I don't know about you but I've always been curious about how Mosaic Records remasters recordings for its box sets. A long-time fan of Mosaic's restorations and attention to detail, I favor the "Producer's Note" that appears at the back of each set's calendar-size liner-notes brochure. Written by Scott Wenzel, Mosaic's award-winning producer, the note often tosses around terms like "metal parts," "lacquer discs" and "second generation LPs." After listening to Mosaic's new Classic Artie Shaw Bluebird and Victor Sessions, I decided enough was enough. I was going to get to the bottom of the process.
My first call was to Scott Wenzel [pictured], who told me how Mosaic's projects begin:
"The entire team here meets to discuss project ideas and the availability of original source material. Once we agree to move forward on a project, we work out a license agreement with the company that owns the masters. Then I research the discographical data and begin to create a master list of each track and the various known takes. I also assign the liner notes to a writer. When I've completed the list, I reach out to one of our superb transfer, sound restoration, and mastering experts to handle the project."
In the case of the Artie Shaw box (and many other Mosaic projects), that person was Andreas Meyer [pictured], who runs Meyer Media. So I called Andreas to find out what steps took place:
JazzWax: Did you simply drive over to Sony's RCA vault and pick out what you needed?
Andreas Meyer: [Laughs] Not quite. Mosaic's Scott Wenzel and I have relationships with the companies that own master recordings. Once Mosaic’s team decides on an artist project and Scott has pulled together the artist discography and track names, we assemble a list of matrix numbers. In the case of Shaw, the period was 1938-1945 and the label was RCA Victor. Then Scott sent over the matrix list to Sony BMG requesting all of the metal parts and test pressings for each Shaw recording we wanted. Sony owns RCA's masters.
JW: For those who don’t know, what is a matrix number?
AM: A matrix number is the record label’s internal serial number. It’s like the Dewey Decimal System for records. The number is an antique system that allows companies to know what they have and find what they need. For example, the matrix number for Shaw’s To a Broadway Rose is #068194. Scott had to research and assemble a matrix number for each single Shaw recorded during the period Mosaic wanted to cover.
JW: But each song recorded may have several takes.
AM: That’s right. And here’s where the magic of finding alternative takes comes in. A recording’s take appears after its matrix number, with a dash. So the first take of To a Broadway Rose was #068194-1. Sometimes multiple takes emerge during this process that you never even knew existed.
JW: How do you know in advance how many takes of a song were recorded?
AM: There are session logs documenting every recording session scheduled. These tell us how many takes of a single title were recorded. But there were mistakes and clerical errors. Time was money, even back then, and when you need to get so many tracks recorded in the allotted studio time, the pace was quick and keeping track of all those takes was difficult. Then there are rejects—sometimes a whole session's worth. These were to be destroyed, but they often found their way out of the studio and onto the company's library shelf. Many of the label's clerical errors have been corrected in very thorough discographies over the years. But we still see discrepancies or new titles just appear.
JW: Once Sony BMG has your list, what happens next?
AM: A Sony archivist pulls the masters from the vaults. The RCA vault is a mile down in a mineshaft in Pennsylvania. The masters are stored in their original sleeves or archival boxes with just the matrix numbers written on them and now a bar code. We ask Sony to pull everything they have with the different matrix numbers. Then they ship all of it to me.
JW: Somehow I imagine a crate the size of an elephant filled with straw, like in Tarzan.
AM: [Laughs] The Shaw material came in 15 to 20 boxes that weighed 60 pounds each. Each box had 10 to 20 different titles, and each title had metal parts and test pressings.
JW: Why do you want everything?
AM: Because you never know which is going to give you the best sound and fewest flaws during the analog-to-digital remastering process.
JW: How are they packed to protect the contents?
AM: The discs are in their archival sleeves or boxes, and the metal parts from the RCA catalog have a wool liner. These are packed with huge amounts of bubble wrap and then fit into a box lined with more bubble wrap.
JW: For those who don’t know, what’s a metal part?
AM: They're just metal discs. When these Shaw sides were recorded in the late 1930’s and 40’s, they were recorded first onto a wax disc. The disc was softened a bit with heat, which allowed a special stylus to cut out the groove as the musicians played and the wax disc turned.
JW: But when the record was finished, the wax disc was too soft to play or press.
AM: That’s right. So they would take the wax disc and make a negative metal master of it.
JW: How did they do that?
AM: Through a plating process. What they did is submerge the wax disc and a metal disc in a chemical bath. The process electrostatically created a reverse impression on the metal disc.
JW: That would mean raised ridges on the metal disc's surface instead of grooves, yes?
AM: Yes. When the disc came out of the bath, the affixed metal “mountains” were covered in wax. Once the wax was peeled off, RCA would have a negative metal “father” master. Of course, you needed to create a positive metal part, or a master with the grooves. So the original "father" disc and a new metal disc would go into another chemical bath to produce a positive "mother."
JW: When the "mother" emerged, what then?
AM: The “mother” could be put on a standard turntable and played. But the “mother” wasn’t created to be played but to serve as the master for “the stamper.” A metal stamper was created through yet another bath process.
