Elephants in the (mastering) room

I was asked to give a talk at the inaugural meeting of the AES Mastering group, last September.  There’s so much I wanted to talk about, but I decided I wanted to talk about a movement, a change that I’ve seen happening in the mastering world for a few decades now.

This is what I talked about, condensed here into some slides and notes. I was talking for 30 min or so, so it’s a long read. Feel free to scroll to the juicy bits 🙂

Click if you want to go faster

Intro

Hi everyone and thanks for coming here today.  I’m going to be talking about the problems I see in audio mastering in the near future, problems that we all know inside, but don’t talk about!

So before we start, I’m just going to give you a few details on my background, so you can see where I’m coming from.

My name is Crispin Herrod-Taylor, and I’m a partner in Crookwood.  We make mastering consoles and monitor controllers, and have been doing so for the last 26 years or so.  We have more models of mastering consoles available than any other manufacturer, and have supplied a wide range of engineers from the elite to the part time engineer.  For this reason, I think I have a unique insight into how the mastering industry has changed, and how it will change.

Before Crookwood, I worked for Focusrite designing the last big format recording console, and for SSL and hi-fi manufacturer Meridian. My old boss, Bob Stewart was giving a talk on MQA yesterday!

I’m not going to give too many plugs for my own kit today, I’m more interested in the industry and where it’s going.  Perhaps this attitude is why I’m not going to be able to retire any time soon!

The problem

So what’s this problem I’m talking about?

 
 

Because we do a lot of semi custom work, I’ve talked to many engineers, and been involved in some great installations all over the planet, and over the many years, I’ve seen the mastering world change.

It’s not really this fellow , the solitary elephant, but rather it’s a herd of elephants.  There are a whole set of forces bearing down on us, I’m seeing a perfect storm of disruption hiding in the corner, but nobody really seems to be talking about it.

 

At the risk of being dramatic, this is what I foresee:

 
 

Ouch!  That’s quite an aggressive statement.  Ok, so it’s partly to wake everybody up in the room, but it also reflects the way I see the industry moving, and I don’t think the 30% figure is far off the mark.

Before the big reveal of the problem, it’s time for a review of where we are now.

It’s always dangerous predicting the future, when I was young the TV show “Tomorrow’s world” predicted that by now, we all have jet-packs and hover cars, I’d go to work in a concorde and have tea on the moon! Clearly this hasn’t happened, although I do fancy a jet-pack…

To understand where we are now, and where it’s all going to go wrong, we need look at how music recording and the world has change over time. 

The first thing to recognize is that while humans have played music forever (there’s physical evidence of musical instruments at least 40,000 years ago), recorded music is a recently new phenomenon. We’ve been mass producing it for only 50-60 years. Before records, the only way of mass distributing music was with sheet music – which incidentally is where we get most of our antiquated copyright laws from. Prior to this musicians were mostly poor and made their money from performances. Sounds familiar?

I’m going to look at some key trends over the last few decades of music, so we can see what’s been going on.  We inherently know most of this information, but have just filed it.  It’s only when you see it together do you start to see where it’s going.

Trends

The barrier to entry has fallen

By this I mean how easy it is to setup as a mastering engineer.  The entry barrier has been lowered because kit is more affordable.

As a simple illustration, the price of a VMS 70 in 1972 was about £50k.  In today’s money that’s the equivalent of £750k!  How many of you could justify that outlay just to be in the mastering game?  (Thanks to Paul Gold in NY for the original price lists!)

And it’s not just hardware, when ProTools was released in 1991, it cost £5k for 4 tracks, that’s about £11k in today’s money! Everything has got cheaper, more affordable, and open to all.

Kit is also a lot easier to use.  The original kit was difficult to use, fragile and unstable, everything was done in real time, you needed to be a trained engineer to use it, and you needed a high level of trained support staff around you.

Today everything (mostly) just works, and you just need basic computer skills to do anything.

The number of music creators has increased

We’re talking about artists here.  Like us, their barriers to entry have fallen, and they can record & mix very easily and cheaply.  Prior to the mid 70’s if you wanted to release a record you needed to be very wealthy, or signed to a label.  Today you can upload to Spotify for free.

This has led to an explosion in music creators, most of whom are keen amateurs, or wannabe’s, and very few who are making a living from music.  Their expectations and budgets distort what you can do and charge for a job.

The number of large studios has decreased

When gear was expensive, and needed looking after, it was small change to employ staff.  You’d often have a mastering room linked with a couple of studios.  That’s all gone, mastering is a solo job, and not all engineers use purpose built rooms.

