Association of Research Libraries (ARL®)

http://www.arl.org/preserv/sound_savings_proceedings/Arhoolie-2.shtml

Sound Savings: Preserving Audio Collections

Archiving the Arhollie Foundation's Strachwitz Frontera Collection of Mexican and Mexican American Recordings

Tom Diamant
Digital Archiving Director
Chris Strachwitz Fontera Collection of Mexican and Mexican American Recordings
Arhoolie Foundation

Background

Chris Strachwitz, a young immigrant from Germany, became enamored of the vernacular music he found when he came to the United States in the 1940s. He started collecting Jazz, Blues, and Country 78 RPM recordings and by the late 1950s was also collecting Mexican and Mexican American recordings. The 78 collecting field had many fans of Blues, Jazz, and early popular music, but Mr. Strachwitz was one of the only collectors interested in Mexican music. He therefore ended up with the largest collection of Mexican and Mexican American commercial recordings. As he continued to seek out 78s, Mr. Strachwitz expanded his collecting to include 45 LPs, cassettes, CDs, videos, photos, and all things relating to Mexican music. In 1996 he donated the recordings of Mexican and Mexican American music to the Arhoolie Foundation. The archive continues to grow as Mr. Strachwitz finds more and more recordings.

The Arhoolie Foundation’s Strachwitz Frontera Collection of Mexican and Mexican American Recordings, or the “Frontera Collection,” as we call it, consists of an archive of almost fifteen thousand 78 RPM records, twenty thousand 45s, three thousand LPs, and numerous cassettes and CDs. In all, the Frontera archive totals over 110,000 individual recorded pieces. It contains many rare recordings including one-of-a-kind acetates and test pressings and many recordings of which we have the only known existing copy. This is a unique archive.

These recordings are all commercial recordings, but unlike many commercial recordings of Jazz, Blues, Country, Classical, Popular, and other forms of “world” music, Mexican and Mexican American music has not been reissued much in the LP and CD era. Since the beginning of the recording era, commercial companies have been making Mexican recordings and selling them to audiences in both the United States and Mexico, but even though Mexican Americans now comprise the largest minority group in the United States, their recorded cultural heritage has been all but ignored by modern society.

Cultural Significance

The Frontera Collection offers a valuable window into the development of this culture as it has blended elements uniquely Mexican with styles, lyrics, vocabulary, social attitudes, forms, struggles, and politics that are clearly the result of the American experience. Within the Frontera Collection, this unique cultural heritage is preserved.

The Frontera Collection consists of corridos, canciones, rancheras, boleros, sones, tangos, banda, polkas, light classical, and dialogues, along with a whole range of American popular music interpreted by Mexican and Mexican Americans. The entire collection represents a window into the changes in American and Mexican American culture during the last one hundred years.

Many of these records never achieved significant distribution on a national scale, which in turn helped them maintain distinctive regional styles. This lack of assimilation led to the development of Tejano and Chicano music that continues to produce themes that are uniquely Mexican American. The lack of distribution and/or wider American influence, however, also contributed to the lack of acceptance of this music as a valuable art form. Produced by scores of small regional labels, many of these recordings have virtually disappeared as the labels went out of business.

Steps to Preservation

The Frontera Collection is housed in a specially built building in the Arhoolie Records complex in El Cerrito, California. El Cerrito is located between two major earthquake faults, with many minor faults running between them. We are keenly aware that one serious earthquake or fire could destroy the archive. As soon as the Arhoolie Foundation took over the archive, preservation became a major concern. The Archiving Process

Cataloging

The first step toward preservation was to catalog the archive. This was accomplished with grants from the National Academy of Recording Arts and Sciences, the National Endowment for the Arts, and Arhoolie Productions. A database was created and all the information from the record labels was entered. This information includes the record label, selection title, artist, catalog number, matrix number (the number given at the time of the recording by the record company), song type, and any other information that might be on the label or album sleeve. Now we knew exactly what we had in the collection.

