The oldest song in the world was written down in musical notation during the 14th century BC. The Hurrian Hymn to Nikkal was found inscribed on clay tablets in the ancient city of Ugarit (now Syria).
Nigel Christensen, President of the Australian Screen Sound Guild, told the story of the oldest song as part of his MetExpo presentation ‘Sound is Story’ on the first day of the three day conference, being held this week in Sydney.
This year the conference is free, register here.
Progressing through the history of sound, Christensen pointed out that song and oral storytelling are the oldest forms of story. In the Lascaux Caves, there are dots that signify the acoustic resonance of sounds painted onto the ancient cave paintings.
Twenty years before Thomas Edison invented his phonograph, the first recognised audio recording in history was made on a ‘phonautograph’ that etched sound waves onto a sheet of paper.
The oldest recorded song in Australia is the Hen Convention by Thomas Rome, a satirical song from 1896, sending up politicians, now available from the National Film & Sound Archive.
Before ‘Talkies’ arrived in the 1927 Al Jolson movie The Jazz Singer, moves were not silent, contrary to popular belief. Musicians played soundtracks and generated sound fx on a range of machines called ‘Photoplayers‘ before sound was invented for movie projectors.
“Every change in technology propelled a change in storytelling,” said Christensen, progressing to the implications of modern day technology on audio creativity. The first recordings of whale songs of the 1970s prompted the environmental activist movement. The invention of Fairlight and Protools technology influenced the art of music and dialog creaton and editing.
“Now that sound is digital it can be manipulated in ways that were never imagined. Everything we enjoy now was built on these foundations… Now we have arrived at a new point, generative Ai.”
He foresees that studios will soon have smart monitors that listen and learn from the way an audio engineer mixes music, dialog and soundtracks. “Machine learning and deep learning will be used to create sound in different ways. It could be a setting as simple as a knob that you can turn between ‘good’ or ‘bad’ mixing settings.”
“The Ai cat is out of the bag and it’s not going back,” said Christensen.
Ai computational power doubles at a rate of about every three months, with content able to be generated entirely from text prompts. “Ai is now moving from utility to creativity,” he said, mentioning Soundflow as an example of an Ai driven sync mixing tool.
“They will augment our capacity to move faster and, possibly, do it better,” said Christensen, citing the example of the Ai generated fake band Velvet Sundown, that has had millions of hits on Spotify.
He is optimistic about the way forward, with Ai extending humans not replacing them. But he warned the audio industry should think carefully about the ethical implications of using Ai in relation to consent, transparency and talent contracts. Talent contracts are now increasingly incorporating ‘no cloning’ and ‘residual rights payments’ in them.
“In our pursuit of efficiency, let’s not devalue human artistry… It is exciting, but also a little scary, the best revolutions usually are.
“Push forward and lean into machine intelligence. As technology changes again we will get new forms of storytelling and new ways to move an audience… Our best stories are yet to be told,” he said.
Earlier on the first day, SMPTE Chair Paul Whybrow thanked sponsors, members and exhibitors for their support of this year’s SMPTE MetExpo.
Conference Chair Doxia Brownlow explained the approach of the conference, which aims to anticipate what people will want to know about technology in the coming year.
This is the first year that all conference sessions are free. “So much work goes into these presentations that we want to give them the audience they deserve,” Brownlow said.
Focus areas for this year’s conference include: AI, security and remote production. There is also a training workshop for the SMPTE ST 2110 suite of standards for transmitting professional media over IP networks. The standard is designed for applications where high quality, flexibility, and reliability are important for broadcast and live production. It allows broadcasters to move away from traditional SDI (Serial Digital Interface) cabling to ethernet networks.
A feed of the conference sessions is available here.
Reporting: Steve Ahern