Edison disc Master Mold playback tests

Sealed Master Mold (bottom view).

• Use the links below to hear results of Edison disc Master Mold and Second Master Record audio playback tests.

• Read a summary analysis of the tests.

Thomas Edison National Historical Park is conducting research to determine best practices to preserve the 9,323 sealed “Master Mold” disc sound recordings held in its collection.

 

Sealed Master Mold (top view).

Containing historic music recorded in New York City and European cities during 1910–1929, Edison used Master Molds in his Phonograph Works factory to mass-produce “Diamond Disc” and “Needle-Type” disc records. Sound on these discs is preserved within audio grooves made of copper and nickel, encased inside a solid copper seal. Each sealed Master Mold represents the closest surviving generation to the original studio master recording, containing high-quality sound.

A sealed Master Mold object, once unsealed, separates into a duplicate pair of two discs:

 

Unsealed Master Mold.

1. An unsealed Master Mold has a negative groove profile made of copper. (“Negative groove profile” means the groove shape is inverted -- it protrudes up from the disc surface, rather than into the surface.) A Master Mold is the one and only recording duplicated directly from the metallic soap (or "wax") Master Record, created via electrotyping copper over the wax Master Record to create a mold. As the wax Master Record is destroyed in the duplication process, this Master Mold is the “preservation master” sound recording. If preserved in excellent condition, the Master Mold retains the highest-quality sound available of an Edison disc recording.
2. A “Second Master Record” has a positive (into the surface) groove profile made of nickel. It was created by electrotyping the surface of the Master Mold.

By design, access to the two groove surfaces requires that the copper seal be cut open. Once the seal is opened, deterioration due to exposure of bare metal groove surfaces to oxygen in the open air becomes a concern. This concern prompted a conservation study in 2011 on how to best preserve the discs once they have been unsealed. Led by a metals conservator, the study involved opening two Diamond Disc-type sealed Master Mold objects as a test.

 

Second Master Record.

As part of this study, we digitized the discs using the traditional stylus-playback method. Stylus playback of the copper negative-grooved Diamond Disc Master Mold proved to be problematic for two reasons:

1. No stylus has ever been designed to traverse the negative groove profile of the vertical-cut (“hill-and-dale”) groove on an Edison Diamond Disc. (“Bifurcated” styli, typically used on negative groove discs, are designed for lateral-cut grooves, not vertical.)
2. Playback of the Master Mold using a bifurcated stylus caused a degree of visible and audible groove wear.

In December 2020, the National Recording Preservation Foundation awarded a grant to The Friends of Thomas Edison National Historical Park to hire the Northeast Document Conservation Center to “IRENE-scan” two pairs of discs that the park opened in 2011. IRENE uses a non-contact approach, which eliminates the possibility of damage caused by mechanical contact of a stylus on fragile media. The resulting image files are then processed through software that translates them into an audio file.

We present the results of these playback tests here. A summary analysis of the results is below the audio file links.

 

---

Edison Diamond Disc matrix 7529-C -- links to audio files

Title: Old pal (Why don't you answer me?), performed by Lewis James, recorded in New York, New York by Thomas A. Edison, Inc. on September 13, 1920.

Master Mold 7529-C (copper surface, negative groove):

Stylus playback test 1:
Date digitized: May 22, 2011
Stylus used: 3.0 bifurcated Stanton
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_ edis-a_20110522_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_copper-master_20110522_declicked-eqd.wav
Attibutes: 48 kHz sample rate ; 16 bit depth ; mono

Stylus playback test 2:
Date digitized: July 1, 2021
Stylus used: 3.0 bifurcated Stanton
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_ copper-master_20210701_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_copper-master_20210701_declikd-eqd.wav
Attibutes: 48 kHz sample rate ; 16 bit depth ; mono

Note on tests 1 & 2 above: For Master Mold 7529-C, using the same stylus-playback method, we digitized the same disc twice -- once during 2011 and again in 2021. The second playback of the disc tested the effect of possible change over time due to surface deterioration. The 2021 "audio not restored" version has more clicks than the 2011 version, perhaps due to stylus wear. The two "restored audio" versions sound very similar to each other.

IRENE scan:
Date digitized: May 5, 2021
Digitized by Northeast Document Conservation Center

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: 7529-C_Copper_MAS.wav
Attributes: 96 kHz sample rate ; 32 bit float depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: 7529-C_Copper_INT.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

 

Second Master Record 7529-C (nickel surface, positive groove)

Stylus playback test 1:
Stylus used: 2.0 FCR Expert Stanton
Date digitized: January 31, 2020
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_nickel-submaster_20200131-1_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_nickel-submaster_20200131-1_declicked-eqd.wav.
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

Stylus playback test 2:
Stylus used: 0.7 spherical Stanton LP
Date digitized: January 31, 2020
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_nickel-submaster_20200131-2_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C_nickel-submaster_20200131-2_declicked-eqd.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

Note on tests 1 & 2 above: For stylus playback of Second Master Record 7529-C, we tested two different sized styli. The two styli produced similar-sounding results.

IRENE scan:
Date digitized: May 5, 2021
Digitized by Northeast Document Conservation Center

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: 7529-C_Nickel_MAS.wav
Attributes: 96 kHz sample rate ; 32 bit float depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: 7529-C_Nickel_INT.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

 

Pressed copy of Diamond Disc 7529-C-1-6

Stylus playback:
Stylus used: 2.0 FCR Expert Stanton
Date digitized: March 2, 2020
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C-1-6_edis-67700_20200302_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_7529-C-1-6_edis-67700_20200302_declicked-eqd.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

[Jump ahead to summary analysis below.]