The beauty of some of the 78 rpm labels found in the Frontera Collection

One of the Arhoolie Foundation board members, Guillermo Hernadez, was not only concerned about the preservation aspect but also the accessibility issue. Over 110,000 recordings on the shelf in a building that only two people have the keys to is not a very accessible archive.

In 1999, Mr. Hernandez was the Director of the Chicano Studies Research Center at the University of California at Los Angeles (UCLA). He was interested in obtaining a copy of the Frontera Archive and depositing it in the Music Library at UCLA. This appealed to us at the Arhoolie Foundation. We would make two copies of the collection. One would be a protection copy to archive and make accessible at our facility and the other copy would be for UCLA to make accessible through their Digital Library system. Through their resources, UCLA could make the collection far more accessible to far more people than the Arhoolie Foundation, with our limited resources, could ever do.Through Mr. Hernandez’s efforts and his association with the Noteño superstar band Los Tigres del Norte, the Chicano Studies Research Center received a grant and the promise of future grants from the Los Tigres del Norte Foundation with part of that grant specifically earmarked for the preservation and accessibility of the Frontera Collection. Los Tigres del Norte is a band that knows its musical roots and is very interested in making these recordings accessible to as many people as possible through UCLA. In 2001 the Arhoolie Foundation received a grant to start the project through the Los Tigres Del Norte Foundation, Chicano Studies Research Center, and the Fund For Folk Culture. This grant would finance the creation of a searchable, digital archive including a visual image of the label, a digital audio copy of the recording, and a discography, all accessible through the UCLA Digital Libraries web site.

In the summer of 2001 we started our project. In collaboration with UCLA Music Librarians Gordon Theil and Stephen Davidson, we created a procedure to digitize the collection. We decided to start with the 78-RPM recordings.

Digitizing 78s

78s particularly are a wonderfully stable medium. They have a marvelous weight and feel to them and the grooves are so wide you can almost see the music in them. Unlike digital archival mediums, they can be played back on the most primitive of devices. You don’t even need electricity. A scientist coming across one a thousand years from now could probably figure out how to get sound out of it with out too much trouble. They are a very stable medium, until you drop or melt one. They are fragile in that respect.

We consulted with experts in the sound recording field, the archiving field, the digitizing field, and the data preservation field about digitizing techniques and procedures. We ended up with many varying opinions. Since this was and still is a new field with many different paths to follow, we realized that we needed to take the best advice we could, make our decisions based on our budget and needs, and just go for it. We knew that although we were a small foundation with a small budget, we could do an excellent and efficient job with careful, thoughtful planning and purchasing.

We decided to make a digital copy of the collection rather than copy it to analog tape. By digitizing it, we could create greater access to the archive, and after decades of experience in the recording industry working with tape, we quickly ruled out the unstable and obsolete medium of reel-to-reel tape. We would record directly onto the computer's hard drive.

Although those of us at the Arhoolie Foundation had great experience in recording 78-RPM records, we realized we need professional help when it came to digitizing them. With the help of Advanced Systems Group of Emeryville, California, a professional audio company, we put together a digitizing studio using the highest quality equipment we could afford.

Visual Documentation

We scan a high-resolution color image (at 300 DPI) of the 78 labels into the computer as well as make a recording of the 78. The scan includes the area around the label to the run-out groove. Often the matrix number, stamper number, or other information is inscribed in the record in this area. This allows the end user to not only hear the recording but also view a scan of the actual artifact and zoom in to read even the smallest of type.

Scanning the 78 label into the computer

With this image on the computer monitor, we compare the information with the information in our already existing database and correct any errors. Preparing the Originals

We then wash the 78 with a specially built record-washing device. For some of the dirtier recordings, we use dish soap and run them under the faucet, taking care not to get the label too wet. The 78 is now ready to be digitized.

Cleaning the 78

Digitizing

We digitize the recordings at the resolution 24 Bits/ 96KHz directly into the computer and save them as uncompressed wav files. The higher resolution was chosen because it simply captures a more complete representation of the audio waveform. It has also been suggested that if these digital files lose or have altered any of their digital bits as they follow their technological, migratory path, there is less chance for deteriation of the audio signal at the higher sampling rate.

