Columbia Master Book - Introductions
Columbia Master Book, Volume I:
U.S. matrix series, 1 through 4999, 1901-1910
This is the first in a series of volumes documenting the disc recordings of the Columbia Phonograph Company. It is the product of a very long journey, and one that is far from over. Research began at Columbia in the 1950s when company librarian Helene Chmura initiated the compilation of lists of recordings made between 1901 and 1934, drawn from company files which no longer exist. These typed lists contained basic information about the recordings: matrix number, title, artist, a production (not recording) date when available. There was limited data on releases, no information on takes or remakes, and many gaps, but, at least, it was a start.
Miraculously, copies of these fascinating lists were made available to a small group of collectors and researchers who started to "fill in the blanks." Copies were made of the copies, and the "Chmura Lists" eventually circulated to many hands. Portions were even published in Record Research magazine in the 1970s. Because of the vast amount of missing information, however, no one attempted to complete the project, much less get it published.
Around 1970 this writer and several colleagues became interested in Columbia, which seemed to be receiving little attention compared to Edison, Victor and other companies. For a time, there was a “Columbia Quartette” of researchers working on the label, then a "Columbia Trio." Between 1975 and 1979 I published several articles about Columbia (see bibliography); one, a guide to matrix series, will be found elsewhere in this volume. Concurrently, William Bryant and I determined to complete the earliest and most difficult series listing, masters 1-4999. Our goal was to include not only matrix, title and artist, but all known takes and issues on any label. Columbia leased its masters to a plethora of labels. By the 1970s no files remained to document any of this, so all remaining information had to be reconstructed from original copies and catalogs. Even starting with the limited framework of the Chmura Lists, it was a daunting task.
Information was first gathered from our own collections and from my extensive library of catalogs and supplements. Then Bill took over with the goal of scouring other collections. Years rolled by as collection after collection was checked and correspondence carried on with researchers around the world. In the early 1980s English discographer Brian Rust joined the project; he extended the listings from 1910 to the early 1930s, using microfilms he had obtained of the Columbia files, which are more complete after 1910. [end p. xi] Greenwood Press expressed interest, and a tentative publication date was set for 1984.
Many delays and diversions ensued however work never stopped entirely. In the early 1990s I began to organize the huge amount of data accumulated for this volume and to prepare it for entry into a computer database. Unfortunately, Bill—a confirmed computerphobe—had barely begun his portion of this work when he died suddenly, in September 1995. The executor of his estate, Stephen Harding, graciously agreed to return his portion of the files so that the project could continue. Much has happened since then. New York collector and researcher Richard Markow pitched in to complete the difficult data entry work, the Association for Recorded Sound Collections generously provided a grant to help defray costs, and numerous experts willingly answered my blizzard of letters and email messages, adding an enormous amount of new data. More than a hundred people, named in the acknowledgments, have contributed in one way or another to the discography you see.
I believe the pre-1910 listings are now virtually complete for issued, domestic, English-language titles. Approximately 3,400 titles and 15,700 releases on 34 labels are documented here. However, the work is certainly not over. Many unlisted takes and alternate-label issues lurk in collections around the world. Some ethnic titles (pressed in very small quantities in the ethnic and export series) are still to be discovered. Many of the missing titles are probably unissued masters that will never be identified. To achieve 100% completeness it would be necessary to inspect every disc issued or licensed by Columbia during this period—in fact, every copy of every disc. Rather than wait for that, it was felt best to publish the large amount of information gathered so far. Readers are urged to check their collections and send any new information to the author, in care of Greenwood Press.
In the course of this research a great deal of information has been gathered about other Columbia series, including trials, personal and foreign recordings. If support for the project is sufficient, it is hoped that this may be published at a later date.
The manuscript was prepared using WordPerfect 6.1, Times Roman font.
Although many hands helped gather this data, the undersigned compiled and edited the material and wrote the accompanying text, and he accepts responsibility for the final result. Corrections and additions are, of course, welcome. Only through a joint effort of collectors and archivists will Columbia’s remarkable contributions to recorded sound be fully known.
