\begin{table}%t1
\caption {\label{tab:processing}Overview of the processing parameters for the three data sets that are handled with SPAM as defined in the test cases.}
%%\centerline

 {\begin{tabular}{l c c c}
\hline\hline
Field name & VLSS J0900+398 (simulated) & VLSS J0900+398 (real) & VLSS J1300-208 (real)\\
\hline
Pixel size$^a$ & 18.9\arcsec & 18.9\arcsec & 11.1\arcsec \\
Number of facets & 347 & 243 & 576 \\
Facet separation & $1.18\degs$ & $1.18\degs$ & $0.62\degs$ \\
SPAM calibration cycles$^b$ & 1 & 1 & 3 \\
Peeling sources & 10$^c$ & 11 & 9 \\
KL model height & 1000 km$^d$ & 200 km & 200 km \\
Fitted KL~model terms & 15 & 15 & 20$^e$ \\
Rejected time intervals & 0 / 464 & 25 / 464 & 86 / 484 \\
Model fit phase rms & $3.0\degs \pm 0.8\degs$ & $21.3\degs \pm 2.4\degs$ & $23.2\degs \pm 3.2\degs$ \\
Peeling corrections applied directly & no & yes & yes\\
 \hline
 \end{tabular}
}
\smallskip
$^a$ {The pixel size for all field-based calibration images is~20\arcsec.}
$^b$ {Adding more cycles did not significantly improve the image quality.}
$^c$ {Arbitrarily limited to mimic a more realistic scenario.}
$^d$ {Increased to improve match with FBC phase screen.}
$^e$ {In this case, 15 terms proved to be insufficient.}
\end{table}