Mere Adequacy is Never Adequate

There are five fundamentals of high-quality digital sound archiving and preservation for analog source material: accuracy, appropriateness, adequacy, consistency, and explicitness.

Previously we examined accuracy (with regard to the source).Today we tackle the next two fundamentals, appropriateness and adequacy (with regard to your processes).

[Image Credit: Sarah Tehan, Aporia, 2013.]

The process of sound archiving and preservation is like scholarly editing. Just as the scholarly edition's basic task is to present a reliable text, the sound preservationist's basic task is to present a reliable digital record of an analog source. In both pursuits, decisions have to be made concerning (re-)construction and representation of the original material to achieve reliability.

Reliability in digital construction and representation of analog sound is a direct result of:

1. Appropriateness: the extent to which a particular process (step) advances accurate sound representation, and the extent which the listener is satisfied with the sound produced.
2. Adequacy: the extent to which a particular process (step) can be shown to be as good as necessary for its purpose or requirement.

The appropriateness and adequacy of the major function groups in the audio archiving and restoration pipeline can be characterized as follows:

• Digital Recording
  Appropriate: Critical first step in the archiving process
  Adequate: 48 KHz minimum sampling rate; 24-bit minimum encoding depth
• Audio Import/Export
  Appropriate: Advances sound data through the processing pipeline
  Adequate: No information lost in transfer
• Format Conversion
  Appropriate: Ultimate sound output in format desired by listener or required by tools
  Adequate: No information lost in conversion; minimal information loss in compression
• Automated Clean/Repair
  Appropriate: Algorithmic processing of common audio artifacts
  Adequate: No artifacts introduced 
• Manual Clean/Repair
  Appropriate: Processing of artifacts unique to a given recording
  Adequate: Minimal audio loss; clipping eliminated; no artifacts introduced
• Track Isolation
  Appropriate: Facilitates playback of individual songs/chapters/passages, etc.
  Adequate: Envelope of silence before and after track
• Volume Normalization
  Appropriate: Addresses volume differences on playlists incorporating multiple titles
  Adequate: Target sound level -0.1 dB for all titles
• Metadata Editing
  Appropriate: Captures relevant information about the original recording
  Adequate: Accurate, and without typographical errors

Another way of characterizing appropriateness and adequacy is that every step in your process must be necessary and your process as a whole must be sufficient. For a step to be necessary, eliminating that step would produce a less-reliable result. A process is sufficient if it is not a failure for having omitted other steps. All magic formulas for success break down on one or both of these criteria. Don't let it happen to you.

             Vinyl-to-Digital Restoration #65            

Artist: Miles Davis
Title: Porgy and Bess
Genre: Jazz
Year: 1958





To quote Dan Epstein: "Take George Gershwin's Porgy & Bess, add Miles Davis and arranger Gil Evans, and what do you get? A classic jazz album that--despite the fact that the material has been rendered almost overly familiar due to countless interpretations--still sounds remarkably fresh four decades after its initial release. Miles' soft yet piercing trumpet style is perfectly suited [TGD: appropriate] to Gershwin's melancholy melodies, Evans' musical direction of his 18-piece orchestra is impeccable [TGD: adequate], and their version of 'Summertime' may well be the finest ever waxed."

© 2014 Thomas G. Dennehy. All rights reserved.

Fast is Fine, But Accuracy is Everything

Previously we proposed there are five fundamentals of high-quality digital sound archiving and preservation for analog source material: accuracy, adequacy, appropriateness, consistency, and explicitness.

Today we tackle the first fundamental, accuracy (with regard to the source).

Consider this. Carole King's The Legendary Demos is now available in 24-bit/96kHz high definition. Demos. In high-def. Does that make any sense? Actually, it makes perfect sense, and illustrates the importance of accuracy.

Demos, as you probably know, are recordings of stripped-down arrangements songwriters use to "pitch" material — to producers, bandmates, potential clients. A promise of a song rather than the song itself. Neither professionally recorded nor mastered, the audio quality of a demo varies widely. And great songwriters are not always great performers (with notable exceptions).

The material on The Legendary Demos is true to the form. Many of the tracks sound as if they are simply Ms. King at the piano in her office. Recorded on 60s vintage analog equipment, the sound level is uneven, the piano not balanced with the vocal (the piano perhaps unmiked). So why go high-def?

The first reason is simply to master the material for release to give it as much polish as possible. (It sounds better here than on what may have been the original tapes aired during an April 2012 Fresh Air interview with Ms. King). Second and more importantly, these recordings document the creative process which produced songs that went on to become big hits. How does one preserve such moments for posterity so they are not lost or forgotten?

The International Association of Sound and Audiovisual Archives (IASA) "Guidelines on the Production and Preservation of Digital Audio Objects" offers these recommendations:

• Sampling rate. When producing digital copies of analogue material IASA recommends a minimum sampling rate of 48 kHz for any material. The unintended and undesirable artefacts in a recording are also part of the sound document [and] must be preserved with utmost accuracy. For certain signals and some types of noise, sampling rates in excess of 48 kHz may be advantageous. IASA recommends 96 kHz as a higher sampling rate, though this is intended only as a guide, not an upper limit.
• Bit depth. IASA recommends an encoding rate of at least 24 bit to capture all analogue materials. For audio digital-original items, the bit depth of the storage technology should at least equal that of the original item. It is important that care is taken in recording to ensure that the transfer process takes advantage of the full dynamic range.

Record analog source material for digital archiving with as many bits as you can muster. Preserve numeric precision (and hence dynamic range) in the audio processing pipeline. Establish and follow a plan that preserves the accuracy of the material presented.

[Originally published 10-May-2012. Revised and expanded.]

             Vinyl-to-Digital Restoration #32            

Artist: Carole King
Title: Tapestry
Genre: Pop
Year: 1971



If you are of a certain age, this is the album your sister and every woman you ever dated (my wife included) had in her collection . At one time, the best selling album in history, and Carole King the only female performer to win four GRAMMY® awards in one year. Legendary.

© 2012 Thomas G. Dennehy. All rights reserved.