Revenue for production has decreased

As the cost of recording has gone down, so have the prices on everything else. I don’t have any specific mastering figures, but when I was working for SSL in the 80’s a big room would go out for at least 2K a day, that’s 8k a day in today’s money! Try charging that now and see how far you’ll get.

Looking around the internet, you can get on-line mastering for £30 a track, that’s perhaps £45 an hour.  You can automated mastering for £5 a track (more on this later).  Yet it costs £100 an hour to service your car at a main dealer. Audio is cheap.

Audio quality in the studio has gone up, and we have plateaued

Everything we do now is to a high reproduction standard.  We don’t have to worry about tape wear or bleed through, our amplifiers are clean and our speakers are reliable and tuned properly.  Digital makes our life very easy and consistent.

We are now chasing ever smaller quality increments in every field apart from speakers.  As an example a mic pre I designed in 1992 has never been bettered, some 26 years later.  You can argue about features, looks, and everybody has preferences, but the fact remains that I got the design right and no magic new technology has appeared in the quarter of century since I designed it. This is perhaps obvious for analogue, but the same is becoming true for digital tech.

I’m not counting gear purposely made to have a “sound” here, that game will live forever.

Audio quality for consumers has gone down

In contrast, peak quality for consumers was with CDs –a perfect consumer audio distribution format, whose sales peaked in 2000. Since then consumers have opted for compressed audio like MP3s (first made available in 1992) as their preferred format. SACD (1999) and DVD-A failed, Flac and other non lossless formats haven’t disrupted the market

One of the biggest red herrings is High Res audio. This is great for the labels, forcing people to pay again for what they own, but the public don’t need it. They are happy with their MP3s. Manufacturers of players are driving HR Audio, not the punters.

It’s not just the delivery formats, when I was a student the first thing you spent your loan on was a kick ass stereo.  You had 2 speakers and they were properly spaced.  Today most home audio is listened to in mono via small wireless speakers. The nearest they will get to stereo is a set of badly tuned headphones, which is not how we make music to be listened to.

Since 1979 Sony invent the Walkman, portable audio and headphones start to be considered normal. More music is listened to on-the-go, in poor environments than sat down in a listening room.  Who even has a listening room today?

We can argue about this later, but the consumers just don’t care about audio quality.  They don’t give a toss.  They have voted with their feet and MP3 is what they want.

More Trends

Stuff we’ve seeing today has been around for a while, just not been commercial

Stereo was invented in 1933, but the first real stereo records were released in 1958.

Amazon was founded in 1994, Deep Blue IBM’s chess AI beat Gary Casperoff in 1997

Spotify is over 10 years old, as is Facebook, You tube and Google are nearly 20.

The point is that a lot of what is coming to pass today is based on old tech.  It works, we’ve just had to wait a few years for it to become affordable or widespread.  Because it’s not new tech, once it gets traction, it can grow very, very quickly

Technicial audio education has gone down (apprenticeships)

Time was, to be an engineer, you needed to be a technician, and you probably wore a lab coat to work.  You were educated on the job, until you had an absolute knowledge of the gear, its idiosyncrasies and ditto with clients.

Now we are either self-taught or you spend a few terms (and a few £k) at an audio school learning general techniques in a classroom.  Not the same, folks.

Number of mastering rooms has increased

As the cost of building a mastering room has come down, and the budget in other areas of recording has fallen, mastering is widely seen as the best place to make a living from music.  You will probably disagree, but the number of rooms and competition has grown wildly since the 90’s.

Price of records has come down

Records were expensive – an LP cost the equivalent of £30 in 1966 – worse than that, there wasn’t a minimum wage, so this was the equivalent of 25% of a teenager’s wage.  Singles cost equivalent of £8.50.

Today, iTunes (RIP) costs 99c a track, and my Spotify family is as low as 0.04p a track.

Recorded music is cheap, disposable.

Number of listeners has gone up

On the other hand, the number of listeners has increased massively, the growth of streaming and the cultural shift to headphones/earbuds and on the go listening, means we’re getting a significant number people listening to music again. Hurray!

Digital has got very good

I’m not going to get into a bun fight here, but we just take digital for granted now.  Once analogue’s been converted, the storage, and transmission is perfect, and the DSP is pretty damn good.

And it cheap, really cheap.  Anybody can afford an edit station. Anybody.

Mastering reputation has gone down (loudness wars) – we’re blamed for this.

Somehow, we’ve taken the blame for this, not the labels, not the artists, not the media, but the MEs.

The generations of care and practice that’s been put into mastering has been/ is being wiped off.  Suddenly we’re just middlemen.  Some of this is valid as there are so many people calling themselves MEs, all trying to get a slice of the pie, most of them without the training or experience to do so well. On the face of it, it’s very difficult to look at a ME website and tell who is real and who is not.