Tim Brooks, January 1998 [end p. xii]
Columbia 10" Matrix Listing
Columbia Master Book, Volume II:
Principal U.S. matrix series, 1910-1924
Compiled by Brian Rust
In the autumn of 1910, the system of numbering 10” masters recorded by Columbia for domestic use had reached 4999 (see Vol. 1), and it was decided that the next block of numbers to be used should start at 19100. When this reached 19999, the next block began at 38100, to continue to 39999. This was followed by a series covering numbers from 45500 to 47499, after which the largest block so far allocated began at 77000 and ended at 81999, the point at which the present volume ends.
Within a year, electric recording was a fact, at least for the "flagship" label of Columbia (its low-priced Harmony subsidiary, which also was established within that year, continued to be acoustically recorded for some three or four years). The striking "flag" Columbia label, introduced in the autumn of 1923, also vanished soon after the end of the present survey; but during the years 1910 to 1924, covered by that survey, the quality of Columbia recording, and the general physical appearance of the records had improved quite outstandingly. There may not have been the extensive roster of "classical" works on 10” Columbia records compared with Victor's, and such concert artists as were under contract seemed called on to record what many would regard as no more than trivia; but in the popular field. Columbia had shrewdly noted the enormous success of black artists recording for OKeh, and were very much aware of the growing market in rural areas for "old-time" and purely folk music—so Columbia supplied it. (Victor made very little attempt to provide anything comparable either in quality or quantity until Columbia and OKeh had been mining this rich source for three or four years.)
The original file cards still exist, and it is upon them that this book is based. All takes made are shown as these cards show them; takes used for issue are underlined. In many cases, two or even more were selected, and are known to have been issued to the public. Although the cards sometimes assert that a recording was rejected, evidently there was a change of mind in the Columbia offices, and pressings were put on sale despite initial rejection. Personnel of accompanying groups are shown if the file cards show them. Pseudonyms and other changes of identity are shown in the same way.
As with most companies, there are examples of recordings having been made for domestic use in America, and which were never issued there, but [end p. 1] which did appear in the British catalogues, sometimes on Columbia, sometimes on its lower-priced subsidiary Regal. The catalog numbers of such expatriated recordings are shown in this text; but if an American issue was made available abroad, only the domestic number is given. It is not possible to determine from catalogs or other publicity material whether a Columbia (or Regal) record is necessarily of American origin in every case without actually inspecting the record in question, and there are examples where both an American and a British recording of the same title were issued under the same catalog number and the same artist credit.
All sessions were held in New York, unless otherwise stated.
Columbia Master Book, Volume III:
Principal U.S. matrix series 1924-1934
Compiled by Brian Rust
This, the third volume of Columbia records made in the USA before the original company was absorbed into the American Record Corporation in 1934, covers the years from 1924 to the end. In those years, great changes took place in the world of recording—and outside; the actual recording technique changed early in 1925 from the rather hit-or-miss method that employed long tapering horns that conveyed the sound waves from the artists to the thick wax block, to a microphone that enabled those sound waves to be extended over a wider spectrum than had been a recording technician’s dream for the preceding three decades.
Within three years, silent films had found a voice, and with them had come musical films that featured songs that every dance band, from Paul Whiteman and Ted Lewis (both had special labels for the Columbia records they made from 1928 onwards) to the newly appointed Ben Selvin as Musical Director, and their orchestras were recording as soon as it was possible for copyright clearance to be obtained. Along with these came the "hits" from Broadway shows that became more and more lavish, not to say exotic, each year. The output of records increased, visits to Atlanta, New Orleans, Dallas and Los Angeles became events of twice yearly occurrence, when recordings of jazz, blues, spirituals, and what were regarded as "country-style" songs and dances were made.
