Magee et al (2020) :question:
measured speech intelligibility in an ideal listening environment, lacking background noise a
Magee and colleagues found that different types of face masks affected acoustic signals differently, with higher frequencies of acoustic signals being attenuated across three types of masks
Keerstock et al. (2020) evaluated how protective masks, background noise, and non-native accents affect speech intelligibility and overall communication. They found speech produced in conversational speaking styles with a paper mask was as intelligible as speech produced without a mask. More specifically, conversational speech produced while wearing a paper mask did not negatively affect speech intelligibility in quiet listening conditions for either native or non-native speakers and occasionally even in the presence of noise for the native speaker.
It has been well established that clear speech enhances speech intelligibility by increasing acoustic-articulatory outcomes (Cook et al., 2014; Smiljanic & Bradlow, 2014; Smiljanic & Gilbert, 2017ab;
Clear speech has been shown to improve speech intelligibility in the various populations: second language learners (Bradlow & Bent, 2002), learning-impaired children (Bradlow et al., 2003), individuals with hearing loss (Ferguson, 2012; Ferguson & Quene, 2014), children with cochlear implants (CIs) (Smiljanic & Sladen, 2013), and adult with CIs (Rodman et al., 2020)
Clear speech benefits on speech intelligibility have been well documented and evaluated by a variety of procedures prior to the COVID-19 pandemic (Bradlow & Bent, 2002; Bradlow et al., 2003; Ferguson, 2012; Ferguson & Quene, 2014; Rodman et al., 2020; Smiljanic & Gilbert, 2017; Smiljanic & Sladen, 2013; and Van Engen et al., 2014)
clear speech resulted in higher speech intelligibility than conversational speech in all types of mask conditions in both background noises and presentation modes
These findings are mostly in agreement with Keerstock et al. (2020). They found that the use of clear speech improved intelligibility produced with no mask and disposable paper masks in the presence of two maskers.