Callas Remastered: "La Divina" Receives Her Due

Move over John, George, Ringo, and Paul. There's another remastering that's come on the scene, and it's every bit as important as the Beatles Mono Edition. It's Warner Classics' high-resolution, 24/96 digital remastering of soprano Maria Callas' entire studio-sourced discography. Consisting of arias, recitals and complete operas recorded 1949–1969, the remasterings reach the international public on September 22, and US music lovers on September 23. Their sound, whether in the 69-CD box set of her entire studio recordings, or HDtracks' 24/96 downloads of its individual components, is revelatory.

Ever since their initial issue on LP, Callas' analog recordings have suffered through numerous digital reissues. Although the last set of her complete studio recordings, which became available in 2007, was touted as definitive, it was only achieved in 16/44.1 sound. In addition, it doctored the mono and stereo originals with the noise-suppressing ART processes, compressed dynamics, and attempted to improve the sound of some of the drier analog originals via the application of artificial reverb.

For this definitive 24/96 remastering, accomplished at the famed Abbey Road Studios with monitoring performed on Bowers & Wilkins 800 series loudspeakers, Warner Classics' three sound engineers returned, whenever possible, to the original master tapes. Though re-equalization was occasionally employed—Robert Gooch, balance engineer for the original LPs of Callas' Il barbiere di Siviglia and stereo Norma recordings, acknowledges that the mastering monitors he used at the time were "rubbish...some commercial speakers"—virtually all of the post-production digital doctoring that degraded previous remasterings was jettisoned. Where mastering instructions were included with the original tapes (often concerning the reduction of overly intrusive street noise), they were respected. [You can view a video that includes Gooch and remastering engineers Andy Walter, Allan Ramsey, and Simon Gibson here .

The results are ear-opening. In advance of the 69-disc box set's release, and staggered release of its individual components, Warner supplied members of the press with a CD that compares the last 16/44.1 remastering of five arias with the new 24/96 versions downsampled to 16/44.1. Even on the Liebestod from Wagner's Tristan und Isolde, released on 78 shellac in 1949 when Callas was not yet 26 years old, the change of source results in far better sound. Callas may not have sung Wagner onstage past 1950, but when you hear the core of her voice open fully toward the end of Isolde's rapture, you will experience as never before the unforgettable fusion of love and suffering that was Callas' art and life.

Once we get to material remastered from the analog tape masters, the restoration of the treble edge of Callas' voice and high-frequency extension of the orchestra, combined with palpably greater richness and body, sounds as though a previously clouded window on Callas' artistry has been scrubbed clean. The differences are astoundingly evident on an excerpt from the 1952 recording of Ponchielli's La Gioconda, and quite noticeable on her 1953 Tosca (one of the greatest opera recordings ever made), 1954 I Pagliacci, and even the 1961 stereo Callas à Paris recital.

Need I even say that the 24/96 downloads of the box set's contents, which appear on HDtracks in three batches (September 23, October 21, and November 11), are the best we're going to get? (Yes, Warner could have chosen to remaster in 24/192 or DSD, but that is not the case, and it's not likely to come to pass, period.) Depending upon the decisions made when the analog masters were first transferred to LP, some of these new remasterings are likely more faithful to the engineers' intentions than the original issues.

ARTICLE CONTENTS

COMMENTS
audiolab's picture

.....this is the beginning of many more such ventures. Hopefully for me Dame Joan Sutherland will be next to receive such treatment. It is a shame that Sutherland/Calas fans are always at odds with each other. I am quite happy to admit she is not my cup of tea at all, but happily acknowledge she maybe someone's cup of tea. The Decca Sutherland recordings are pretty good anyway, to have an honest remastering could prove revelatory. Here's hoping.

Rick Tomaszewicz's picture

...it'll sell well in Europe. Here in the "new world" we're often suspicious of and dislike anything too sophisticated. We generally prefer common culture. Opera's hard to understand, an acquired taste, requires patience, is expensive and suggests you're trying hard to impress. Too bad. With enough label support, convenient formats and press coverage perhaps more people will give it a try. Those getting bored with pop music, who have enough intellectual curiosity to explore something outside their own tribe's tunes, may be surprised to find themselves emotional at the end of a great opera. Welcome to western culture's highest achievement. Mountains aren't supposed to be easy to climb.

Jason Victor Serinus's picture

This is THE Callas DVD every opera lover must see:
Maria Callas At Covent Garden, 1962 & 1964
http://www.arkivmusic.com/classical/album.jsp?album_id=74679
Filmed in a concert setting, the Act II Tosca with Gobbi and Cioni is one of the greatest live opera performances ever recorded.

