Fig. 8From: Synthesis of soundfields through irregular loudspeaker arrays based on convolutional neural networksNormalized reproduction error (NRE) distribution in \(\text {dB}\) for a source placed in \(\textbf{r}=[0.99\ \text {m}, 2.88\ \text {m}, 0\ \text {m}]^T\) at \(f=1007\ \text {Hz}\) when using: \(\text {MR}\)Â (a), \(\text {CNN}\)Â (b), \(\text {PM}\)Â (c), \(\text {AWFS}\)Â (d), and \(\text {AMR}\)Â (e). Black loudspeakers represent the geometry of the chosen arrayBack to article page