insanely good. And its tools are crazy.

(50:26):
Absolutely crazy. There's one, they're like weird
magic Photoshop select tools where one, you
can drag it along what looks like the fundamental and it will
grab it and all the harmonics. So you can grab a
piano note with all of its harmonics and pull
it out manually and copy it to a different layer.

(50:47):
And when it separates things, if you don't mess with it bit for bit,
it goes back together. There's zero audio loss. Every bit is
somewhere. That's wild. It's ridiculously cool. I mixed a
record where a lot of the drums were recorded, where the
overheads weren't cymbal mics, they were absolutely the kit and that was
a big part of the drum sound. But on a couple of the songs there

(51:09):
was some symbol that was crazy loud and I needed
to turn the cymbals down for 16 bars, split
the overheads into cymbals, kick, snare, turn the cymbals down.
What? And the cake and snare don't change at all.
And this is for stereo mixing. A lot of people talk about the demixing to
make yourself more source material for atmos or whatever. Spectral

(51:31):
layers might be the best money I've spent in years.
It's incredible. That is. That is amazing. Well, I'm
glad that we pushed that question into the end. Even though
I almost let it go. I didn't answer your question.
I mean, look, the idea of being able to train your own stuff is
incredible. None of us have enough material to make it actually work

(51:54):
very well. I never thought about that thing that sucks. Like with the Beatles
stuff that they did, separating
the things so that they could do the Atmos mixes it is
a much more specific data set because they had the recording.
But then if they wanted to split the bass out, they had every single
performance of that bass part. So

(52:16):
it knows what order the notes are coming in, and that's how they trained it.
So the Peter Jackson thing is trained with every,
like, that model only knows about that song, period.
That's nuts. That's it. But it can pick the bass
out because it's always Paul playing the bass, same
line, everything's the same. That's wild. Do you think that Peter Jackson

(52:37):
Tech will ever enter the world?
I don't know. I don't know. There's always talk of it. But again,
I think that it's not generalized tech,
it's specific. So you'd have to go train on how to train it
to get anything done. And how many of us have datasets that
are the right thing? And I'm going by hearsay. I've never talked to

(53:00):
anybody, but I've heard that that's what it does.
They trained it by other performances of the same part.
So what do you have as a job where that would
work? Nothing. I've never in
my life had something where that would work. That's
true. So, you know, that's amazing. All right, well,

(53:22):
I know you've got to get back to your day. I've got the last two
questions that I hit everybody with, which I was. Supposed to look at, and I
didn't. Well, if you didn't, it makes it more fun.
Was there a time in your career that you ever chose to redefine what success
meant to you? Well, we already talked about it, actually. Those
projects where it was like, if this isn't huge, then everything's wrong

(53:43):
with the world and all of that. And, yeah, I've tried, and
it's really difficult, but I have tried
constantly to make the success that I
get to work on some stuff that I still want to listen to when I'm
done that I'm actually really proud of,
and that's hard, but it's, you know,

(54:04):
when it happens, it. That's what. That's why we're doing it. None
of us are doing it to get rich. You know, I bought a modular, so
I'm doing all right. But, you know, you're not doing it to be on MTV
Cribs, which anyone under the age of 30 won't even know what that is.
We're doing it to make music that we love. And
we're in a position, the people who do what we do, where we can't

(54:26):
make the music ourselves. So when we can work on a project with a
musician or an artist that is making the music we wish we could
do, that's the best. I mean, what's better than that? Nothing. Nothing.
Nothing. I think people think that when you're at the top of
your career that every day you're doing, like, your favorite song ever. But
that's just not. That's not the case. Oh, no. Absolutely not. I

(54:48):
mean, I'm so lucky to have the low Roar project because
for six albums over, I mean, I
work. I've been working with Ryan for twelve or 13 years or something like
that. I've had album after album. That is my
favorite music I've ever worked on. But that's six
records. Exactly. Only six. And there's other stuff I've worked on

(55:10):
I love. I mean, there's lots of stuff I've been very lucky to work on,
stuff that I like, but. Yeah, but the other part is you
do do this to pay your bills as well. But if you can't
take that part of it, then do something else to pay your bills so
that you still love it when you work doing what we're doing. Because if
you start not liking this, what's the point? It's

(55:33):
brutal. It is a brutal, horrible job where everything you
do is completely subjective. You can think you've done the best thing ever
that took you five weeks, and they just go, nope, I don't like it.
Yes, yes. It reminds me of my January.
You're just like, this is the. I love this. And they're like, we're not gonna
use it. You're like, anytime. I really like it. They hate it.

(55:57):
Yeah. I can almost say that as well. Yeah. Every time. It's my
favorite. It's not theirs. Yeah. Okay, so last
question before you go. What is your current biggest goal and what's the next smallest
step you're going to take to go towards it? I really should have
read the questions beforehand. Like, Tony
Maserati has this great thing where he has a yearly

(56:19):
review where he does exactly what you're talking about. Like, okay, so what are my
goals? How am I going to get there? And he's got like, okay, here are
the steps I'm going to take. And then he has a meeting with himself in
a year to say, like, okay, have I done that? Have my goals changed? What
did I, you know, which I think is great. I mean, a short
term goal, just selfishly, is

(56:40):
I'm going to try and get bounce factory out there a little bit more
because I love working on it. I actually love the coding because it isn't
subjective and I would love for that to kind of
cushion. So I don't have to work on records. I don't want to. And, you
know, again, people think you're all day, every day. That is
absolutely not the case. I need to work. I have to work

(57:02):
because I like electricity and water and food.
So, you know, like, everybody, again, in a good
position. So I don't want to get, you know, trolled over this. And then
the next thing is to just try
and find, like, what is the next low roar thing going
to be for me, if there is one? There may not be. I mean, a

(57:23):
lot of people would never even have one, so I'm
super lucky. So I've got, what am I going to
do with the legacy of Loroar? Like, I'm going to be doing a vinyl box
set and things like that, but I don't want it to just disappear. So
I have no idea what the first step is with that, though. That's unknown. So
that doesn't answer your question at all. It's good, though. I'm with it.

(57:45):
Well, I think it highlights diversity, and I feel like it doesn't
matter how long you've done this. Having diversity in
income allows you to do the records you love because the record you
love might not be the one that's going to pay your bills. So if something
else does, whether it's music related or not,
that makes your life better as a musician? Yeah, absolutely. I mean,

(58:08):
I love making the plugins because I'm making plugins that I actually wish
existed, but at the same time, that's a little bit of extra money and
get enough of those things, and it just takes some pressure
off. Yeah. Which is good, because I think all of us can smell a
gig that's going to really suck, and it may have nothing to do with the
source material or the music. It's usually about the people

(58:30):
and, you know, going in that you really wish you could say no. Yeah.
Yeah. Everybody's been there. Everybody's been there. If you enjoyed this one and
it's got you thinking about how you can. If you could better build your sonic
identity as a producer or engineer, then you're gonna want to check out my interview
with Tony Hoffer, the producer mixer behind Beck, Midnight Vultures, the kooks,
inside in, inside out. M 83, midnight city and many more.

All Episodes

Andrew Scheps - A Conversation About the Future of Audio and The Skills You Need to Be a Mix Engineer

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Andrew Scheps - A Conversation About the Future of Audio and The Skills You Need to Be a Mix Engineer