From the scrapping of the live orchestra and language rule, to the recent use of pre-recorded backing vocals, the Eurovision Song Contest is constantly evolving and changing. But will autotune ever have a place in the competition?
It truly wouldn’t be national final season without a debate about vocal effects and their place at Eurovision. Last year, autotune was thrust into the spotlight when Spanish hopeful, Luna Ki, bowed out of Benidorm Fest due to being unable to use the effect if she represented Spain in Turin.
This year, autotune is once again a focus of the national finals, as Stig Karlsen – project manager of Norwegian national final MGP – has confirmed that all acts will be able to use autotune in their performances.
In and of itself, this may not be so controversial: Sanremo frequently allows their acts to use autotune as a vocal effect. After all, autotune is an instrument! You wouldn’t ask Alexander Rybak to perform without his violin so why should artists who use it as an integral part of their sound be any different? However, the controversy surrounding Karlsen’s decision seems to stem more from the rationale behind it. In an interview with Norwegian newspaper VG, Karlsen said: “A vocalist is also carried by the individual character of the voice. Autotune is just one of several tools for correcting and optimising the sound experience.” Many fans have interpreted this to suggest that MGP artists will not only be permitted to use autotune as a vocal effect, but also to mask sour notes. Given that Eurovision still doesn’t permit the use of “lead vocals, lead dubs, or any other voice that has the effect of unduly replacing or assisting to lead vocals during the live performance on stage”, could this mean MGP send an entrant to Liverpool whose performance can’t hold up live?
Debate around vocal effects at Eurovision is nothing new but will autotune ever be accepted? To find out we’re going to look at the history and evolution of autotune, the use of other vocal effects in the competition and what some of the Phoenix team think about its future in ESC.
A Brief History of Autotune
Autotune was invented in September 1997 by Ph.D. research engineer Andy Hildebrand, who had already had great success in the oil industry. When a colleague’s wife joked that a device should be invented to help her sing in tune, Hildebrand was inspired: what if he could invent a device that would allow off-key singers to sound as if they were perfectly in tune? It would revolutionise the entire music industry!
Before 1997, correcting vocal errors was no easy task. In the early years of recording music, it could only be achieved by hauling the entire band back into the studio to re-record the track. The invention of multitrack recording in the mid-50s made things a little easier, since musicians could now record their parts separately and put them together afterwards. If the singer was off-pitch they could simply go back to the studio alone and overdub the bum notes. However, while this cut down on studio time, it still cost serious money. And if the vocalist couldn’t sing live? Well, the only solution was for them to lip sync!
Autotune seemed to be the solution to many a record producer’s problem, offering them the chance to correct pitch in real-time, add harmonies without the need to hire back-up singers and it could even be applied to live performances if needed.
However, one day in October 1998, an experiment with autotune would change the face of music forever…
The Evolution of Autotune
Music producers Mark Taylor and Brian Rawling had been tasked to work with the ‘Goddess of Pop’ Cher. Their mission was to make a dance record that wouldn’t alienate her long-time fans so they began working on a modern-day disco number called ‘Believe’. Whilst in the studio, Taylor and Rawling began tweaking the use of autotune, discovering that by setting it “on its most aggressive setting, so that it corrected the pitch at the exact moment it received the signal, the result was an unsettlingly robotic tone.” The producers liked what they heard so they kept it on the track.
It was an instant success, with ‘Believe’ topping the charts in 21 countries worldwide. The effect was unlike anything listeners had heard before and everyone was desperate to know where that strange, synthetic sound had come from. Taylor and Rawling initially tried to keep it secret; they claimed the effect had been achieved using a vocoder, a device which synthesises the human voice and makes it sound robotic. However, soon their secret was out and all genres were experimenting with autotune. The phenomenon of distorting vocals quickly became known as ‘the Cher effect’. As Pitchfork contributor Simon Reynolds points out, “it always felt like a gimmick, something forever on the brink of falling from public favour. But autotune proved to be the fad that just wouldn’t fade. Its use is now more entrenched than ever.”
However, while undoubtedly popular, this doesn’t mean autotune is without controversy. Rapper Jay-Z titled the lead single of his 2009 album The Blueprint 3 as ‘D.O.A. (Death of Auto-Tune)’ with his lyrics attacking artists who used the software: “I know we facing a recession/But the music y’all making gon’ make it the Great Depression.” Others felt that autotune was cheating, arguing that it could make the most terrible singer sound good, with Time journalist Josh Tyrangiel dubbing it “Photoshop for the human voice.” Some argued that it took the soul and raw power from music, with producer Rick Rubin saying that “Right now, if you listen to pop, everything is in perfect pitch, perfect time and perfect tune. That’s how ubiquitous autotune is.”
