Intercomparisons

Figures - plot various of the quality measures against one another. The most obvious result is that the dynamic range is not very useful - images with the same DR may differ by more than a factor 10 in other, more accurate measures of their quality. As expected, the snapshot observations yield much poorer images, by any measure, than the corresponding long syntheses. All measures but the DR also show that a snapshot observation of any of the sources produces an image much worse than that produced by a long synthesis of any of the sources. For complicated sources this is all reasonable. Finally, the two means are correlated but with a large scatter (about a factor five), while the median peak FI and the peak SNR agree very well indeed.