Patrick Butler's picture

What a wonderful article. Thank you Jason.

Greg Mitchell's picture

I do wish writers would check their facts. The 1958 walk out was at the Rome Opera, not La Scala, the Rome Opera clearly at fault for not providing a stand-in, a fact clearly established in a later court case.
Her last performance in an opera was indeed Tosca at Covent Garden, but she completed the whole opera. She had been scheduled for three performances, but ill health caused her to cancel all but one, a royal gala in front of the Queen. Considering the parlous state of her voice at that last performance, the decision to cancel all but one was a decision well taken.
A few weeks earlier she had been forced to abandon after Act I a performance of Norma at the Paris Opera due to ill health.

volvic's picture

I concur a well written article, thank you Jason.

Jason Victor Serinus's picture

Some readers have undoubtedly noted that the new Callas 24/96 remasters have been available in Mastered for iTunes format for some time now. There have been many claims about "Mastered for iTunes," including that it often sounds better than the actual CD release. Here is mastering engineer Robert (Bob) Ludwig's take on Mastered for iTunes. Note his final comparison to 24/96 downloads. The first of the three batches of Callas 24/96 downloads becomes available on HDTracks.com on September 23.

A well-done, auditioned and A-B'd Mastered for iTunes (MFiT) file CAN sound better than the 16bit CD some times, but here is the catch:

1. Apple has software (afclip) which measures both on-sample and inter-sample clips. The record companies request that masters we submit for MFiT have either zero clips, or the ones that are there have been auditioned and approved as inconsequential (inaudible). This is done by lowering the level into the encoder, typically -1dB. Apple is not the clip police, and there are MFiT examples I can show you that have a quarter-million clips in 3 minutes.

2. Apple asks that the engineer A-B the MFiT file with the 24bit master and determine the most accurate encoding level.

3. Even some top mastering engineers I speak to do not do this, they do not listen! They merely have their assistants submit a 24 bit file for Apple that has no clips on it.

4. While the top end usually sounds decent, the AAC and MP3 encoders window the bass frequencies in a way that even a few extra 1/10ths of a dB can dramatically increase the space and "air" around the kick and bass instruments. It is ironic that the bass is more of an issue than the top end.

5. I routinely have my engineers prepare a listening playlist on the workstations with the original 24 bit master and AAC encodes at -0.7dB, -1, -1.2 and -1.5dB and A-B EVERY album I master for iTunes. Assuming, for example, that the afclip shows no clips (digital "overs") when lowering the level into the encoder -0.5dB I often find that the bass frequencies and everything else sounds more accurate by further lowering the level -1.2 or even -1.5dB into the encoder, and that is what we send to Apple. It is like focusing a camera.

What I'm saying is, it is possible the 'average' MFiT file one compares to a CD may not sound as good, but if an MFiT encode file is carefully made, I think you will find, as my audience voted at two different AES presentations I made, the MFiT file sounds closer to the 24 bit source than the CD.

Google: 'Apple mastered for iTunes' and you will find the encoding applications Apple has made available to ANY-one and you can experiment and prove it for yourself.
I'm looking forward to Apple releasing 24bit iOS devices.

Of course, none of what I'm saying deals with 96kHz and above which improves the sound further by putting the all-too-audible low-pass filters up an octave. MFiT can be closer to the 44kHz 24 bit master that makes the encodes, but seldom to a 96kHz /24bit master.

Bob Ludwig
Gateway Mastering Studios, Inc.
Portland, ME

bernardperu's picture

Ludwig's thesis is unclear. He seems to say that mastered for Itunes can sound better than CD-lossless. But only if properly done. How about a CD that has been properly mastered? It has to sound better than the lossy mastered for itunes version, right?

prof's picture

He seems to say that mastered for Itunes can sound better than CD-lossless.

That does seem to be what he's saying.

How about a CD that has been properly mastered?

From the text, he's clearly talking about his own CDs which, I assume, he took similar care to master as "properly" as he could.

It has to sound better than the lossy mastered for itunes version, right?

What he seems to be saying is that the greater dynamic range of MFiT offers enough of an advantage over 16bit Redbook CDs that it outweighs whatever downsides there may be to AAC encoding.

Don't get hung up on "lossy" versus "lossless." Every digital encoding (and, for that matter, every analog recording) loses something. It's all in the tradeoffs ...

X