If you didn't catch the first part of this interview, step on over to the following link and get caught up!
And now…….Part 2.
PR: Since MP3 has become essentially the 'in demand' format, has this placed any extra changes or demands on the work of mastering engineers?
BK: I think that the biggest thing that is going on is that more and more listeners are listening to individual songs and not whole albums, which I'm kind of sad about because I think that creating a terrific album is an art. And the fact is that if you're singles-oriented, the usual approach is to take each song and master it with some degree or a greater degree of compression to make it seem loud, assuming it's going to be listened to next to who knows what, maybe next to the loudest song on earth, rather than in an album format.
So the indirect answer to your question about MP3 is really about downloads in general and a singles-orientation as opposed to an album orientation. So people are tending to treat MP3s as singles rather than like albums.
PR: So in an ideal situation, is there a benefit for creating the MP3 from the original mastered audio, or just taking it from the end product audio CD itself?
BK: Ideally, extracted from the audio because when you compare the two, you can get a marginally better sounding MP3 if you take it from the 24-bit file that precedes the 16-bit file on the CD. The MP3 encoders can do it a little bit better if they are given a higher resolution as the source. I believe this to be the case, though I can't call it a fact. But I do believe it to be the case.
PR: So would that then become the job of the mastering engineer to take the original mastered audio and exact the MP3 from it, or is that typically just done after the fact?
BK: Usually after the fact. Most of our clients will make their own MP3s because there's an extra cost, and most clients have the tools to do that.
PR: While we are on the subject of alternate formats, what do you think is the future or fate of higher resolution audio formats like SACD and others?
BK: If we're talking about audio-only formats, well, history repeats itself. I'm not that optimistic about it. I think that it's going downhill. I think that most people's home hi-fi systems consist of a home theater system with a video monitor. They usually want to see picture as well. If there's not an image associated with it, chances are the fate lies in a relative niche market. In other words, listening to high resolution audio such as SACD as long as it continues to last, or even HDDVD as it takes over; if it doesn't have a picture, it's likely doomed to a very small audience.
And that is sad because what I do all day long is listen to really great sounding, usually stereo recordings on a fantastic stereo system, and I turn off the video.
PR: For somebody wanting to get started in mastering, what kind of things should they focus on learning?
BK: The first thing is to listen to a LOT of live music performed without any sound reinforcement. Good luck finding that! But I'm very serious about it. If you don't know what real instruments sound like, then you're not ready to become a mastering engineer unless all you want to do is hip-hop. And I'm not saying that hip-hop is all about making artificial sounds. I'm just giving an example of how limited your sights can be if you don't know what real music sounds like, and it'll keep you from over-equalizing.
But nevertheless, you have to know many different styles as well. So there's the old saying "too much ain't enough". If plus 9db at 30Hz is what turns on a particular rock recording, even if it doesn't sound natural, that's fine. But you should be aware of what the natural response is so you can know how much is too much when you make it too much. That's my feeling that you need to have that before you become a mastering engineer.
The second thing is that you need to become an audiophile in the sense that you need to know what well-recorded music played back on a wide range of reproduction systems sounds like. And have that preferably in your home. If you go to one of the audio schools and all you're exposed to is 4-inch woofers and 1-inch tweeters sitting on the meter bridge of a console, you're not destined to become a good mastering engineer because mastering is a place where we listen and make judgments on audio microscopes that reveal very small errors and issues that keep us from doing the wrong things with recordings. And unfortunately, if all you have is a low resolution system typical of the near-fields that people have been exposed to more and more and don't have stereo systems in their home, and only have a good car system, you're not going to make a great mastering engineer in my opinion.
I've done a dozen, 15, or 20 seminars where I spend a lot of time setting up a really good, high headroom reproduction system, and I play back a wide variety of material for the audience. And most of them come out of there blown away, having heard many differences and problems they were never aware of in recording until now.
PR: So for somebody who runs a home studio, wearing all the hats due to budget or whatever reason, is there any hope? How would you recommend them to proceed?
BK: Get a great stereo system, and listen to all the major recordings of the past several years. Some of the interns that come in here, the first time they have ever heard a great stereo system is when they enter the mastering room!
PR: So for somebody who does feel like going it alone, would you say that the monitoring system is the most important thing they can spend time and money improving?
BK: And the room. But note, that if you're trying to do mastering, generally a room dedicated to mixing is going to be the wrong room for mastering. There aren't many dual-purpose rooms around that can actually do a good job at both.
In general, all the mixing consoles, racks of gear, etc. can even act as acoustic obstacles. So keep that in mind.
PR: So talk to us about the K-System. You're widely known as it's creator. Can you give a general idea about the K-System?
BK: First, to be clear, there was no ego in naming it that way. I think it's just a good abbreviation that makes it quicker and easier and means a lot more than if you…..it has to have a name. But I do not receive any profit whatsoever from any of these meters, and I don't manufacture any of them. All the specs are public domain and open. I offer my help in implementation for free when needed, so this is not a commercial venture for me. I just want to be clear on that.
So let's talk about the benefits. First of all, the most important concepts that a lot of people initially are not aware of is that the K-System is an integrated metering and monitoring system. In a really good monitoring environment, with high headroom loudspeakers, located at a certain distance that have been calibrated, you don't need any meters at all. When I master in my mastering room, the vast majority of the time I don't even look at the meters.
So the K-System, if you want to start learning about it, drop in at the honor roll at digido.com and see if you can pick up as many as possible of those recordings and use them as references. Notice how they are ordered by order of monitor gain or attenuation setting, and that for recordings that have high transients that are up-tempo and rhythmic and have a lot of snap to them, that are normalized to full-scale peak; with those two requirements, you can fairly well predict that the lower the monitor gain that you have to use, the more compressed that they will be.