Autotune as an Instrument
One artist credited with popularising the heavy use of autotune in music is T-Pain. Not only did T-Pain use autotune on his own songs but he inspired a host of famous rap and hip-hop artists to do the same. In 2008, T-Pain was asked to consult Kanye West on how to use the technology on his upcoming album 808s and Heartbreak. Dealing with the recent loss of his beloved mother and the end of his relationship with then-fiancée Alexis Phifer, West wanted a distorted, electronic sound to reflect his state of mind. In this way, autotune was as much of an instrument on 808s and Heartbreak as the titular Roland TR-808 drum machine.
T-Pain had used autotune from the beginning of his rap career to make his voice stand out, even when it drew criticism from his peers. Jay-Z’s aforementioned ‘D.O.A’ is rumoured to be directed at T-Pain and Usher once told him he had “f***ed up music for real singers”, spiralling him into a deep depression. However, it is unfair to say T-Pain is not a ‘real singer’, as anyone who has watched his NPR Tiny Desk Music Concert can attest to. In this footage, T-Pain performs with only a piano accompaniment, leaving his gorgeous, soulful singing voice to do the talking and proving his haters wrong. Therefore, like West, it is fair to say that T-Pain uses autotune as an instrument, not a cover-up for bad singing.
Vocal Effects in Eurovision
Nowadays most songs will feature some form of pitch correction, including the studio versions of Eurovision songs we know and love. However, no competing acts have used autotune live at the contest, although other vocal effects have been used. In 2000, Danish act Olsen Brothers won the contest with their song ‘Fly on the Wings of Love’. In the song’s final chorus a vocoder is used, giving the vocal a robotic effect. Russia, the runners-up, protested the win but Denmark were found not to be in breach of any rules and their win was upheld.
In 2017, Norway’s JOWST used vocal samples in his song ‘Grab the Moment’, despite live backing vocals still being required in ESC at the time. In a lengthy statement on his Facebook page, JOWST argued that these samples were a “digital instrument” and did not count as vocals. He was permitted to keep them in and placed 10th in the final.
On the other hand, in 2018 Israel’s Netta was forbidden from using a looper onstage during her performance of ‘Toy’ as the EBU deemed this a live instrument, which are not allowed to be played onstage during ESC. However, this did not deter Netta and she instead used live backing vocalists in the looper’s place, going on to win the competition.
Most notably, the use of pre-recorded backing vocals was permitted in 2021 due to ongoing concerns as a result of the COVID-19 pandemic. These were used prominently by Norway’s 2022 act Subwoolfer, who placed 10th and were one of the more memorable acts of the night. Could this have influenced Karlsen’s decision to allow autotune at MGP?
Autotune in Eurovision: Yay or Nay?
We’ve seen that the use of autotune sparks debate among musicians, producers and fans alike. What do Eurofans think about its potential future in ESC? In a Reddit thread sparked by Luna Ki’s withdrawal last year, they did not seem keen on the idea. LN_TheDudess said, “I think the problem is, although some people may use the autotune for stylistic choices, there’s going to be others who abuse the autotune to help them out. People already hide behind live and pre-recorded backing vocals so they probably would hide behind autotune as well if they can’t sing.” Others agreed, with user bo-tvt saying, “The title is ‘song contest’ but it’s also a singing contest and to do that you have to actually sing the song. No autotune please.”
Some of our very own Phoenix writers were more positive about the potential for autotune in future contests. Sophie argued that “Autotune is a part of modern music, and we now see more and more artists in the charts use it as a stylistic approach rather than to hide any vocal mishaps! Eurovision has done a brilliant job of modernising in the last few years or so but this is a huge stumbling block for the EBU. To really turn into a modern contest they need to find a way to include these types of vocals.” AJ agreed, saying that “Plenty of songs use autotune as an effect – it doesn’t mean that the artist in question is covering up their lack of ability. I think we could see some potentially interesting songs in the contest if they allow the use of autotune.”
Liv was also happy to see autotune have a place in ESC as long as it is “a stylistic choice and not one to cover up bad vocal ability.” She feels “there needs to be some regulation around it so that it doesn’t start being over-used or used in an attempt to push people who can’t sing live well to sound ‘better’. At the end of the day Eurovision is a live music contest; if you’re not great live then it’s just unfortunate and autotune or pre-recorded vocals shouldn’t be used to cover that fact.”
What do you think? Would you welcome the use of autotune in future ESC entries if it was a stylistic device or should it stay far away from Eurovision? How do you feel about MGP allowing its use? Have your say in the comments and keep the debate going by following Phoenix on our socials.
Pingback: 🇳🇴 Sneak Peek: Melodi Grand Prix Heat 1 – Phoenix
Pingback: 🇪🇸 Sneak Peek: Benidorm Fest 2023 Semi Final 2 – Phoenix