IN FEBRUARY 2006 a news report echoed around the internet, purporting to play back 6,500-year-old voices and other sounds from a clay pot. The pot allegedly had waveforms etched into a groove as a potter incised a line with a stylus while the pot spun. It turned out to be Belgian television's offering for April fool's the previous year. But Patrick Feaster's rendition of 1,000-year-old audio is no jape. In May he regaled the annual conference of the Association for Recorded Sound Collections with the latest in what he dubs paleospectrophony.
It was only with Thomas Edison's invention and commercialisation of the phonograph in the late 19th century that etching transient sounds reliably in solid matter for future aural reproduction became possible. Earlier efforts involved trying to capture noises on paper, by hand or using various contraptions. Back then sound scribes reasoned that if they transcribed audio in the right way, others would be able to replay it in their heads, says Dr Feaster, just as trained musicians look at a score and hear the music. It did not work out that way. Most human brains are not, it seems, quite plastic enough to make the leap from a visual representation to an aural one.
Dr Feaster has been working for years to recover sounds from early attempts to transcribe them in tangible form. He and his colleagues have discovered troves of pre-Edison sound writings, and developed new techniques that ease the translation of assorted squiggles into recognisable noises. Take phonautograms, churned out by a device called, logically, a phonautograph. Invented by �douard-L�on Scott de Martinville, it took sound carried through a horn and moved a stylus to draw a waveform in lampblack, a sort of powdered pigment, onto paper or glass. This is more or less how modern equipment captures sound, but the variation and fidelity of the phonautograph necessarily produced poor recordings. The results are nonetheless recognisable noises. (It was not until the invention of the spectrogram during the second world war that both pitch and intensity?ie, the sound waves' frequency and their amplitude?could be captured in a form that could be interpreted visually.)
In 2008 Dr Feaster joined forces with audio expert David Giovannoni and scientists at the Lawrence Berkeley National Laboratory to decipher one of Scott de Martinville's better preserved phonautograms. They used a process the lab had developed for teasing sounds out of the recorded grooves made in fragile wax records or broken disks from digital photos or scans of high-resolution pictures rather than by physical contact.
Source: http://www.economist.com/blogs/babbage/2011/07/ancient-audio?fsrc=rss
No comments:
Post a Comment