Then came the Wall Street disaster of October 1929. The demand for records of all kinds began slowly but steadily to dwindle down to a mere fraction of what it had been in the boom years of the 1920s. In 1928, many dozens of titles of all kinds were recorded every month; in 1932 through 1934, there were months when one session was held—sometimes none at all. There were no more road-trips after 1931; soon the "race" and "country" series of records were abandoned, with a few very carefully-chosen items being issued on Columbia’s OKeh label, acquired from the General Phonograph Corporation in November 1926, in its long established, but soon to be discontinued series of similar issues.
It was in the decade covered by this volume that the dance bands, many firmly established in popular esteem, began featuring vocalists—some being members of the bands, some engaged by the leaders simply as singers. Bing Crosby’s first recording was made for Columbia, although anonymously, and his first as a named solo singer followed on Columbia just over two years later. The original files are somewhat inconsistent in the naming of these vocalists, often showing them as anonymous “vocal refrain” artists, but sometimes when they are named, there will be some surprises for those whose interests lie in this direction! A curious procedure can be seen in the naming of vocalists on the Harmony label, introduced in the fall of 1925 at 35 cents (compared with the 75cents price asked for Columbia records): the same pseudonym was used on records by more than one singer, and one singer on a four-title session might be re-named two or even three times. One such was Irving Kaufman, whose tenor voice and perfect enunciation made him an ideal performer on acoustic and electric recordings alike, who appeared as Tom Frawley, Ned Cobben, Arthur Seelig, Robert Wood and Marvin Young, as well as himself. (His wife is said to have jokingly asked him who he was going to be that day. "I don’t mind," said the cheerful Mr. Kaufman, "as long as my real name is on the check.")
Some recordings were never issued in the United States, but nevertheless were added to the British and Australian Columbia catalogs or to their Regal or Regal-Zonophone lists. The series found on records shown in this book are as follows: Table 14a.
Columbia Master Book, Volume IV,
U.S. twelve-inch matrix series, 1906-1931
Brian Rust and Tim Brooks
Volumes I through III of The Columbia Master Book document 7" and 10" disc recordings made by Columbia between 1901 and 1934. Volume IV covers the 12" format, from its inception in 1905 until domestic recording in that format was suspended in 1931, during the Depression.
This volume is divided into two sections, covering 1906-1912 (the 30,000 matrix series) and 1912-1931 (matrix 36350 and higher). The two sections employ somewhat different conventions, in part because of different Columbia practices during the two periods, but also because they were compiled by different authors on two sides of the Atlantic. We trust that these minor inconsistencies will not inconvenience the reader.
A detailed introduction to Columbia's history, its matrix series and numbering conventions, is found in Volume I. Full acknowledgments will also be found there, however I would be remiss if I did not express my sincere gratitude to Brian Rust, with whom I have worked on this project for many years. He is one of the giants of discography, as well as a true gentleman, a rare combination indeed.
Part I: 1906-1912
The first Columbia 12” disc was released in June 1905, numbered 3141 in the regular 10” domestic series. After two additional 12” issues in that series (3358, 3377), a new numerical series was begun for the format in 1906, starting at 30000. (The first three issues were renumbered into this series, as 30000-30002). Thereafter all Columbia 12” recordings had their own matrix and catalog series.
In addition to the normal matrix number, some Columbia pressings from 1904-1908 bear an “M” number in the wax, whose purpose is not certain (see Volume I for further discussion). M-numbers, where known, are shown here in brackets following the title.
All dates are from lists compiled in the 1950s from the Columbia files. Although these lists call them “recording dates,” their meaning is not always clear. In some cases they may be true recording dates. In most cases they are believed to reflect the date on which the original master was shipped from the recording laboratory to the factory for processing, which was normally a day or two after recording, but occasionally later than that. It was at this time that the matrix number was assigned. If a date is much later than those of surrounding masters it probably refers to a later take (see, for example, mxs. 30747-30750, released in August 1911 but “recorded” on April 4, 1912!).
Other dating anomalies occur within the first few hundred numbers, including dates that are much earlier than those of surrounding masters (e.g., 30132-30137, 30210-30211). These presumably reflect recordings made and shipped at the earlier date but for some reason not assigned a matrix number until much later.
