Mastering Log

INTRODUCTION

After loading up the individual mixes, most mastering sessions start by evaluating the mixes, both individually and as they relate each other so we can settle on an approach will which will eventually become "the album". When working with quality mixes, our main job is to deliver an album that can be listened to in one pass, without the listener feeling uncomfortable due to volume imbalances or unplanned tonal differences. In most cases, the mixes are enhanced, sometimes just slightly or sometime quite a lot, in order to finish the mixer's work and take the album to the final level. A good mixer usually leaves us some room to manipulate the mix in order to complete our "balancing act". As a general rule, leaving "room" means that the mix is not quite as bright, compressed or as loud as the mastered version will be.

On some sessions, the mastering engineer is called upon to do some tricks, i.e., make some extra magic happen. At the time, when it works, the clients are usually amazed and sometimes, so is the engineer!

Other times, we have some insight on why a session, or part of it, went well, or didn't, and would like to share some thoughts with you. So, the idea of this website page was to write down some observations that might be interesting to our readers.

As I do my work everyday, I still find lots of interesting things to learn and situations that cause me to do more research in order to explain a new observation so I'm hoping that some of these thoughts will be helpful to you as you get ready for your next mastering session. I'll continue to add to this page as often as I can.

If you have any questions, comments or ideas to make this a better forum, please e-mail me at andy@silverbirchprod.com.

RANDOM TOPICS IN NO PARTICULAR ORDER:

HOW LONG WILL IT TAKE TO MASTER MY ALBUM? (also see FAQs for a more concise explanation)

We are often asked to estimate the length of time it takes to master an album. So, besides the daily routine of trying to help people figure out the cost of an album over the phone, the inspiration for this topic came from two recent, consecutive mastering days.

Day one was an pop album that needed eight songs mastered with the addition of two previously mastered songs from a different album. Day two was an eleven song album from a pop-metal band. The first album took seven hours and the second one took four and a half and yet the performances, recording, mixing and production were very professional.

So, to start at the beginning, the first point to make is that mastering, like mixing, is an artistic endeavour and so, if you want the best job possible, you should not limit the hours of a mastering engineer. I.e., someone will call us and say that they had their last twenty-one song album mastered in four hours and why does Silverbirch think it will probably take seven or eight? Well, there are reoccurring patterns from job to job and so it is possible to make an educated guess re time needed.

Let's assume that your mixes come to us on CD-ROM (data disc) or CD Audio. These are the most common and fastest format to load. We also get DATs and various reel to reel tape formats, which have to be loaded in real time, but see these formats far less often. Mixes delivered on DAT will add approximately an hour to any average album session. Tape can be more costly to transfer, especially considering the three different tape widths which can be used, plus the possibility of being encoded with Dolby SR.

We start by loading and preparing the files for mastering. This includes checking and trimming the front and ends of each file, if necessary (called top & tailing). I also quickly analyze the peaks of the files and sometimes adjust the overall volume so that each song hits my analog gear with at a similar level. I then mark two to three key points in each song, i.e., in a pop production, a verse and out chorus and any other section that may be useful to loop or refer to during mastering process. This preparation process takes an hour to and hour and a half.

After the prep work is done and before starting the actual audio work, I will have a dialog with the client to see what is wanted or expected from the mastering process. This might include explaining what goes on in a typical mastering session. If my clients are interested and haven't attended a session before, I will go into some detail. The education part is important because if you don't explain the possibilities, the artist, producer, etc., will sometimes not offer enough insight into some of the problems they had, or a still having, with the mix. We need to know how faithful you want to stay to the mix, whether keeping dynamics versus a high volume output is your goal and a host of other important topics. There is definitely more than one way to master an album! In fact, some major labels will pay for two or three good mastering studios to master the same album and then pick the one they like best. Of course, that is the exception because most mastering engineers are chosen because of their good sounding albums, track record and/or recommendations by those who like their work.

Although not all mastering engineers share my philosophy, here is how I like to work. I am hired to be the "fresh, unbiased ears" and am responsible for taking your mix professionally to the final production CD. In the process, I respect that fact that the people in the room, i.e, artists, producers, engineers, etc., have spent far more time on the project that I have. There is a very good chance that they will have some very useful insights about their project that would help me to give them the best mastering job. Based on what you say, I will try to give you the sound you are looking for because there is more than one way to master an album (an important observation to keep in mind!). If any of your suggestions will move the album technically or commercially out of the norm, I will advise you, especially if your suggestion(s) has any negative ramifications to the sound! However, you will always have the final say.

