It’s that time of year!
Lately I’ve seen lots of comments in
audiophile discussion groups and websites about how bad mastering is the source
of modern music’s sonic ills. But how many audiophiles have actually met or
talked with a mastering engineer? Not many I’ll wager.
To help clear away the fogs of ignorance
I see rolling across the Internet, Audiophile Review decided to ask a genuine
working mastering engineer some questions.
David Glasser is the one of the founders
and the chief engineers at Airshow Mastering in Boulder, Colorado (Airshow has
another facility in Tokoma Park Maryland run by Airshow’s co-founder, Charlie
Pilzer.) Glasser has over thirty years of experience, and has mastered
thousands of recordings, including more than 80 nominated for Grammys. He has
earned two Grammy awards for mastering and restoration. The first was for the
1997 Anthology of American Folk Music and the second for the 2002
release, Screamin’ and Hollerin’ the Blues: The Worlds of Charlie Patton.
Glasser began his audio
career with the Boston Symphony Orchestra, followed by eight years of recording
and production for National Public Radio. After founding Airshow in 1983 to
offer location recording services, he shifted to full-time mastering in 1990.
Airshow Mastering expanded to a second facility in Boulder, Colorado, in 1997.
Since then Airshow has grown into one of the largest mastering facilities in
AR: What does a mastering engineer do?
Why would someone use a mastering engineer?
DG: Simply put, a mastering engineer puts
the final touches on a recording. That could include: dealing with levels –
both song-to-song balance, and overall level of an album project; tonal
adjustments – examples might include adding presence or high end to an
otherwise good mix, or reducing muddiness and boominess. A lot of the job is
working to present the mix in the best possible light, taking into account: the
vision of the artist & producer, and other similar recordings. Also,
mundane but important work including quality control, encoding for lossy
formats, and metadata encoding.
Why do musicians use a mastering
engineer? It’s a fresh set of unbiased ears – hopefully working with the proper
equipment in a well-designed room, and with experience to know what to do or
AR: What forms do music files come in
nowadays? Do you work with multi-channel or two-channel files? And what
resolutions do you normally use?
DG: Things have actually gotten simpler
as far as file formats go. PCM mixes come in as wav or aiff files, almost
always 24 bit, and with sample rates from 44.1 to 96k (quite common) to 192k
(occasionally). Mixes can also come in as DSD files, which I import to a
Sonoma workstation, or on analog tape.
We have well-maintained Ampex ATR and
Studio 820 analog recorders with a variety of headstacks and replay electronics
to choose from for tape-based projects. I always master at the highest
available resolution and downsample as required for release. If there are
multiple release formats – CD, LP, download – we encourage the client to cut
the vinyl from our high-resolution files. Same for download – Apple now accepts
high-resolution files for iTunes encoding. We also deal with legacy formats
including PCM-F1, Genex, 1630, cassettes, DATs, and others, and we are equipped
to make analog to digital transfers optimized for Plangent Processing.
AR: Walk us through what would be
involved in a mastering session.
DG: It always begins with discussions
with the artist, engineer, producer, and whoever else is involved in the
creative process. We’ll talk about what the project needs – sometimes this
means very detailed instructions and suggestions. Other times I have carte
blanche to do what I think is called for. I listen to the mixes – maybe not all
the way through – but enough to get a sense of the musical direction. Then I
try several approaches – different equalizers & compressors, analog and/or
digital signal paths, etc. until I find something that works. I work on each
song individually, in album order to maintain continuity.
Sometimes the initial settings serve as a
template for the whole record, sometimes not. It’s all program dependent. After
this processing stage the songs are edited and assembled as needed, and the
required masters are created. The client always receives a reference for
approval before the masters are shipped. All masters go through a multi-step
quality check process.
If it’s a surround session, or a
lock-to-picture project for DVD or BluRay, things get more complicated, as
there are more options for surround, but the basic idea is the same: present
the music in the best possible light.
AR: What was involved in the Grateful
Dead Europe 72 box set project? How much time did you spend on it? Did you
actually listen to every second of the recordings?
DG: Europe ’72 was a dream project! I
spent 7 months working on this – 3 to 4 days per week, occasionally more. I
worked on each 3-4 CD show as soon as Jeffrey Norman completed the mixes and
The Europe 72 tour was recorded on
2″ 16-track tape. These multi-track tapes (2 pallets worth) were
transferred by Jamie Howarth. The tracks were all put through the Plangent
Process, which eliminates wow and flutter. I think this contributed to the
clarity we were able to maintain. The Plangent-processed files were shipped to
Jeffrey Norman who mixed the project. Jeffrey sent the 96k 24 Bit mixes to me,
and after mastering, I sent references to him and producer David Lemieux for
Occasionally revisions were requested.
The masters were delivered electronically to Rhino Records via a secure
DigiDelivery server. My workflow and mastering approach was established in
consultation with Jeffrey and David during some initial test sessions, and
variations on that basic signal flow were used for the entire 73-disc
project. And yes, I did listen to every note. Several times!
AR: What is your take on the “Loudness
Wars” issue? How do you handle clients who want to adjust/reduce the dynamic
range for MP3s?
DG: “Loudness” is something
that is usually discussed with the client for every project. Ultimately it’s
the client’s project and his or her decision. My job is to give them the master
they want, and if they insist on a heavily compressed master, so be it.
However, things are looking brighter. Apple has just released their
“Mastered For iTunes” guidelines, which contain 2 significant
recommendations. The first is Apple’s acceptance of high-resolution files (up
to 192k 32 Bit) for submission to -the- iTunes.
Now Apple is not selling high-resolution
audio (yet), but they acknowledge that their AAC encoder works best when fed by
a high resolution source. The second important recommendation is to allow
adequate headroom in iTunes masters. Lossy codecs can produce levels exceeding “0”
resulting in audible distortion. Apple has released some software tools (Sonnox
also has a similar software plug-in) that can help to optimize masters for
better AAC encoding. Apple is also saying what we have all long known: highly
compressed music sounds worse when converted to MP3 or AAC.
AR: What’s your latest project? Why and
how is it challenging?
DG: I don’t have anything as massive as
last year’s Europe 72, but I’m working on a couple cool things. Look for a CD
from the Bruce Kaphan Quartet this spring. Bruce is a pedal steel player
(Sheryl Crow, Thomas Dolby, David Byrne) and this record is a very cool
jazz-fusion project that’s unlike anything I’ve heard. Also, I worked on the
recently released Dave Carter/Tracy Grammar CD “Little Blue Egg,”
which contains unreleased songs and demos recorded before Dave’s passing.