Can You Generate a Completely AI Produced Song?

A-MNEMONIC’s Toby Jarvis explores whether artificial intelligence can truly create music that emotes, and has personality and believability.

Having watched the rise of artificial intelligence – both during the creative process and, with the listener or consumer, things are getting interesting.

Is it possible to generate an utterly AI-produced song? Not only the lyrics but the chord structure, the melody, all the band member’s instruments – and the lead singer?

Alan Turing experimented with computer-generated melodies in 1951, David Bowie with randomised lyric writing and many, many others since have used AI to augment their creative/ recording process.

There are a plethora of artificial intelligence, prose, lyric and songwriting tools out there. Some are unnervingly clever. There are chord progression, and melodic generators too. They suggest any number of chord sequences and musicologically solid routes forward.

I’ve been using a ‘virtual drummer’ for ages. It obediently plays along while you’re working out ideas, it’ll follow your tempo perfectly, and even decide on where to correctly play any fills.

There are hundreds of programmes like Riffer, which will endlessly generate melodies or arpeggios or rhythms. Given you dictate the key/tempo and sound, these many programs are still very much under the creative control of the composer.

The only AI programme we found to generate anything near 100% original was ‘Melobytes AI Song’. However, scoring high with creative originality it scored very low in how it sounded. Terrible!

For singing voice generation, we sometimes use Emvoiceapp. You input the melody, either by picking it out with one finger on a keyboard or inputting the notes on a grid. You then type in your lyric text, then choose one of four pre-set voices. Press enter and… Bingo! Well, sort of. One of the voice options sounded like a really pissed-off Cher. One of the male voices sounded quite creepy. No character, spark or believability. We sometimes use it to work out backing vocal harmonies, in advance of the singer doing it for real.

All these creative aids have become standard use in music creation. However, all require significant human input and judgement. None of these everyday tools can create anything like original music. Yet. It’s still down to the writer.

If a particular song has been successful, machine learning can be used to scan existing models and create new versions. Lately, AI-penned (or modelled) songs have come into their own. Indeed there is a top 10! In most cases, the lyrics are written using predictive text. Artist, “Botnik” used predictive text 50% trained in Morrissey lyrics, and 50% in Amazon customer reviews!

However, aside from the lyrics, these ‘tribute’ songs are based on an existing artist’s lyric style and sound. They’re all sung by humans with human people playing. Not truly AI-generated.

So, is it possible to generate a completely AI-created song? Just input your desired mood, and the emotions you want to hear and feel.

You say to Alexa: “Make me a track in the style of the Beatles, Swedish House Mafia, Beethoven with a 10% of dubstep, oh and it’s rainy and a robot took my job today. Please create a song that fits that criteria for me.”

Some production music libraries are heading this way – Soundraw, Jukebox and Amper are a few. However, they seem to rely on vast amounts of pre-recorded (and pre-composed) audio.

Because AI has to be pre-programmed with options to create specific musical pieces, it often produces music that is predictable. Desperately bland and characterless. Because it is essentially governed by an algorithm. It doesn’t feel believable.

An audience will listen or engage with music if it emotes, and has personality and believability. As far as I can see, pure AI music creation can’t do this yet. Not by a long way.

A-MNEMONIC create new audio branding for Getir with Neverland Agency

A-MNEMONIC have created a music branding package for Getir – the app-based home delivery service. The first time Getir has embarked on such a wide-reaching audio branding exercise.
Getir identified the need for a ‘characterful brand sound’ to be used across all their media assets worldwide. Together with agency Neverland, they commissioned A-MNEMONIC to create a sung mnemonic, ownable and unmistakably attributable to Getir, that anyone could sing, whistle, hum or mimic… and work in any language.

Getir, founded in Turkey in 2015, operates in more than 50 cities in its home market, and has expanded over the past year into eight other countries, including the UK and US.
A big player in the UK’s fast-growing rapid delivery market, Getir aims to bring groceries to customers in as little as 10 minutes.
Getir (the name means “bring” in Turkish) offers customers a selection of some 2,000 grocery items, ordered online and delivered to your door within minutes – at any time of day!

The A-MNEMONIC music team went through various stages of brainstorming and creative workshops, working with various composers, musicians and singers. The final result was a sung mnemonic using the strapline ‘You’ve Got It – Getir’.

Getir Mnemonic

A-MNEMONIC produce music for ‘I Literally Just Told You’ on Channel 4

Richard Bacon’s new Channel 4 creation ‘I Literally Just Told You’ has literally turned the quiz show format on its head.

