One of the recent trends in the world of mastering is what we mastering engineers call "The Loudness War." In the past couple of years there has been a steady increase in the "loudness" of commercial releases from virtually all the major labels and in almost all styles of music. We are going to take a look at the history, the listener's perception, and the pitfalls associated with this trend, as well as investigate how the marketplace pressures artists to turn their great record into merely the next loudest record. History
How
do we perceive loudness? There are two different ways to measure the loudness of a sound. The first is to find the peak level that a sound reaches; the other is to average the sound level over a period of time. The ratio of peak to average level is called the crest factor. In general our ears respond to the average levels, not the peak levels when judging loudness. An example of this is found in listening to different kinds of music. When you compare the peak levels of classical music with commercial rock, they both peak at the same level, but perceived loudness of the rock music will be much greater because the average level is much higher. One generally accepted rule is that the louder sound will always grab our attention and, for short periods of time, sound better to us. This is why when the record company A&R guy is listening to twenty CDs of new bands, the loudest one will grab his attention. This is great for a single, but as we have found in many of the current crop of hypercompressed CDs, this "full on all the time" approach ultimately fatigues the listener. A variety of dynamics will keep listeners on their toes and make them continue listening. The film community has used this dramatic dynamic for years to keep viewers on the edge of their seats. Five large explosions in a row do not have the same impact as four little ones and one huge one. The
origins of the loudness race With the advent of the compact disc in the 80's, a whole new trend was started. The CD's increased dynamic range and absence of rumble, ticks, and pops caused the artistic community to embrace dynamic range. Many extremely dynamic titles were produced in this period. Several record companies even went so far as to put disclaimers on a CD saying they were not responsible for speaker damage caused by the extreme dynamic range of the particular disc! With the release of the first 5-disc CD carousels with shuffle play in the late 80's, the race started all over again. Many mastering engineers consider the early 90's the golden age of mastering, because decisions made about the loudness of a disc were made for aesthetic reasons and not by marketing people. Many of these records are still alive and kicking today, but if you compare the level of some of the loudest records of the day, like Nirvana's Nevermind or the Smashing Pumpkins' Siamese Dream, these records are 6 to 8 dB quieter than virtually all the commercial rock today. (Just as a note, 6 dB is perceived half as loud). This trend is not limited to the rock world. Virtually all genres have fallen victim to the loudness war. Hip Hop, R&B, AAA, AC and even folk/Americana are making records that have all the music squashed into the top 5 dB of a medium that has more than 90 dB of dynamic range. Moreover, the formats and the locations that we listen to music have changed. This has fueled the loudness war. Many people today listen to their music via mp3 and AAC on their computer, in the car, and on the Walkman/iPod. Because of the limited dynamic range and frequency response inherent in these systems, the data compressed versions of the music sound better. Also, compressed music sounds better in the car, because it gets the sound up over the ambient motor and road noise. Whereas in the early 90's we were mastering records with the assumption that they would be listened to on a consumer playback system in the living room, today we need to take into account the other locations people listen to music. How
did you make it that loud? I am frequently asked, "How do I get my record to sound like the latest and greatest record on the radio?" Unfortunately, there is no easy answer to this question. Most of the records on the radio are productions that cost $100,000 and up. Also, they are done by people with the experience of many records under their belt. But there are several things to think about when you want to make your record loud. I always tell people, "ItÕs not how loud you make it, it's how you make it loud." The big difference between major productions and a home demo is in the way that you go about compressing things. I get masters in all the time that someone has run through the latest and greatest plugin, but complains that it doesn't have the impact or punch that the latest million seller has. What I tell them is that to get a record to be loud without sounding squashed, you have to think about it in a systematic way. It all starts when the first sound is committed to tape or disk. Gentle compression at all the different stages of the production (tracking, mixing, and mastering) will yield better results than doing it all at the end. Makes sense, huh? Yet