 

---

Edison Diamond Disc matrix 4897-C -- links to audio files

Title: Mighty lak' a rose, performed by Anna Case ; recorded in New York, New York by Thomas A. Edison, Incorporated on July 17, 1916.

Master Mold 4897-C (copper surface, negative groove profile):

Stylus playback:
Stylus used: 3.0 bifurcated Stanton
Date digitized: September 9, 2011
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_4897-C_copper-master_20110921_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_4897-C_copper-master_20110921_declicked-eqd.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

IRENE scan:
Date digitized: May 5, 2021
Digitized by Northeast Document Conservation Center

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: 4897-C_Copper_MAS.wav
Attributes: 96 kHz sample rate ; 32 bit float depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: 4897-C_Copper_INT.wav.
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

 

Second Master Record 4897-C (nickel surface, positive groove profile)

Stylus playback:
Stylus used: 2.0 FCR Expert Stanton
Date digitized: January 31, 2020
Digitized by Thomas Edison National Historical Park

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_4897-C_nickel-submaster_20200131_mono.wav
Attributes: 96 kHz sample rate ; 24 bit depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: tenhp_edison_d_4897-C_nickel-submaster_20200131_declicked-eqd.wav
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

IRENE scan:
Date digitized: May 5, 2021
Digitized by Northeast Document Conservation Center

Audio not restored (Left-click to open the audio file in a new tab.)
Filename: 4897-C_Nickel_MAS.wav
Attributes: 96 kHz sample rate ; 32 bit float depth ; mono

Restored audio (Left-click to open the audio file in a new tab.)
Filename: 4897-C_Nickel_INT.wav.
Attributes: 48 kHz sample rate ; 16 bit depth ; mono

 

---

Summary analysis of playback test results:

• The ideal playback option would produce excellent sound, while causing no trace of wear on the disc object. After comparing all of the recordings posted above, I believe most listeners will agree that stylus-playback of the copper negative Master Molds resulted in the best sound quality. Though, stylus-playback of the Master Molds caused a small degree of visible and audible wear on the copper groove surfaces.
• Stylus-playback of Master Mold 7529-C resulted in better sound-quality than stylus-playback of the excellent-condition pressed copy of Diamond Disc 7529-C-1-6. (These two discs contain the exact same recorded take.) This comparison illustrates the value of the Master Molds collection, as it indicates that a sealed Master Mold has the potential to provide the highest-quality sound available, compared with other copies of the same recording.
• Stylus-playback of the nickel positive Second Master Record caused no apparent wear, but the sound was marred by high/mid-range surface noise. If the nickel surfaces were in excellent condition, they should produce excellent sound, very similar to the Master Mold. (See "further testing" note below.)
• Initial stylus-playback tests of the Second Master Records produced many unwanted high-pitched "dink" sounds. We determined that these "dink"-sounds can be avoided via anti-static treatment, using an anti-static pistol (such as the "Milty Zerostat 3") on the groove surface prior to playback. All of the Second Master Record tests on this page include anti-static treatment.
• Audio quality from the IRENE-scans is lesser than stylus-playback of the Master Mold, though IRENE-scans have the advantage of producing no groove wear.

-Jerry Fabris, Museum Curator, Thomas Edison National Historical Park, August 2021

 

---

Further audio digitization testing still to do and options to consider:

Further playback tests are needed to determine best practices. We plan to explore the following alternate methods of playback:

• Conduct polishing tests on the Second Master Record. As was done historically, the Second Master Records may need to be polished prior to stylus-playback. (See for example Victor's "Command Performance" at 12:20 minutes into video.) Not polishing the Second Master Records prior to playback could be the cause of poor sound experienced in our tests, rather than surface deterioration.
• Design a bifurcated stylus specifically for vertically-cut Diamond Disc Master Mold grooves. Rather than a diamond, use a softer material (perhaps sapphire), to lessen the amount of wear caused by the stylus during playback. (Note: It may be problematic to find someone who has the required skills, equipment and willingness to make this custom-designed stylus.)
• Rather than play the Master Mold with a stylus at the prescribed speed of 80 rpm, playback at 40 or 20 rpm to exert less force. (Less force on the groove may produce less wear during playback.)
• Attempt to mold a new disc from the Master Mold using modern materials.
• Research the possibility of creating a vinyl pressing directly from the copper Master Mold. (Note: It is our assumption that copper is too soft of a metal to use for pressing without damaging the Master Mold. We assume that it is not avisable to use a Master Mold as a stamper for vinyl records in a commercial record press. Though, these assumptions need verification.)
• Research and test scanning options other than IRENE.

 

---

Acknowledgments

Thomas Edison National Historical Park gratefully acknowledges the following individuals and organizations that contributed time, expertise, and information to support these Edison disc Master Mold playback tests:

Rachael Arenstein, A.M. Art Conservation
Joannie Bottkol, Historic Architecture Conservation and Engineering Center, National Park Service
Frank Cunningham, Northeast Document Conservation Center
Michael Devecka, Antique-Sound.com
Larry Fast, The Friends of Thomas Edison National Historical Park
David Giovannoni, FirstSounds.org
Naomi Kroll Hassebroek, Historic Architecture Conservation and Engineering Center, National Park Service
John Levin, Barrington Media
Duncan Miller, Vulcan Cylinder Record Company
Northeast Document Conservation Center
Sarah Nunberg, The Objects Conservation Studio
Bryce Roe, Northeast Document Conservation Center
Gerald Seligman, National Recording Preservation Foundation
The Friends of Thomas Edison National Historical Park
Don Wilson, Wilson Records

The Friends of Thomas Edison National Historical Park gratefully acknowledge financial support for this project provided by the National Recording Preservation Foundation.