Because we are supposed to be the "final authority", many mastering engineers are just a little too rigid in their approach and tend to feel that they know how the album should be treated and therefore are not too interested in hearing their client's opinions. I know this from personal experience as I was a producer for a number of years and attended a few mastering sessions as a client! I still hear of mastering engineers who charge extra if a client wants to be present at the session. As an ex producer, I find that philosophy arrogant and perhaps foolish. If you want to hear a tale of woe, ask any experienced producer how many times he has received the production copy of a CD, where he has not been present at the mastering, and has been disappointed in the mastering job. Needless to say, we welcome our clients to join us at the mastering session.

Anyway, I digress, so back to actual mastering session. The sonic work on the album begins by picking one track as a "template" track, for example in a pop production, one of the loud, dense tracks but not one that is atypical of the rest of the album. I will work on the sound until I have something that I think is worth auditioning. At that point, I involve the people in the room and we listen the unmastered and mastered versions, to see what was done, and also, may listen to some excerpts of good sounding major label CDs that are approximately of the same style or idiom. This gives us all an insight to how we may want the album to sound in comparison to the "competition". I usually spend about three quarters of an hour on this particular segment.

After we all agree on the final sound for the "template" track and create the master, the next track we work on is the one that is sonically closest to it. Mastering in musical groups is the fastest and most economical way to work. That is because we, like most pro mastering studios, use a fair amount of class-A outboard gear and it is not easy to "save" the settings. So, I like to move the knobs as little as possible as I move from track to track. With my method, the last track to get mastered is usually the slowest, most sparse or the most different track on the album. As for time, after the "template" track, most tracks take between twenty to thirty minutes to master.

The last step is finalizing the spacing between songs and cutting the master. This takes about 30 minutes.

So using the examples from above, let's calculate the time it may take to master a eleven track, well recorded pop album. Take an hour to load and prep, and forty-five for the first track, twenty minutes for each subsequent track and thirty minutes to finish. That works out to about five and a half hours.

Variations on this are as follows. A pop album with less musical variations that is recorded and mixed with very a consistent sound can sometimes take less time master (see Day two in the second paragraph of this topic). I've actually mastered albums in as little as four hours but this is certainly not the norm. An album with sparse instrumentation, such as piano and voice, will usually average out at three and a half to four hours simply because it is usually easier to work on that type of album.

On the other hand, many albums take seven or eight hours and sometimes much longer. This is because there can be technical problems (pops and clicks caused by the recording studio that had to be removed), editing work, noise reduction or simply continual requests to try to improve the mix because the clients weren't particularly happy with what they received from the studio. Sometimes, a nice sounding collection of mixes can prove difficult to turn into a well-balanced album.

Requests such as "can you raise the vocals", "give us more kick without more bass", "turn down the hi-hat" are just some of a long list of things you can try to do at the mastering level. Of course they should have been done when the clients had access to the multitrack volume, eq and compression, but for many reasons, could not get what they wanted or could not afford the time or money to go back. Some of these requests can be successful at the mastering level, but there are no guarantees! In other cases, a few of my clients keep coming back to fine tune the mastering. The most extreme example was seven touch-up sessions! The record here at Silverbirch is thirty-two hours! That was a choir director who decided to have me re-conduct the group for her. Most of the time was spent editing.

If you have special requests sequencing requests, such as cross-fades or audio hidden in pre-gaps, add another 15 to 30 minutes.

A final thought on time. As a rule, the more people you bring to the session, the longer it takes. Musicians tend to listen to their own instruments more than the whole production and one can get caught up in trying to clarify or change something that sounds fine in the context of the whole but perhaps not to that musician. Experienced producers usually have a good overview on what it takes to finish the album and know when to quit polishing!

***************************************************************

SOME NOTES ON THOSE TROUBLESOME TOPS AND TAILS

The inspiration for this topic came from mastering a very poorly recorded demo. However, the problems outlined below happen far more often than you would think, and sometimes with otherwise very professional mixes!

One of those unsung jobs that mastering engineers do every day is check, and correct if necessary, the beginnings and endings of each song on your album. Ideally we would like a clean and noise-free start and a smooth, noiseless ending. How often does this happen? On a consistent basis throughout the album, maybe only about 30% of the time!

It seems that the catchword these days is to "fix it in the mastering". Well, here's what is easy to fix. A smooth, noise-free ending that is a little too long. An ending that is on the long side is no problem to work with, and even desirable, because for the first time, we have the album in its correct order and can begin to think about the transitions between songs so a little extra is not a bad idea. However, if the engineer is experienced and knows what he is doing, it is perfectly OK to do all the top and tailing before the mastering session and save a half an hour to an hour of studio time.