this is a concept most beginners fail to take to heart. The analogy I make is with archeology. There are two ways to dig up an ancient ruin: with a bulldozer or with small handtools. They both do the same thing, but knowing which tool to choose is the hard part. I'd like to mention a couple of pitfalls that people often fall into when they are mixing a record. The first is how, when, and why to use a compressor on the stereo mix buss. One thing I get all the time is a mix that has been over compressed before coming to the mastering studio. I ask how the compression was applied and invariably hear that the band was comparing the mix to a commercial CD and it didn't sound as loud, so they compressed to stereo mix at the end of the session to bring the level up. This is a patently bad idea. If you are going to use a mix compressor, it should be part of the monitor chain. It should always be there. All the mix decisions about sound and levels will be affected by the compressor, so switching it in and out, I find, often confuses the matter. If you have any question as to whether or not to use this compressor, I would say do not use it. While the mastering engineer can cure a great number of problems in mastering, the one thing that can never be undone is overcompression. Also, your mastering engineer has very specialized tools for stereo compression that can probably do a much better job than the generic compressor at the mix studio. Another thing to think about is what CD you should be comparing your mix to and why. Remember that the CD you are comparing your mix to has already been mastered and if it is a current release, will sound much louder. It is a problem because the mastered version often is radically different from the original mixes. CHOOSE YOUR COMPARISON MATERIAL CAREFULLY! Once you have finished your mix, there is no replacement for the experience and tools of a mastering engineer. While many people today have access to plugins and workstations in their home studio or wherever they are mixing, it takes years of experience to realize that the "if a lot is good, than too much must be better" school of thought does not apply to mastering. I have seen many perfectly good sounding projects ruined through overcompression. There is no substitute for experience. A typical mastering engineer sees about 150 to 200 records every year in the same controlled environment of the mastering studio. Also, an intimate knowledge of the highly specialized tools in the mastering suite cannot be understated. I personally did not feel completely comfortable in the mastering suite until I had mastered about 500 records over four years. The right choices in mastering can raise your project to the next level. It's up to you to decide whether your CD will be crushingly loud or if you will perfect the subtle nuance of the release of a reverb tail. © Mark Donahue
In the early days of the LP, record mastering was simply the process of cutting a lacquer master disc from a finished master tape. As time progressed, producers, artists and record companies began to realize that mastering was an important step in the production process. Skilled disc mastering engineers knew how to get the most out of the master tape onto the disc master. Of course, in the LP days, every record had to be mastered, otherwise copies could not be pressed. The advent of the compact disc and the revolution in computer technology has fueled the home studio and made sophisticated music production financially possible for the masses. It has also changed mastering from a simple transcription process to an active role in the creative process of music making. In the production process, once you are ready to mix your record, it's also time to start thinking about what you need to do to prepare for your mastering session. The first thing to ask yourself is: To what format should I commit my final mix? There are many variables. Do I mix digital or analog? This is a difficult question because both formats have advantages and disadvantages. In a perfect world, you would mix to both and choose at the mastering studio which you liked more. However, back in the real world, you will probably have to choose one or the other before that point. Let's talk about analog first. Analog 2 track is usually only available in the more expensive mixing studios. Why? New analog master recorders cost anywhere from $5,000 to $20,000 and require regular service by a knowledgeable engineer to operate at peak performance. Also, they are very expensive to feed. A reel of 1/2" tape costs upwards of $50 and at 30ips is only 16 minutes long. However, there is a reason why most engineers, when given the opportunity, will mix to 1/2" analog tape: the sound. Analog tape has a characteristic sound that is both revealing and forgiving. And depending on how you set up the machine and how hard you hit the tape, analog can add that special color that takes your mix to the next level. With some very basic understanding of the characteristics of analog tape, you can choose what is right for your project. The first question to address is what speed you should run the tape. There is no correct