Quick Links

Who We Are

Read Our Research

Watch Our Videos

Hear Our Sounds

Use Our Work

How We
Preserve & Serve

How We Convert Images to Sounds

Scott de Martinville

The Patrimony of All Mankind

In The News

Help Wanted

Contact Us

RSS Feed






















Notes regarding the initial playback of Smithsonian NMAH 287654.11


The playback documentation released by Lawrence Berkeley Labs does not note the constant linear speed at which this recording was played.  A 20% reduction in speed sounds more natural to my ear.  There are two recording events between which the lathe was stopped and re-started.  Power to the lathe was removed at the end of each event while recording; similarly, recording of the second event began before the lathe was up to speed.


I laid a 1 kHz tone next to LBL’s eduction to document temporal adjustments.  The 20% reduction in playback speed lowered the reference tone to 800 Hz.  Then I roughly corrected for the start and stop of the lathe during the second recording event. These changes can be seen in the graph below and heard in this reference file.

To aid comprehension a bit I applied steep bandpass filters at 60 Hz and 4,200 Hz.


Flash player required

The Flash plug-in is required to listen to the audio. Download Flash now.

Get Adobe Flash player


[...] by inventors Sumner Tainter and H. G. Rogers.  It’s the eleventh day of March, eighteen hundred and eighty five.  [Trilled R] How is this for high?  Mary had a little lamb, and its fleece was [...], and […] — oh, fuck.

[Trilled R] {during power-up}
Mary had a littler lamb, and its fleece was [...], and wherever Mary went, the little lamb was sure to go.  How is this for height? 
[...] {during power-down}

In the two places where “white as snow” is usually found I can hear “black as soot,” “black as sin,” “black as coal” – and sometimes even “white as snow”.  I was initially concerned that my (and others’) hearing of the expletive was an anachronistic imposition of modern familiarity; however, repeated playing and close analysis imbue high confidence in this hearing.


I leave open the option that the speaker replaced “white as snow” with other words as a test of comprehensibility.  Lab notes report that associates were in the habit of comparing word-for-word what was heard on the receiving end of a telephonic or photophonic transmission against what was actually said.  In this way they could assess the accuracy of the transmission.  With this in mind, the speaker may have devised alternate recitations of the nursery rhyme as a fairer test of the technologies used in the recording.

Something interrupted the first and second recordings that caused the experimenter to voice an expression of frustration, stop the lathe (creating a splotch between the tracks) and begin anew.  What caused the interruption is not clear.  Perhaps the speaker bungled a script designed to test specific sounds: he may have intended to say “black as sin” but said “black as soot”, for instance.  If he did in fact use a different phrase the second time this hypothesis would be supported.  Or perhaps he noticed a technical problem mid recitation, stopped to make an adjustment, then continued the experiment.  The second track is slightly clearer than the first, which provides support for this hypothesis.

I hear “oh fuck” as an honest expression of frustration, neither planned nor intended to be heard outside of the Lab’s circle of associates.  We cannot say the speaker would have chosen his word more carefully had he known his utterances would be heard generations hence.  But we can fairly note that his experiment succeeded in an unplanned dimension.  More accurately than it registered his speech, his machinery captured the very real human emotion of frustration – perhaps for the first time in any sound recording.

– David Giovannoni