Hosted by Jimmy Carr, four contestants have to answer questions that have been written LIVE as the show is being filmed, from a celebrity cameo to an off the cuff joke, or even something one of the contestants has shared about themselves. 

The show ducks and dives with tension and drama, and A-MNEMONIC helped create that extra tension with the music – writing 40 minutes of music that’s used throughout the show, as well as composing the theme music.

The final two contestants hilariously have to write the questions for each other for the last round, fighting to win a cash prize of up to £25,000 

Never Mind The Buzzcocks gets an A-MNEMONIC music update

A-MNEMONIC have updated the Music Theme, as part of the rebranding of Never Mind The Buzzcocks.

The renowned comedy show has returned to Sky with brand new host Greg Davies. Returning to the show, Noel Fielding is a team captain opposite Daisy May Cooper, with Jamali Maddix featuring as a regular guest.

Each week the host and team captains are joined by a roster of top music and comedy stars to compete against each other. Viewers can expect the unpredictability of the original series, with some of the best known and loved rounds remaining, alongside some new surprises thrown in along the way.

A-MNEMONIC produce the new Audio Branding for The Guardian Podcasts

Despite being the leading commercial podcast publisher in the UK, The Guardian has only recently turned their attention to the task of branding their podcast offering

Over 15 million people in the UK listened to podcasts in 2020, with the figure set to rise by a cool five million in the next four years. That’s a lot of listeners! 

For The Guardian, A-MNEMONIC worked with the idea outlined in the brief: reflecting The Guardian’s core values and the concept of ‘diverse voices in conversation’. 

Besides creating a sonic identity for the suite of podcasts, A-MNEMONIC also had to search and cast for five male and female voices that could veritably represent the sound of contemporary Britain.

Redknapps Big Night Out

Music theme produced for ‘Redknapp’s Big Night Out’

A-MNEMONIC were asked by Sky One and Expectation to produce a music theme ‘with a wall of attitude’ – for the new Father-and-son duo Harry and Jamie Redknapp topical sports-come-chat show ‘Redknapp’s Big Night Out’,  on Sky One.

Following the success of their ‘Redknapp’s Home Fixture’ last May, the father-and-son duo have reunited with comedian Tom Davis for the new six-part topical sports-come-chat show, which airs on Sky One and TV streaming service NOW.

A-MNEMONIC Dominates Saturday Night Prime Time TV producing Music for both ‘I Can See your Voice’ and ‘Game of Talents’

A-MNEMONIC has gone head to head and produced both the music themes and music branding package for ‘I Can See Your Voice’ on BBC One, and ‘Game of Talents’ on ITV.

I Can See Your Voice is hosted by Paddy McGuinness who is joined by celebrity investigators Jimmy Carr, Alison Hammond, and Amanda Holden.

Each week the celebrities, joined by a special guest, will help a team of two players tell the difference between good and bad mystery singers, without ever hearing them sing a note.

Then over at ITV, you can watch ‘Game of Talents’ – branded ‘the most judgemental TV show’ and presented by Vernon Kay in the new seven-part series which features eight different performers every week, each showcasing their own talent.

There’s one big catch though – contestants in the studio and viewers watching at home have no idea what those talents are and have to make a prediction based on the performer’s appearance.

A-MNEMONIC helps to being a little softness to Charlie Banana diapers commercial.

We’re excited to share with you our latest music and sound design on this beautiful commercial for Charlie Banana. With music composed by Stuart Hancock, and Sound Design by Toby Jarvis.
We teamed up with NERD Productions London and USA based Agency Curiosity, to bring to life Charlie Banana’s mission of making the world a better place for future generations and to celebrate motherhood and all-around sweetness. Check out the hand-crafted, stop motion piece of beauty! ‘A little softness’, Directed by Hayley Morris.

Martin-lewis-money-show-news

A-MNEMONIC commissioned to produce a new Title Theme and music package for the rebrand of the ‘The Martin Lewis Money Show’ on ITV.

Now in its 10th series, and presented by Martin Lewis and Angellica Bell, the show is the UK’s most-watched current affairs programme averaging just under four million viewers. 

In preparation for the new series, the Martin Lewis production team were tasked to update the entire look and feel, including commissioning new music identity. 

A-MNEMONIC were invited to compose a new theme for the show, but with an entertainment twist – fun but authoritative. This music package included opening titles, end titles, music beds and stings.