answer. At 30 inches per second, the high frequency performance improves. However, at 15 inches per second, the low frequency performance improves. Also, at 15 ips the noise increases, so you should think about noise reduction, which today is usually Dolby SR. I usually tell my clients that they should do a mix and print it at both 15 and 30 ips, with and without noise reduction, and chose the one that sounds the best. When the engineer is setting up the machine, you should ask what the operating level is. With today's high output tapes, it is normal to see levels of +6 and +9. What does this mean to you? The simple answer is: At +6, you have 3 dB more headroom before tape compression sets in. So, if you are recording music that has lots of dynamic range, like jazz or folk, and you donÕt want lots of tape compression, use +6. If you are recording heavily compressed rock or urban music, use +9. However, I always say listen and choose what sounds best. Finally, insist on having full calibration tones (1k, 10k, 100, 50 Hz) for each mixing session so the mastering studio can properly align the machine. THIS IS AN ABSOLUTE MUST. The other way to record your mix is digital. In the last five years there have been great advances in digital technology that have made it possible for even the smallest project studio to record in high resolution digital. Today's moderately priced analog to digital converters perform as well or better than the megabuck converters of even five years ago. Here is my advice on digital masters: You should record at the highest bit rate possible. This sounds simplistic; however, the number of bits has a much greater impact on the sound of digital masters than does the sampling rate. A master recorded at 24 bit 44.1 kHz will sound much better than a 16 bit 88.2 kHz recording. In an ideal world, 88.2 kHz and 24 bit is even better. The use of good outboard analog to digital converters also makes a big difference in the sound of a digital recording. This is a simple case of economics. A $2000 A to D converter will sound better than the internal converters on a $1000 DAT machine. If it doesn't, somebody has done something very wrong. People today are working more and more from their computers, and the norm is now to have soundcards that have 24 bit resolution. You should try to keep data in the computer at the highest possible resolution at all times, and when it comes time to go to mastering, make 24bit SDII, AIFF or .WAV files on a CD-ROM. This way the mastering engineer has the highest resolution material to work with. If you must output to CD-R or DAT (only 16 bit,) you should pay special attention to the dither settings in your software or digital mixer. Dither should be used anytime that you are reducing the number of bits. There is usually a dither setup page in the I/O preferences of most software and digital mixers. The sampling rate you choose is also very important. When mastering for CD, you should always record at 44.1 kHz or the integer multiple of 88.2 kHz. When you record at 48 or 96 kHz, the difficulty of the math involved in the sample rate conversion causes the output to not sound as good as 44.1 or 88.2. The final thing about digital masters is to ALWAYS make safety copies. At the end of the day, tape and CD-R's are cheap compared with mastering and mixing time. The next
thing to think about: What do you need to bring to mastering in terms
of different mixes? The suggestion I make is that you should record
the following mixes: There are a number of reasons to print all these mixes, the main reason being that you can recreate any mix from the individual elements. If you need to take out profanity for a radio version, you can edit the solo voice track and then add it to the TV mix to duplicate the final version. The 30 minutes it takes to print the additional mixes is much cheaper than having to go and recall a mix to bring the vocal up 2 dB. The last mix issue to talk about is the use of 2-mix compression. When the mix is finished, the worst thing you can do to a mix that will be mastered is throw a compressor on to make it a little louder. Compression is the one thing that canÕt be undone in the mastering studio. Also, the compressors and limiters at the mastering studio are usually of much higher quality than those usually found in mixing studios, and can usually do a better job of making the mix louder. If you are going to use 2-mix compression, the compressor should be an ever-present part of the monitoring chain. All the decisions you make about level and EQ will be affected by the compressor. A soft ratio (less than 2:1) is best. And finally, when you are finished mixing, book your session with the mastering engineer, but give yourself a good week or two to listen to the mixes, come up with a final order, and catalog where all the mixes you want to use are located on the tapes. This will save you time and money when you are in the mastering studio. What is difference between mixing and mastering? Mastering is both an art and a science. It is the last step in the creative process and the first step in the manufacturing process and thus requires the mastering engineer to