As an aid in dating recordings made during this confusing period, the month of release is shown after the earliest issue in the right-hand column. The reader is advised to use all dates with caution, and to study both release dates and surrounding matrix dates in order to estimate when a given matrix and take was made.
During this period Columbia leased its masters to numerous smaller labels, and a great many were released by the company’s English affiliate. In fact, some matrices or takes were only released in England. Because early Columbia matrices may appear on many different labels, all known 78 rpm-era releases are shown in this section.
Only fragments now remain of the original Columbia files from this period, so much of the information on the 30000 series had to be reconstructed from catalogs and copies of the records. Additional information from readers is solicited, in care of the publisher.
Part I was compiled and edited by Tim Brooks, based on his own research and that of the late William Bryant and Helene Chmura. Other contributors are gratefully acknowledged in Volume I. Production assistance was provided by Rich Markow and David Giovannoni, to whom sincere thanks are due.
Part II: 1912-1931
By 1912 Columbia 12” discs were being produced on a regular basis, using more regularized procedures. During the boom years of the 1910s, in fact, some were best sellers, rivaling the sales of 10” discs.
Fewer dating anomalies are found, and the dates of remakes are sometimes known. However there are still minor inconsistencies in the files. Slightly different dates are sometimes given in different places for the same matrix (e.g., on the matrix card and on the artist card for the same recording). No recording ledgers exist to resolve these discrepancies. The date shown here is generally the date the master was shipped to the factory for processing; not until the late 1920s did the files begin to routinely specify a recording date. Again, readers are advised to treat all dates as approximations of the true recording date.
The Columbia files covering this period are divided into three parts: release cards, artist cards and matrix cards. The release cards show the matrices assigned to each catalog number, and when that disc was released. There are usually no takes shown, and cards generally exist only for Columbia, not for other labels. The artist cards summarize the issued and unissued recordings made by a given artist. They are often incomplete, and contain little detailed information on the recording sessions, being basically a cross-reference. Few release or artist cards exist prior to 1910, and some are missing after that.
The most detailed recording information is found on the matrix cards, which show for each matrix number the takes made, the date each take was shipped to the factory, and its disposition (released, rejected, etc.). Unfortunately actual recording dates were not shown on these cards until the late 1920s (they very occasionally appear on earlier release or artist cards). Original copies of these valuable matrix cards survive for 10” masters from about 1910 onward (19,100s and up), however for 12” recordings, very few exist prior to 1921 (98,000s). Therefore the take information shown in this book prior to that date was obtained from actual discs inspected. Throughout the book, takes known to have been issued (or in the 98,000s, indicated in the files as issued) are underlined.
Columbia ceased leasing masters to other labels in the early 1910s. From this point on virtually all releases are on Columbia itself, so the label name is not shown in Part II unless it is other than Columbia. Foreign releases are not shown unless they represent the only release of a given master. Thus in most cases English Columbia issues are not shown. Various Spanish-language “export” series (C1000s, H1000s, S5000s) do turn up in the U.S., especially in Spanish-speaking regions, and are shown, as is the regular domestic U.S. foreign-language series (E5000s). Additional information on these obscure series is welcome.
Readers wishing more information on British issues of U.S. Columbia masters are referred to the excellent Columbia Twelve-Inch Records in the United Kingdom, 1906-1930 by Ronald Taylor (East Bamet, UK: Symposium Records, 1994). Columbia’s domestic foreign language recordings are listed in Ethnic Music on Records by Richard K. Spottswood (University of Illinois Press, 1990).
Many operatic recordings were released in single-faced format during the 1910s and early 1920s. Single-face discs, which used the matrix number as the catalog number, are shown in the issues column where applicable.
Part II was compiled and edited by Brian Rust, based principally on data from the Columbia files, with supplementary information and indexing by Tim Brooks. Production assistance was provided by Victor and Jane Rust, Tim Brooks and David Giovannoni.
The Columbia Master Book Discography, 4 Volumes, Complied by Brian Rust and Tim Brooks. Reprinted by permission.