To get sound into our computer, we needed the best analog to digital converter. This takes the analog sound and changes it into a digital signal that the computer can capture. Our biggest single expense was the external analog to digital converter. (See the Appendix for the equipment list)

Slating

Upon recording each side of a 78, we created individual computer files of each recording in the WAV format. We decided to verbally slate each recording at the beginning by reading the Label, Catalog Number, Matrix Number, Title, and Artist. This has proven to be both a great help and a bother. It is helpful since each individual file has the complete label information spoken at the beginning. Even without the discographical database, the recording is identifiable. When listening to a series of recordings in a row, it is also helpful to hear what each one is. However, if you are quickly going through the recordings using the interactive database, it becomes tiresome to hear the spoken slate. Fortunately, you learn how to skip over it.

Data Documentation

Since we started this project we have upgraded our software to use the Broadcast Wave Format. This allows for all the information to be added to a text header in each individual sound file. We now not only list what information is on the label but the metadata of all the equipment used in the digitizing process right down to the stylus size.

Recording the 78

Selecting Styli and EQ Curve

Our digital archivist tests the 78 with different diameters styli before choosing which one achieves the least noise and the best original sound out of the grooves. We use a pre-amp that can be adjusted to recreate a variety of equalization curves. There was no standard eq curve for mastering 78s. Play it back with the wrong eq curve, and it can muffle the sound or make it shrill and unnatural. Some eq curves are known for some of the major labels, but for the smaller, independent labels, and many of the major ones as well, careful, knowledgeable listening is the only way to know what setting to use.

The most important qualification for a digital archivist is a good set of ears. We hired a musician, Antonio Cuellar, not for his knowledge of formal archiving, but for his knowledge of what musical instruments and the human voice should sound like when they are played back through a good pair of speakers.

Keywords

As the recording is being played, the archivist listens and notes keywords from a list we have created. These keywords tell us what the main idea of a song is and which instruments are featured in an instrumental. We have keywords such as labor, migrant labor, revolution, murder, contraband, intoxication, and saxophone. Since most of the songs are love songs, we have many qualifiying words for love, such as praise, betrayal, jealousy, and murder. These keywords will be used by end users to search for recordings under a variety of topics.

Digital Archivist Antonio Cuellar logging keywords into the database. (As of July 2004 Mr. Cuellar has listened to over 20,000 recordings from the Frontera Collection)

Notation Field

A notes field is used to log additional information. Any anomaly of the record is noted, such as the labels on the wrong side, or incorrectly labeled all together, whether it is chipped or cracked, or excessively worn or too damaged to be recorded. If the recording is of a corrido, the Mexican narrative ballad form, further details are noted about the song, particularly if the song is about a true event. Due to limited time and the expectations of our funder, we cannot be as detailed as we would like in our notes field. Ideally, we could have more details about each song.

Results

We end up with a label scan, digitized sound file, and a database with discographical information and key words. How do we archive these at the Arhoolie Foundation and how do we send them down to UCLA to add to their digital library?

Data Archiving and Data Transfer Method

At the suggestion of our sound consultants, Advanced Systems Group, we chose to burn the files to DVD-R. For transferring them to UCLA, DVD-Rs are a very stable medium. Unlike a portable hard drive, you can drop DVD-Rs and there is no damage. They are not affected by magnetism and they hold 4.7 gigabytes. They are a standard storage medium that any modern computer DVD drive can read on any platform. UCLA then copies the information from the DVD-R and onto its Digital Library mainframe computer.

Storage at The Arhoolie Foundation

For the Arhoolie Foundation Archive, we wanted a medium that entail the least trouble and expense to store the 30,000 individual recordings that comprised the 78 collection. Hard drives were too expensive at the time and required too much maintenance. CD-Rs did not store enough per disc. We decided on DVD-R, which holds seven times more information than a CD-R.