have an understanding of the science of sound and the aesthetics of music. When you start to think about it, mastering is the exact opposite of mixing. In mixing, you are taking the individual elements and combining them into a single picture. In mastering, you look at the big picture and make changes that affect the entire picture. What makes a mastering studio different from a recording studio? The most important difference is the listening environment. An extremely accurate, full range listening environment is the mastering engineer's sharpest tool. When it comes down to it, if you can't hear all the subtle information in a mix, all the gear in the world won't help you create a good master. The processing that the mastering engineer uses is very simple: equalization, compression, limiting and stereo width adjustment. While the processing may be simple, the tools the mastering engineer uses are very specialized. They must be very precise and resettable, and usually are of a higher resolution than those found in mixing. A noisy Pultec EQ may be great on the vocal track, but would you want to run your whole mix through it? It seems today every studio advertises that they do mastering. While it is possible to make a master in whatever computer workstation is handy, the finer points of mastering are a very specialized skill learned through long and hard experience. I often joke with my clients that "the first couple of hundred masters are hard - after that it gets much easier." While mixing engineers get to hear a single track for days or even weeks, the mastering engineer hears an entire album in a day. That means forty or fifty mixes in a week! The mastering engineer brings a different perspective to the project, and a fresh set of ears with no preconceived notion of how the mix sounds. Ultimately you get what you pay for in mastering. A good mastering engineer can bring your mix up to the next level. An inexperienced mastering engineer can do more harm than good. A good mastering engineer can listen to many different kinds of music and make an informed decision as to what can make a good mix sound great, regardless of genre. A typical mastering session consists of three elements. The first part is what is actually called the "mastering". This is the process of making any sonic changes to the original mix master. Normally I like to work in the final program order, so there is a sense of flow from one song to the next. Once all the mastering is complete, the next step is sequencing. This is the process of putting the songs in order, trimming the noise off the head of the song, doing the fades and setting the space between tracks. There is no such thing as the "standard" spacing. All spaces are done by feel, with the tempo and fade of the previous song dictating where the next track should start. The final step is where the creative process ends and the manufacturing process begins: the creation of production masters. These are the tapes or discs that are sent to the factory for replication. Realistically, it takes at least 6 hours to master a full length CD. If you think about the process above, it takes 3 times the running time of the program just to copy and listen to each song. The last thing you want to do is to rush through the mastering process. Once a master is done and the disc is replicated, there is no going back. The cost of full service mastering facilities can vary widely, from about $100 to $400 per hour. What you will find at a full service mastering facility is the ability to play virtually any format that a mix would come in on (1/4", 1/2" analog w/wo noise reduction, DAT, CD, AIFF, SDII, .WAV ...) and the ability to make analyzed production masters in whatever format required by the record company or replicator (PCM-1630, DDP, PMCD, PCM-9000...). The analysis of the data integrity of a production master is absolutely necessary. Also they should have a single purpose monitoring environment with appropriate mastering tools, both analog and digital. © Mark Donahue
Duping
Do's and Don'ts PCM-1630 This is a professional format that you will find only in mastering facilities. The advantage of this format is that it is analyzed for errors before leaving the mastering facility and is verified at the factory. This format has been the industry standard for nearly twenty years. This is the only format that most manufacturers guarantee leaving the building. If you are going to spend the money of professional mastering, this is a very cheap insurance policy. Any replicator that will not replicate from a 1630 should be avoided. CD-R This has become the most popular delivery format in the past few years. However, there are a number of issues that you need to be concerned with, as all CD-Rs are not created equal. The CD-R that you get at the mastering facility is called a PMCD or pre-master CD. This is a Redbook standard CD with an additional data block at the head of the disc containing all of the track information as well as ISRC and UPC/EAN codes. If you are sending a PMCD for replication, ask your mastering house to analyze the disc for you. This printed analysis is essential for heading off problems