And don’t get me started on the varying quality of room acoustics!

So in summary, this is where we today.

 
 

OK, so what’s the massive problem, why will things change?

The Elephants

So let’s name all these elephants sitting in the corner:

  • Music Streamers
  • AI, or more accurately machine learning
  • Crowd funding – new ways of working with money

Each elephant by itself doesn’t pose a threat, but together they’re creating shifts in our perceptions of what is normal, changing markets and values and breaking up the status quo. 

They all have two things in common however.  1) Their financial models.  In order for them to work, they need scale, they need to become massive, and there are a lot of investors willing to take the risk, lose money now and wait to see if they win.

2) The second thing is that are disruptive.  They essentially look to remove multiple middle men, and replace everybody with just them as the only middle man between content and user. 

 
 

From early beginnings, most music is now rented and streamed rather than purchased.  It’s enabled more people to have access to more music, and to find new artists.  For a punter is it is a good thing, and personally I love Spotify, and I’ve discovered loads of music I’d forgotten, let alone new music I haven’t experienced yet.

Money has been pouring into the big streamers, and investor modelling predicts that in the next few year with a few niche exceptions, the only streamers left will be Spotify, Apple music and possibly Amazon, every other one will bite the dust, and many already have.

However for this model to work long term, they need to lower their costs, and this is where they affect us. The obvious way to lower cost is not to pay so much for the content, and while they can bully content providers a bit, the more interesting approach is to remove the labels completely from the process.  I and many others predicted this would happen a while ago, and low and behold we see Spotify trialing this. Artists can now directly upload material to their systems for free, and they will provide free tools to enable this.  The next angle is to let artists promote their work in paid for playlist positioning. 

This should be a concern for us because

  • Artists can save money and get promoted in a one stop shop
  • Most modern music is transient, it exist sonly for a few months before being overwritten. Is it worth producing properly?
  • We are the next middleman
 

The rise of the robots!

Artificial intelligence of course doesn’t exist, and probably won’t in my lifetime, but it’s the tag we all use to describe machine learning, and machine learning is real.  It has now passed a threshold where it’s affordable, and progressing rapidly. The key word here is the learning bit.  Every time they do something they slowly get better.  And they can do a lot of things a lot faster than a human can.

For AI, mastering is the low hanging fruit.  If you were to look at any part of the music production process to automate, you’d choose mastering.  Why?

  • Artists understand recording and mixing, but mastering is black magic. AI is magic too.
  • There’s a small number of channels, 2 in, 2 out. Much less processing power required.
  • Finite number of processes involved, basically level adjustments (EQ, Fader, Compressor), not like music creation or editing
  • They can analyze all existing music for trends and patterns – the end product has no other processing added since mastering
  • Simple processing can produce audible results for the artists
  • Artists aren’t involved in mastering the way they are in recording or mixing ( we don’t like attended sessions)
  • It’s at the end of the process when the artist has run out of money, or is desperate to release the product.  They can provide cheap, rapid (instant) mastering 24/7
  • The artist can iterate the mastering process, feel in control and feel creative
  • The consumer mostly doesn’t care about quality

The key area AI is focused on now is artists – content providers who don’t have access to studio tools.  They want to enable them to produce good results without spending a lot of money or having to learn how to use technology. A lot of their focus is improving the quality of youtube type recordings, and let the artist iterate the mastering until they are pleased with it

This sounds great doesn’t it? It shows how important mastering is and it will benefit us all. Except that:

  • Although poor now the machines will get better, and will threaten human mastering
  • It sets the low end price for a track, now $5, and will start a race to the bottom, one only they can win, their cost being fractions of a penny
  • It gives power back to the artist, who sees mastering as a creative process they should own, not one they should abdicate to an ME
  • The idea that as the grow, they’ll ditch the machine and come to us (the trickle up effect) just doesn’t happen in the numbers we expect
  • Audio is a commodity, customers don’t care

Furthermore the long term ambition for the AI guys is to build in copyright and revenue tracking into their machines, so artists can get compensated. A sort of 21st century watermarking.  To get the critical mass to let this happen, they are looking to provide APIs to the tech so that other manufacturers can add the features to their products.  This is not just apps, but platforms and hardware too. 

If the next audio interface an artist buys has an auto master feature built in, so they can hear the final product while they work, if they’re happy, why sent it anywhere else?

 
 

The final nail is Crowdfunding and alt finance.

It’s great. I used it to finance and market prove our SoundBucket speakers.  It lets you communicate with an audience, progress your ideas and market test your offer.  For artists this opens new ways of working.