The question was how long would a DVD-R last? If you listen to the manufacturers, they will tell you up to 100 years. I can tell you that the ones we burned three years ago still work. Selecting the medium to use for storage was one of the most difficult to determine. What is commonly assumed is that any files we create will have to migrate to the newest and latest storage device periodically to keep the storage medium from becoming obsolete. For now we burn a DVD-R and put the copy on the shelf. At some time down the line, we will probably have to copy them all to a new medium.

We have the luxury of knowing that UCLA will be putting the files on their Digital Library Computer System, backed up regularly by the UCLA IT staff. Even if there were a disaster at the Arhoolie Foundation, there would be the copy at UCLA and vice versa. This is one of the key advantages of digitizing an archive. Multiple copies can be easily made and stored in multiple locations guaranteeing that at least a digital form of the archive will always be preserved.

User Access

The UCLA Library has developed a searchable, interactive web site http://digital.library.ucla.edu/frontera/ where a user, logging in from the UCLA internet domain, can have complete access to the discography, sound recordings, and label scans. This allows an end user to search on a variety of criteria, read the discographical information, listen to the recording, and view the label. Due to copyright laws, a user logging in from outside UCLA can listen to only the first sixty seconds of the recording but can still view the label and discographical information.

UCLA is continuing to explore ways to make the archive more accessible. Researchers, students, and ordinary fans of the music, both at UCLA and around the world, are already using this web site. By the time we are finished digitizing the 78s in the Spring of 2005, UCLA will have almost 30,000 individual recordings accessible from this web site.

Home page for the UCLA Frontera Web Site: http://digital.library.ucla.edu/frontera/

In-house Database at The Arhoolie Foundation

The Arhoolie Foundation has created an accessible database for in-house use. We have converted all of the sound files into MP3s so that the file size is small enough to fit on a large single hard drive. We have also made medium resolution JPGs of the label scans. We have designed a searchable database. The database takes full use of the information in each data record, the label scans, and the sound files. When you find a record in the database, there are buttons to click on to view the label and hear the recording. You can search on any of the fields: artist, keyword, composer, musical genre, label, etc., and sort by these as well.

This makes studying a collection—once only accessible by going to the shelf and listening to one 78 at a time—far more easy. If you are looking for corridos about migrant labor, performed by Los Madrugadores on the Vocalion label, you can find it, listen to i,t and view the label as fast as you can type the query. Or you can quickly browse and listen to any recordings found by using any criteria.

An example on one record from The Arhoolie Foundation’s in-house database.

One of the most interesting benefits of digitizing our collection is that we are actually listening to every 78 in the archive. This collection, like many other archives, was often added to by obtaining large quantities of 78s in bulk. Until now, no one has ever listened to each individual recording. Through the process of digitizing and noting keywords, we are carefully listening to each and every recording in the collection. The result of this is now we really know exactly what is in the archive and we are discovering many hidden gems. By having, for the most part, one person, Antonio Cuellar, listening to all of these recordings, we are also creating quite an expert in the field on Mexican and Mexican American Recordings.

The Future: A Digital Enyclopedia

As we digitized the Frontera Collection, we realized that this was just the beginning of a digital encyclopedia on Mexican and Mexican American recordings. We have in our collections hundreds of photos of musicians that are being scanned and added to the archive. We have artists' biographies, record label histories, catalogs, recording logs, lyrics, articles about the music, posters, and videos. We are working on digitizing all this information and linking it together so that as you listen to a recording, you can view photos of the artist, read a biography of the artist, a history of the record label, view a poster of a concert, look up the recording log and see how much the musicians were paid for the recording session, read the lyrics, and even view a video or listen to an interview of the recording artist, getting an entire view of the music and where it came from. This way of examining an archive gives the end user as complete an educational and emotional experience as a computerized experience can be. It may not be the same as holding the actual artifact in your hand and listening to it on a turntable, but in some ways it can be much richer.

Eli Oberstien Recording Lydia Mendoza in 1936

Appendix

Equipment used for the Arhoolie Foundation’s Strachwitz Collection of Mexican and Mexican American Records Digital Archiving Project

Software:

© 2004 Tom Diamant