Interview with mastering engineer David Else of Digital Media Duplication (www.digitalmediaduplication.com ) as told to Chris Garvey at a3music, March 2006

First off, what exactly do you do to mix's that clients bring you (the process)?

The main goals are to obtain an appropriate frequency response for the the genre of music, and to increase the loudness to a commercial level. Other important tasks include finding any key problems and trying to eliminate them.

For an example of frequency response, drum and bass needs a very different EQ to classical. You can only make sound judgments on EQ with high quality flat response speakers and an acoustically treated room. Every room has standing waves that effect mostly the bass response of speakers. Someone maybe hearing what they imagine is awesome bass, but it might well be the room tricking their ears. You can try using frequency analysis software or high quality headphones to attempt to compensate, but you need to be sure by getting it mastered properly.

Many mixes are now done in peoples bedrooms rather than professional studios. With the advent of high quality plugins and DAW software like Cubase SX, Logic and Pro-tools anyone now has access to the tools to make full quality commercial music! We are in the age of the bedroom recording studio. The only downside about this is the musician is now expected to be a recording and mix engineer as well. Often great music is let down by inexperienced mixing. Mastering can often help a lot in these cases, but the real cure is to get a professional to do the mix in the first place.

There is a lot of moaning going on from professional mastering engineers about the so called 'loudness war'. This is the problem that people want their music to be as loud as humanly possible to be heard above everyone else's. The current trend of extremely compressed music is argued to be at the expense of musical fidelity and dynamics. Often tracks loaded into a sound editor will look like a large square rectangle as they have been so highly compressed. Great music from the past never had this treatment, and many people say it sounds much better for it. For example, load an ABBA song into a sound editor and have a look... nobody can say they were not a commercial success!

I will ask a client if they want their music to be as loud as possible. Most people do. I can do my best to get it as loud as possible while minimizing the side effects, but I would not be happy limiting a song to death like some engineers have done.

What's an example of something that creates problems for you when mastering a tune?

First problem, people seem to insist on sticking their mixes through waves L1 limiter before they send it to me! However much i tell people, it seems an addiction out there! You should leave all that up to the mastering engineer. The mix should not be clipped in anyway, but as loud as possible with no limiting on the mix bus or put on afterwards

Some things are obvious and can be fixed right away with no side effects. Other things are tricky as they involve trading things off against each other. For example, last week I had a song that had a very muddy acoustic guitar with a quite thinnish sounding voice. Cutting the mud from the guitar made it sound great, but then the voice was too thin sounding, so a balance had to be struck. A decision had to be made as to which was the greater evil and a compromise sought

What general techniques do you use on every master you do?

EQ is the first port of call. Sometimes a recording will be near perfect, sometimes they can be really messed up. I did a dance track someone gave me recently that needed about 8dB cutting from the middle frequencies to even sound listenable! Everyone is at different skill levels of engineering, it doesn't mean the song sucks.

It is very rare than recordings don't sound better with multiband compression on the bass. Everything can be made more solid and punchy. Any music with a kick drum and bass instrument can benefit. For dance music it is mandatory. The tick is to find where the best place for the cross over frequency is, and adjust the attack to let the right amount of bass drum through. The release should be as quick as possible without distortion becoming a problem. The wave length of a 60 hz sine wave is 16.66ms. This is often where dance music kick drums live. A faster attack/release than this will mean the waveform itself will be effected. It's important to remember too fast an attack/release for bass frequencies will mean distortion, but sometimes a little distortion is good. Every track is different.

What equipment do you use and what's your favorite bit of gear?

I don't want to give away all the secrets of gear, as this is part of why engineers sound like they do. Also it can bias how people will imagine the sound of the mastering. For example, I might say I master onto a vintage studar 1/4” tape that was modified by Bob Katz to make every bit of music passing thru it into a million selling hit. Now people are amazed! If I said i use a freeware plugin that no one has heard of, people might think it is going to sound rubbish.

It really is not about the tools, but knowing how to do what you want with what you have. I think there are 2 types of plugin. Good ones and bad ones. What they cost is not the main factor. Try Voxengo plugins like HarmoniEQ and PSPaudio's Vintage Warmer. These are high quality. Now try the built in EQ of Cubase SX. This is poor sound quality. What I am trying to say is the process the plugin performs is often subjective, but the bad quality ones screw up the sound.

Is there a reason why you chose digital equipment over analogue equipment?

Yes. It is 10x more convenient and flexible. All the settings are remembered, so I can come back the next day to review my decisions. The sound quality is now as good as analogue, if not better. An emulation of something is always an emulation. However close it is, the original is the real thing. The point is that often 'the real thing' is not what is required in the first place. Digital is no longer the 2^nd rate system.

As CPU power increases you will eventually be able to model any hardware with 99.9% accuracy. It is not quite there yet, but it is so good that most people can't hear the difference. EQ and multiband compressors used for mastering are different from other music software, you usually don't want to hear them! Things like linear phase EQ are amazing, you can change the eq balance without losing sound quality like you would on analogue.

As opposed to mastering software, I do think that digital emulations of synths can sound weak and false compared to their analogue counterparts. The best emulations of analogue synths in my opinion are the Korg Legacy Collection MS20/Polysix and Arturia CS80 1.5. Stuff like Native instruments Pro 53 sounds like the old generation now, really weak in comparison.

There is a trend at the moment for branding plugins from famous name analogue equipment manufacturers. This is typical marketing bullshit. If you want a vintage neve desk sound, go and hire one. These brands are famous for their time and their sound. The mix is going to sound like the sum of all its parts, everyone is looking for the short cut to a great commercial sounding mix. The bad news is it takes a lot of experience, not a lot of plugins!

The music equipment market is just the same as any other. People seem to forget that as they are so involved in the software and hardware. The forums are full of people ranting and raving about this and that... i am sure some of the posts are put there by the companies. It makes me angry to think of people dreaming away about their next big purchase instead of getting down and writing some amazing new music. I only say that, because I have been guilty myself in the past.

It is always annoying to hear the old cliches, but everyone needs reminding sometimes. Use your ears, not your eyes or assumptions. Often the mastering engineers most important role is that of the trusted 2^nd opinion.