Loudness of the recording in front of the volume control is part of the K-System. And so if you recognize that a monitor level control is like a water faucet, the more water that comes into your faucet, the more you have to close it. If you're water faucet was market numerically, you could judge the pressure that must be coming into the faucet by the position of the faucet and the pressure of the water that's coming out. And that's what the K-System is mostly about.
If you calibrate your monitor and you know that it's set to -6dB, and it sounds loud, chances are that it's a very dynamic recording and relatively clean, and it's around a K-14.
So the monitoring should integrate with the metering, using an SPL meter with pink noise calibration. The reason is because meters are typically approximate and not very accurate in measuring loudness. Also, you have to realize that the ears are not absolute instruments either. For instance, you hear differently in the morning than in the evening. Or if you listen go loud music for 5 minutes, it doesn't sound as loud. So this makes it more important to have metering and monitoring integrated and use them together.
The K-System encourages producing cleaner recordings that have good headroom and have some bounce, impact, and punch to them. And that's important, isn't it?
PR: Of course, and a lot of people miss that.
BK: Oh, but more and more people are beginning to get that. You see on the Gearslutz board when people try the system, and they say they are doing better recordings than ever.
PR: Is the K-System meant to replace other metering systems such as RMS and peak metering, or supplement them?
BK: I think that a scale that's centered around a zero point is the most natural kind of scale to use. In the old days, zero was the point we were going for. And the K-System is working around a zero point for forte passages, or +3 or +4 for fortissimo passage, or mezzoforte maybe -1 or -2. So it actually has kind of a meaning. If an RMS metering system is measured in values relative to full scale, with full scale called zero, then you start thinking in terms of a minus when talking about your RMS levels, such as having an RMS level of negative 14 or negative 12. I personally like to think of that as my RMS is 0dB at forte. It feels more comfortable because all of a sudden, +1 is more, -1 is less, based on the new reference point. It makes it easier to understand.
So you can work with the other meters, but it's just harder to grasp around your head. I mean, when it goes into the red, that means something.
PR: Basically, going into the red means you're radio ready, right?
BK: (laughs) The more red lights you see, the better it will sound on the radio, right? (sarcasm)
PR: Are you aware of any labels, major or otherwise, who are trying to combat the loudness race?
BK: I get sporadic A&R directors at certain record labels that understand it and get it, and we work together to make better sounding recordings that aren't so squashed.
Major labels? Ha! The definition of major label right now is to lose billions of dollars. We don't know if there will be any majors, so-to-speak. But there are some independent labels who are doing it.
PR: Are you aware of any existing technologies that offer some hope in combating the loudness race?
BK: Yes. I believe the future of reproduction in the home is not the CD player, but the server. More and more music is being transmitted as downloads. And there is, for example, a server-client system made by a company called Sling Devices. And they have two players that are really easy to use, almost plug-n-play. The smaller one of them is called the Squeezebox, and you plug it into your stereo system and it can play internet radio without a computer. Or it can play your entire music collection from the server, with full control just as with a CD player. Both this device and the bigger brother, the Transporter, can play double sampled music and downsample it to play directly.
So what does this have to do with your question? There is a technology called 'loudness normalization' where there is an identity bit in the ID3 tag that indicates the apparent loudness in the recording. It's a little like the dial norm number that gets applied to motion pictures.
But anyways, what the loudness normalization does is that if you go to the menu on your Squeezebox and set it to turn on the loudness normalization, it will read the tag and know what level to play so all your music comes across at a consistent level. So as more and more producers begin to realize that no matter how loud they make a recording, that the playback system that the consumer uses is going to control it down, then there is hope that they will stop overcompressing their recordings.
This is also present in iTunes. You can control the playback level on the consumer level by just reading a tag. So this is all a step in the right direction.
PR: Do you find that people who are good at mixing are generally able to learn mastering well?
BK: It's hard to cross the line. You have to be schizophrenic. It's possible to be good at both, but mixing requires a totally different head than mastering. But you have to make a little mental change in your head before you switch from one mode to another. But both are completely separate skills that you have to develop.
PR: Talking about software, some effects market themselves as a "mastering effect". For instance, some compressors claim to be more suited for mastering than mixing, or vice-versa. Is there really a difference needed between these tools, or are they relatively the same?
BK: Yes and no. There are a number of compressors that were originally used in mixing studios that have gained wide acceptance on the mastering side, and vice-versa. But to be a good mastering compressor, you need, for instance, certain time constants that you may not find in a mixing tool. And also, it helps, especially in mastering, the wider the frequency range of music you're dealing with, the more frequency-sensitive controls you need in your compressor. If your bass drum has too much energy, it can really screw up a wide band compressor, for instance.
PR: So you're known by many as the author of the popular book "Mastering Audio". Is there anything you would like to talk about concerning the book, or perhaps a bit of a plug?
BK: I would rather let others talk about it.
PR: Well, it's been referred to as "the" guide when it comes to understanding what the mastering process is all about.
BK: Well, that makes me really happy that I've struck a chord. That I'll say. Thank you to everyone who has resonated at the same frequency that I have.
I will say that I think it's more than just for mastering engineers, but also mixing and recording engineers. The title is a play on words. It's not just about audio mastering, but mastering the art of audio itself. It also helps to prepare recordings and mixes to create better masters.
We would like to say a very big thanks to Bob Katz for taking the time out of his schedule to do this interview. It was truly an honor to be able to do. So thank you!
NOTE: And at Bob's note, I do recommend checking out his book, "Mastering Audio". This is a genuine recommendation and we get nothing from it, but I have this book in my own personal library, just as many other professionals do. If you want to take the things Bob talks about further, it's probably the best resource you can have on your shelf. Bob's website is http://www.digido.com where you will find lots of great information on the book, the K-System, and much more.