Endings that are not easy to fix are endings that are noisy or too short. Noisy ending come in several variations. A common problem begins at the tracking session where one or more of the musicians have moved or hit their instrument before all the instruments have had a chance to fade out naturally. Sometimes, in spite of all the warnings, the tracking engineer's comments just aren't absorbed and then the "magic take" is recorded and the noisy ending is accepted. In this case, the mix engineer can help by muting or fading all of the tracks individually to cut out as much noise as possible. This is generally a good practice to follow at all times.

If you don't have a great monitoring system, use good headphones for this purpose. In fact, using headphones to check a mix for clicks, pops and other problems is a really good idea. All too often, clients are noticing for the first time the various technical flaws when they come to master their album. The reason is good mastering suites have really excellent monitoring and so all of a sudden, the problems are noticed by them or the mastering engineer.

Sometimes, a band will come to mastering and think that an ending can be cut digitally and sound natural. Well, anyone with a digital workstation can try cutting a file and will discover instantly that that is a most unnatural act!

Another common problem is compressor noise. This seems to happen more often with plug-ins and digital workstations. A compressor used on an individual track, or tracks, can become noisy when there is no signal being sent to it. The result is a hiss on quiet spots in the song and in the tail. Amps and less-than-stellar mixing boards can also cause hum as well as hiss. If the mastering engineer tries to make a smooth fade of a track that combines music, hiss and hum, the result is usually less than satisfactory.

So, here are some of the techniques we use.

The first one is, of course, to try to refade the ending. We use the E-magic controller surface which has faders and allows a more musical approach to fading. The fades that are created by hand can then be edited -- the best of both worlds. Of course the auto fade is handy at times but is rarely used!

Sometime, a reverb unit can be used extend an ending. First we create a reverb setting that sounds like the overall reverb of the track. We then cut the noisy ending short and let the last clean sound of all the instruments playing together hit the reverb unit and decay naturally. This is then blended into the rest of the track with automation or editing. We use Digidesign's Reverb One or the Altaverb by Audioease.

A noisy ending can be treated with noise reduction gear, both for broadband noise and hum. If noise is not a problem throughout the track, we will just treat the ending. This is usually done with automation. We use the Waves Restoration Noise Reduction Toolkit.

Occasionally, using a "time-stretching' program can be helpful. A typical example is an ending with power chords that is too short or perhaps too noisy to use as is. The program will actually sample the sound and make it longer. We can then make a more natural fade now that we have more length to work with. We use a program called Pitch'N'Time.

The above examples all refer to "natural endings", i.e., endings where the band, or some part of it, is ending together. Faded endings (sometimes called a "board fade" sometimes come to us with one of two problems. The first is that the end of the fade drops off too suddenly or is cut off. The other problem is ending in an unmusical place. The fastest and easiest fix is to try an smooth out a problem fade with a subtle refade. A more unusual and drastic approach would be to find an appropriate eight to sixteen bars (or a loop) and replace the old fade with the new material. Once the material is edited into place, a new smoother fade can be done.

The subject of the tops, or beginnings of songs, is shorter because there are less problems to deal with.

The most common problem is noise before the track. This is usually very easy to cut. However, if there are any background noise problems to deal with, either within the track or at the end, please, please leave us a sample in front of the track. This is a case where it is not helpful for the mix engineer to cut the track close.

When it comes to noise before a track, I always ask my clients if that was intended. Some engineers and producers like the sound of the room or amp before the band starts, feeling that it sounds more natural than a clean start.

Sometime the track seems to start abruptly. In this case, sometime a very slight fade-up can make it smoother.

The last point to cover is cross-fades. When taking a creative approach to sequencing an album, a client wants a song to be fading out as the next one fades in. We have effective ways to create a variety of cross-fades; however, you should budget extra time for this process as it does make the mastering process more complex. One of the reasons for this is that it is sometimes the two adjacent tracks will need different mastering settings, in order to balance the album, and therefore when cross-fading between songs, there has to be some thought put into to how to blend the cross-fades together so that they sound natural. Besides the sonic work, it does take extra time to audition each cross-fade until everyone is happy with the results.

One also has to make a decision re the I.D. point for each cross-faded track. There is usually several choices and sometimes none of them are completely satisfactory! One also should consider at least consider how random play will affect the top and tail of each track. The last issue to point out is that all CD players interpret the start of each I.D. differently. The most extreme example is a fifteen year old player vs a brand new version. In most cases, the old one will start later than the new one. The mastering engineer has no control over how the various CD players are going to pick up the I.D. and therefore it we have a choice, we will try and pick a point that will work well a little before or after the chosen spot. With cross-fades, the client has to accept that in order to achieve a seamless sound from track to track, one may have to accept compromise in regards to the start of each track and the effect of random play on the CD.