Simply it lets artists raise money to finance their work without having to re-mortgage the house or deal with labels, while building an audience.  If they are serious about moving forwards they recognize that they need to promote themselves, and have contact with their audience. Crowdfunding lets them get an advance off their future sales to their audience.

What’s wrong with this? Well,

  • As the artist gains more power and authority over their work we’ll have to deal with more individuals rather than a few labels, how do we reach them, how do we establish relationships?
  • Crowdfunding typically generates less money than other methods and teaches the artist to be more responsible, and watch the pennies.  Less money for us.

So what to do?

Solutions

 

Strangely, the solution isn’t to go and buy lots of Crookwood gear, although if you wanted to, please don’t let me stop you!

The solution is actually a really boring, non audio one.  Go back to basics and learn about how to run and market your business.

Most of us got into audio engineering through love, not money, but you are now running a business, and if you don’t attend to this, you will go under.  Go back to basics. All the way back. 

 
 

You need to know, in great detail, what makes your business tick, why people choose you, and why others choose your competitors. You need a full market audit.

  • How are you doing – how much biz do you have, what do you charge?
  • What are your direct and variable costs?
  • Where are your competitors, what do they charge, what do they offer?
  • Where do you get your customers, how do they know about you?
  • Are the number of customers increasing, decreasing?
  • What’s their demographics / psychographics?
  • Would they miss you if you disappear?

And so on.  There are plenty of web references around about how to perform a marketing audit, so if you’re not sure what to do, use these resources to start a plan.

When you’ve got a handle on this

 
 

This for me the central reason for doing what you do.  You are not offering mastering services, you are offering a kind, honest ear to your clients work.  This empathy with what they want to achieve, together with honest feedback and praise for the artist’s work is uniquely human and individual.  No machine will ever be able to replace this.  Simple mastering yes, but human interaction?  No chance.

And with more and more artists working without professional help, you may be the first & last music authority they meet. Understand their needs and solve their problems.

Your choice of room, hardware gear, lounge and everything is a human element, as much as you are.  Your opinions, your conclusions, irrational or not.  This is human, and machines don’t stand a chance.

Bland, vanilla, same as everybody else, ITB, faceless mastering is just asking for Skynet to take over.

 
 

USP stands for “Unique Selling Point”. The thing that makes you different from everybody else.

Not many people are really offering a proper service, so you could concentrate on just that, however there are still enough people that you need to offer something different. 

It’s better to provide 100% satisfaction to a small number of people, than 50% to a larger group.  By focusing on a small group of people, you can provide exactly what they need and get a name for yourself.

Although it’s unusual, you could offer a niche in music genre – you are the man to go to for Mongolian throat music, or wider for all dance genres. 

Your Niche could be in turnaround time, streaming and CD masters for the same price, location, packaging, delivery, label type, quality, anything.  Just something you can say, is unique to me.  And then find people to whom this is important.

 
 

There are two points here.  You need to become a brand.  Not another faceless mastering studio.  When people think of Debbie’s Master’s they need to think of what Debbie does for their music.  This brand gives you leverage and positioning. Make sure you get mastering credits, offer discount to ensure it if needs be. People need to know about you.

Secondly, put your prices up!  For Pete’s sake please put them up!  I know there’s pressure and everybody around you is doing deals, but you have to have pride in what you do.  You are worth the money, or why are you doing it?  All around you inflation is making things more expensive.  You can’t do a deal with your food, energy, rent, transport etc, but they go up by inflation each year relentlessly.

As an example, if you charged £100 for a job 5 years ago, typical inflation means that you should be charging £110 today.  Put your prices up a bit every year.  Put them up now.  Just do it.

If you feel nervous, try it on a few customers, or offer discounts for prompt payments, credits, or slow Tuesdays. But don’t start off cheap. Please.

 
 

With all of this extra income you’ve got coming in ( 🙂 ), you need to do something with it.  You cannot stay still, the world is moving too fast.  So allocate some time every day to plan and better yourself.

Read books, check online, enrol in an online course.  Perfect your room, so you can perform to a higher standard, and invest in efficiency, so you can do more with the same amount of time.  Look at your back office, and look at your gear, do more for less.  Small plug: Crookwood stuff lets you work much faster than before.

 
 

As a breed, we’re pretty useless at talking to each other.  We’re loners in our little rooms, working by ourselves.  But these changes are affecting us all.  This AES group is great to meet people and listen to their issues.  Perhaps somebodies already solved the problem you face?  Join Facebook and Linked in groups.  Join forums.  Go to AES meets. Talk to people and learn.

And finally

 

I started doing all this because I love music.  Don’t lose sight of what makes life fun for you, don’t let it be a chore.  Life is exciting, music is exciting. 

Enjoy it. And thanks for listening.