JW: What did they do with the stamper?
AM: In many cases, multiple stampers were created and shipped to the presses around the country. The reason they went through this process was for longevity. You always wanted to protect the "father." That’s your true master and the version that’s closest to the original.
JW: Then what's a test pressing?
AM: Record companies made test pressings in order to listen to what they had before they released it. This way they could be sure they had an accurate picture of the recording and decide if it was good enough for release.
JW: They could have used the "mother" for this, no?
AM: Yes, but the less handling and use of a metal part back then, the better. So they'd make test pressings before actually going to press for mass production. Once LPs came on line in 1950, RCA started to reissue 78-rpms on 10-inch LPs. The engineers at the time made PVC [polyvinyl chloride] or vinyl test pressings directly from the "fathers" and the stampers.
JW: How do they sound today?
AM: The test pressings are often some of the best sources for reissue today. I think this is because vinyl was a much softer material than shellac, which was used for most of the test pressings before 1950. Vinyl captured more sonic information.
JW: So you hauled in everything Sony BMG had on Artie Shaw between 1938 and 1945?
AM: Yes, I always request everything, and a lot of times we get the metal “father” masters as well, which is kind of interesting.
JW: But what can you do with a father, since it has raised ridges rather than grooves.
AM: I actually have a turntable that can play them. In fact you need special turntables and needles for the "father," "mother" and test pressings to maximize the signal. [Pictured: Andreas Meyer at work on a metal part]
JW: Were the Shaw parts in good shape?
AM: Over the years, RCA Victor took very good care of their metal parts—"mothers," "fathers" and stampers.
JW: I assume you’re not ordering these special turntables on Amazon?
AM: [Laughs] Half the remastering industry uses vintage gear that has been reconditioned. The other half uses brand new customized gear. In fact, I just got off the phone with a technician who’s building me a reconditioned Technics turntable I had designed to play forward and backward, which plays negatives. This turntable [pictured] has a speed range from 30 to about 96-rpm. It will be placed in a slate stone plinth from Oswalds Mill Audio. The plinth is large enough to house two tone arms—one for positives and one for negatives. Then all my styli are custom built in a large range of shapes and sizes.
JW: What was your biggest challenge with the Shaw box?
AM: Keeping the titles organized. There were over 300 masters and alternate takes. That’s an enormous amount of material. You have to move with extreme care to ensure that titles and numbers don’t get mixed up.
JW: What happens during the remastering process?
AM: First I had to make sure the part and stylus I chose was going to produce the best transfer to my computer. Once I've chosen the best metal part or pressing, I'm down to esthetic decisions. It’s a question of how far do you go with the restoration and the mastering process.
AM: You can make a mistake by not touching the original quality at all and winding up with all the noise and flaws. Or you can err on the other side by cleaning up a recording to the point that it sounds artificial. Scott and I usually discuss the right balance.
JW: What’s your preference?
AM: Usually I try to retain the sound of the recording the way it was when the disc was issued at the time—in mint condition. I don’t want to take a recording out of its historic
JW: As a jazz fan, was it great listening to the Shaw metal parts?
AM: Amazing. It was like having Artie Shaw’s band in the room while I was working on the box. The fidelity was that rich.
JazzWax tracks: The Classic Artie Shaw Bluebird and Victor Sessions box from Mosaic Records is available here.
JazzWax clip: To see the RCA Victor record-making process in 1942, go here for Part 1 and here for Part 2. Astonishing what record companies went through to make your foot go up and down. Talk about a labor of love!
Marc,
As an an audiophile as well as a lover of jazz (especially in the vinyl LP format), your interview with Mr. Meyer was most interesting. Thanks to both of you!
Posted by: Doug Zielke | November 03, 2009 at 09:05 PM
Thank you very much for this great article. This is by far one of the best music articles I've read this year.
Posted by: Rab Hines | November 05, 2009 at 12:34 AM
Several corrections to Andres Meyer. When he likens the matrix number of a recording to the Dewey Decimal number of a book he has misunderstood what the Dewey number is. It only identifies the category of the book, not the book. Thousands and thousands of books can have the same number. This would mean that nearly every jazz record could have the same number because most every jazz book has the same DDN with slight modifications to whether it is a history, biography, discography, etc.
He is also wrong when he describes the electroplating process as "submerge the wax disc and a metal disc in a chemical bath." The metal is part of the liquid chemical bath, not in disc form. The metal molecules are formed into a disc only when they adhere to the surface of the wax disc. They do not start with a metal disc.
There is a photo error at the description of recording on wax which is softened. The photo shows the creation of a "Flow Coat" which is melting a layer of wax onto a heated metal plate. That is a different process than what he describes, which is recording on a thick solid wex platter. Victor was using BOTH of those different systems at that time, as well as also using lacquer coated discs.
Posted by: Michael Biel | November 20, 2009 at 05:46 PM
Thank you Michael for that correction. Yes the metal particles in the bath solution electrostatically adhere to the disc - thus creating a metal part. THis process is still used today at "modern" pressing facilities.
Cheers,
Andreas
Posted by: Andreas | April 13, 2010 at 08:35 PM