later in the process. If you are making your CD-R master at home, there are a number of things to which you need to pay close attention. The only disc that you can send to the factory for an audio-only CD is a Redbook or Disc at once CD-R. These discs are created in a single pass and contain a complete table of contents. The CD-R created in a stand-alone CD recorder can not be used to create CD masters. These recorders create what is called an Orange Book or Track at once CD. Normally what happens at the factory is that an extra fee is charged to transfer your Orange Book CD to a Redbook master. Hybrid CD or a CD that contains both audio and data sections are called Bluebook CDs. When creating these discs, the same issues for the audio and data portions of the disc apply. Each section needs to be written in a single session, with the audio portion always going first. DDP Once you have made your master, go through it with a fine toothed comb. Listen to it over and over on different stereos. Listen for even the tiniest click, pop, dropout, etc. After all, you spent so much time and energy, not to mention money, that you don't want to rush things at this stage. Remember, once you have sent your master off to the duplicator, it canÕt be changed and there is nobody at the factory to catch your mistakes. The last issue dealing with the CD production master is that of safety copies. You should always have three copies: master, safety, and reference. The master gets sent to the factory, the safety sits on a shelf until the finished CDs come back from the factory, and the reference is for you to audition. Role of the Broker Most of the people who advertise replication in places like Performer are brokers. They do not own CD factories. They are essentially re-sellers of the factories' services and play a very important role in the replication process. The broker offers packages that range from the most basic (you provide films for your artwork and a CD master) to all-inclusive deals (including graphic design, mastering, all films, etc.). One of the things that you have to remember here is the rule "You get what you pay for." Often a package that sounds too good to be true is just that. Always ask your broker what, if any, additional charges there are. Shop many different brokers. They will all have packages that will suit your needs, but you will find that the bottom lines vary greatly. Many inexpensive replication packages often include add-on charges for glass mastering, film and separation creation, assembly, shrink-wrap, and so forth. More expensive packages include all the above add-ons as well as value added services like graphics assistance at no extra charge. Also be aware of the differences between replication of Production CDs and duplication of CD-Rs. While CD-Rs are a great deal for a couple of one-offs, there are many things to be concerned with: replicated CDs are significantly more reliable and durable than do-it-yourself one-offs Š which are literally "burned" rather than injection-molded like the replicated CDs that major label releases use. In addition, replicated CDs are dramatically less expensive than CD-Rs in anything other than the smallest quantities. Shop around and Caveat Emptor. Once your project has gone to the factory, there are only a few things of which you need be aware. The only part of the replication process that has any bearing on the sound of the final product is the speed at which your glass master is cut. This is a contentious subject for many people in the replication industry. While the data on the final product may be identical, there are marked differences between the sound of discs where the glass master is made in real time (1x) versus faster than real time (2x, 4x). The culprit is called "jitter." Jitter is defined as a variation in the timing of digital data. Without going into a technical discussion (which is well beyond the scope of this article) we can discuss how jitter affects the sound of your CD. Jitter is very insidious because unless you have something to compare your replicated CDs with, like a safety copy, it is very difficult to hear the problem. Some telltale signs of high speed replication are: the high frequency information will sound more sibilant. The low frequency sounds will lack definition and " punch" (especially things like kick drum and percussive bass sounds). Things that are mono (lead vocal, etc.) will no longer be a pinpoint source between the two speakers, but will spread into an undefined area. These differences are subtle but noticeable. The other problem is that these effects are most obvious when the disc is played back on inexpensive CD players, which comprise about 98% of all CD players sold. The way to avoid this problem is to request real-time glass mastering. This usually involves a small additional charge ($100 - $300) but is well worth the money. If a broker or replicator refuses to respond to your request for real-time glass mastering, you should consider taking your business elsewhere. You may also wish to remind them that they are in a service industry, and this service is important to you. © Mark Donahue
|
| |
|
| |
|
| |
|