Sing softly and don't shout to reduce the risk of Covid-19 spread, new research suggests, offering a ray of hope for musicians who have been restricted from performing in public.

新研究指出，轻声歌唱、不喊叫可以降低新冠病毒传播风险，这给那些被限制在公共场合表演的音乐家带来了一丝希望。

"It is not about the vocalisation – whether it's singing or speaking – it is about the volume," said Jonathan Reid, a professor of physical chemistry at the University of Bristol and a co-author of the research. "Just by singing a little bit more softly you really reduce the risk."

In a study that has yet to be peer-reviewed, the team report how they asked 25 professional singers to breathe, speak, cough and sing into funnels. They then measured the mass of tiny droplets suspended in the air, known as aerosols, that were produced. The experiments were set up in an orthopaedic operating theatre, a setting chosen for its lack of background aerosols.

While one route by which Covid-19 spreads is via big droplets, largely produced when someone coughs and which fall to the ground within a couple of metres, Reid said aerosols were another possible route, noting such tiny droplets can linger in the air.

The team found the results of their study varied across participants, however at the lowest volume singing and speaking generated a similar mass of aerosols as breathing.

But when the team asked participants to recite Happy Birthday at different volumes, they found the loudest singing and speaking – 90-100dB – produced about 36 and 24 times the mass of aerosols respectively as generated by breathing.

"The volume of the activity, whether it is speaking or singing softly or speaking or singing loudly, that is really the main factor in governing the aerosol mass that is generated," said Reid, adding that while singing generates a slightly higher mass of aerosols than speaking, at least when loud, the difference is very small compared with the effect of volume.

In other words singing in a cathedral might pose a lower risk for spreading Covid-19 than shouting across a crowded pub. "Intuitively that would seem to be the case, assuming people are speaking or singing at the same sorts of volume," said Costello.

However the research has limitations: it only measured aerosols produced by one individual at a time, while it did not look at how much virus was contained in the aerosol or the actual infection risk posed by the levels of aerosol produced.

The team say their findings have already contributed to new guidance for England released on 15 August by the Department for Digital, Culture, Media & Sport.