In July, Declan Costello – a.k.a The Voice Doctor – talked to Bachtrack about the PERFORM study, which looked at the quantity of potentially infective particles (“droplets” and “aerosol”) exhaled during singing and the playing of wind instruments, the purpose being to shed light on what would be appropriate conditions for a return to live performance. Rarely have the results of a scientific study been so eagerly awaited.

Declan Costello

The results for voice were published on 19th August in a “pre-print” form (they have not yet been peer reviewed). Costello describes the key findings: “Singing doesn't generate vastly more aerosol than speaking. In fact, the biggest factor in terms of the variability of aerosol production is the volume with which you're using your voice. There is a huge amount more aerosol generated at high volumes than at low volumes, a 20 to 30 fold increase.” The results show little variation across genre (singer were involved from opera, choral, soul, rock and other genres), or the sex or BMI (body mass index) of the singer. The single highest level in the study came from the strongest cough.

In that case, I ask, will greater use of amplification enable singers to sing more softly and thus cause lower risk? Costello doesn’t think so: “ in the context of 20 or 30 people on the stage and maybe a few hundred in the audience, I think the marginal difference of being able to sing 10dB quieter because you’ve got amplification is probably going to be a relatively small area for improvement.”

HGO's Sāvitri in the gardens of Lauderdale House, London
© Laurent Compagnon

The UK government got to see these and other results before the rest of us, so on 13th August, to huge sighs of relief from the performing arts industry, they amended the guidance for the Performing Arts to permit the restarting of live performances – albeit with restricted audiences for indoor events and a raft of guidance for promoters and venues as to how to ensure that performances are as safe as possible. A number of braver organisations have sifted through the recommendations and gone ahead with staging events of various types: I’ve now attended one opera and one multi-genre mini-festival, both outdoors and both with strict controls to enforce social distancing. In the next week, I expect to be joining my first amateur singing event – again, in a format that’s very different to its pre-Covid equivalent, with a carefully laid out room and a raft of rules about behaviour.

Essentially, what the PERFORM study has achieved is to debunk comprehensively the idea that singing is in some way uniquely dangerous compared to all other activities. The idea stems from a choir practice on 10th March in Skagit County, Washington in which somewhere between 30 and 53 of the 61 attendees become infected, after which it gained a great deal of currency in public health circles and in the broader media. The US Center for Disease Control gave a measured response (recommending better ventilation and 6 feet distancing), but others, including Public Health England, leapt to a more extreme position. PERFORM also debunks many of the more outlandish ideas about how classical singing must be far more dangerous than other genres (or far less dangerous, depending on who you asked).

Socially distanced audience for Precipice at The Grange Festival
© Joe Low

So is that “job done”? Not quite, says Costello. “This paper is yet to be peer reviewed and published. But in this strange world we're living in, rather than waiting for a definitive publication before you go public with things, scientists are tending to publish things before they’ve gone through that peer review process. And we need to publish the data with regards to wind and brass players; there's a huge amount of data that we've collected that needs to be analysed and put into a paper there. So yes, there's an awful lot more yet to do; it’s not necessarily job done.” There are “various logistical problems” in getting the wind and brass data to the same state as the voice data, but the PERFORM researchers are hoping to do so in the next few weeks.

And while this research has enabled the tentative restart of performances, it’s hard to argue against the view that it’s only a single piece of a very large jigsaw. One obvious area for research is the effect of different types of ventilation systems in different types of buildings: there are mathematical models to do this based on various fluid dynamics algorithms, Costello tells me, but they’re complex: a building like Westminster Abbey will behave very differently from Cadogan Hall or St John’s Smith Square. Turning those models into a broadly applicable guide which doesn’t need specialist ventilation consultants would be a godsend to venue operators. The University of Colorado at Boulder has been prominent, with Professor Mark Hernandez leading research into how the virus moves indoors and how it responds various filtration and disinfecting techniques.

The elephants in the room remain: how much infective material is contained in the droplets produced in singing (or speaking or, for that matter, breathing)? How much infective material constitutes a significant likelihood of infection? And what, for that matter, is the infection mechanism within humans? There are many studies likely to come out over the coming weeks, and it’s notable how cross-disciplinary the problem is: the paper investigating the Skagit Valley Chorale incident by CU Boulder’s Shelley Miller lists authors from various departments of Mechanical Engineering, Civil & Environmental Engineering, Chemistry and Microbiology as well as public health officials.

Looking at the PERFORM results, and looking at the way that the UK’s Covid-19 infection rates changed from a downward path to an upwards one in early July, I find it hard to escape the conclusion that the re-opening of pubs and bars in England on 4th July was a really terrible idea. Costello isn’t so sure: “Well, I have a certain amount of sympathy for the politicians here, because if you purely looked at the public health aspects of things, then you would probably keep the whole society locked down until we were well on top of this and we have vaccines and treatments and all the rest of it. But actually, that's not a politically viable option. You can't keep the economy locked down forever. I really don’t envy politicians in having to make these decisions: they are terribly, terribly difficult.”

For now, let’s celebrate the fact that even though the UK’s number of recorded Covid-19 cases is increasing, the UK’s rate of Covid-19 deaths has dropped to 10 per day – breast cancer kills three times that number. Let’s hope that the decline can be maintained and that improved understanding of safety measures can lead to a steady return of music performance.