\begin{table}%t3 \caption{\label{tab-summary1b}Spectral fit parameters for X-ray binaries.} \par \small%\centerline { \begin{tabular}{ll|ccccccc} \hline\hline Object & Instru- & \multicolumn{7}{c}{Spectral Model Fit Results$^a$}\\ & ment &$\Gamma_1$ & $E_{\rm cutoff}$ & $\Gamma_2$&Norm & Norm$_2$ & $\chi^2$/d.o.f. & F-test\\ & & & $[$keV$]$ & &$[$keV$^{-1}$cm$^{-2}$s$^{-1}]$ &$[10^{-6}$ keV$^{-1}$ cm$^{-2}$ s$^{-1}]$ & & prob.$^b$\\ \hline & SPI & 1.36 $\pm$ 0.09 & 206 $\pm$ 50 & -- & 0.031 $\pm$ 0.010 & -- &1.6 & -- \\ \raisebox{1.5ex}[-1.5ex]{XTE J1550-564$^c$}& ISGRI &1.00 $\pm$ 0.05 & 108 $\pm$ 9 & -- & 0.009 $\pm$ 0.002& -- & 1.3 & -- \\ \hline & SPI & 3.3 $\pm$ 0.2 & -- & -- & 33 $\pm$ 21 & -- & 0.6 & -- \\ \raisebox{1.5ex}[-1.5ex]{4U 1630-47} & ISGRI$^d$ & 2.3 $\pm$ 0.2 & -- & -- & 0.043 $\pm$ 0.038 & -- & 1.0 & -- \\ \hline & SPI & 1.23 $\pm$ 0.07 & 23.9 $\pm$ 3.2 & 1.27 $\pm$ 0.58 & 0.16 $\pm$ 0.04 & 2.1 $\pm$ 1.0 & 0.7 & 0.015\\ \raisebox{1.5ex}[-1.5ex]{OAO 1657-415$^c$}& ISGRI & 1.42 $\pm$ 0.31 & 24.7 $\pm$ 3.6 & -- & 0.23 $\pm$ 0.21 & -- & 0.4 & -- \\ \hline & SPI & 1.58 $\pm$ 0.11 & 100 & 0.97 $\pm$ 0.63 & 0.081 $\pm$ 0.026 & 3.8 $\pm$ 2.4 & 0.8 & 0.0035\\ \raisebox{1.5ex}[-1.5ex]{GX 339-4$^c$}& ISGRI & 1.71 $\pm$ 0.10 & 100 & 1.0 & 0.038 $\pm$ 0.011 & 1.9 $\pm$ 0.8 & 0.7 & -- \\ \hline & SPI & 2.00 $\pm$ 0.08 & 47.5 $\pm$ 3.9 & -- & 3.14 $\pm$ 0.71 & -- & 1.6 & -- \\ \raisebox{1.5ex}[-1.5ex]{4U 1700-377$^c$}& ISGRI & 2.31 $\pm$ 0.17 & 74.1 $\pm$ 14.1 & -- & 6.4 $\pm$ 3.2 & -- & 0.9 & -- \\ \hline & SPI & 1.75 $\pm$ 0.20 & -- & -- & 0.016 $\pm$ 0.013 & -- & 0.4 & -- \\ \raisebox{1.5ex}[-1.5ex]{IGR J17091-3624$^c$}&ISGRI & 1.93 $\pm$ 0.04 & -- & -- & 0.034 $\pm$ 0.005 & -- & 1.2 & --\\ \hline & SPI & 3.26 $\pm$ 0.06 & -- & -- & 26.4 $\pm$ 5.9 & -- & 0.7 & --\\ \raisebox{1.5ex}[-1.5ex]{GX 354-0$^c$}& ISGRI & 3.53 $\pm$ 0.02 & -- & -- & 43.3 $\pm$ 3.0 & -- & 1.0 & --\\ \hline & SPI & 1.80 $\pm$ 0.06 & 113.4 & -- & 0.28 $\pm$ 0.06 & -- & 1.9 & -- \\ \raisebox{1.5ex}[-1.5ex]{1E 1740.7-2942$^c$}&ISGRI & 1.44 $\pm$ 0.04 & 113.4 $\pm$ 7.6 & -- & 0.040 $\pm$ 0.004 & -- & 0.7 & -- \\ \hline & SPI & 2.82 $\pm$ 0.14 & -- & -- & 2.1 $\pm$ 1.1 & -- & 1.1 & -- \\ \raisebox{1.5ex}[-1.5ex]{IGR J17464-3213}&ISGRI & 2.49 $\pm$ 0.02 & -- & -- & 0.63 $\pm$ 0.04 & -- & 0.9 & -- \\ \hline & SPI & 1.54 $\pm$ 0.07 & 178 $\pm$ 28 & -- & 0.12 $\pm$ 0.027 & -- & 1.7 & -- \\ \raisebox{1.5ex}[-1.5ex]{GRS 1758-258$^c$}&ISGRI & 1.36 $\pm$ 0.02 & 123 $\pm$ 5 & -- & 0.06 $\pm$ 0.004 & -- & 1.9 & -- \\ \hline & SPI & 1.51 $\pm$ 0.14 & 48.4 $\pm$ 7.8 & 0.82 $\pm$ 0.69 & 0.24 $\pm$ 0.09 & 2.1 $\pm$ 1.8 & 0.5 & 0.002 \\ \raisebox{1.5ex}[-1.5ex]{Ginga 1826-24$^c$}&ISGRI & 1.78 $\pm$ 0.04 & 68.5 $\pm$ 3.5 & -- & 0.42 $\pm$ 0.05 & -- & 1.5 & -- \\ \hline & SPI & 1.7 $\pm$ 1.7 & 12.5 $\pm$ 9.2 & 2.76 $\pm$ 0.18 & 4 $\pm$ 15 & 1.4 $\pm$ 0.1 & 0.6 & 9.2$\times$10$^{-5}$ \\ \raisebox{1.5ex}[-1.5ex]{GRS 1915+105$^c$}&ISGRI & 0.3 $\pm$ 1.1 & 7.6 $\pm$ 2.1 & 2.87 $\pm$ 0.05 & 0.09 $\pm$ 0.25 & 1.17 $\pm$ 0.04 & 0.7 & 7.8$\times$10$^{-9}$ \\ \hline & SPI & 1.28 $\pm$ 0.19 & 95.9 $\pm$ 15.5 & 2.11 $\pm$ 0.09 & 0.43 $\pm$ 0.35 & 99 $\pm$ 18 & 0.9 & 9.4$\times$10$^{-6}$ \\ \raisebox{1.5ex}[-1.5ex]{Cyg X-1$^c$}& ISGRI & 1.44 $\pm$ 0.14 & 115 $\pm$ 18 & 2.20 $\pm$ 0.14 & 0.90 $\pm$ 0.53 & 36 $\pm$ 23 & 1.5 & -- \\ \hline & SPI & 2.0 $\pm$ 0.7 & 20 $\pm$ 10 & 2.4 $\pm$ 0.3 & 3.5 $\pm$ 6.6 & 9.0 $\pm$ 1.1 &0.8 & 3.5$\times$10$^{-3}$\\ \raisebox{1.5ex}[-1.5ex]{Cyg X-3$^c$}& ISGRI & 2.00 $\pm$ 0.08 & 20 & 2.6 $\pm$ 0.2 & 3.4 $\pm$ 0.7 & 5.3 $\pm$ 0.5 & 1.7 & -- \\ \hline \end{tabular}} \medskip $^a$~Spectral parameters not used in the model for a particular object are marked as~``$-$''. \\ $^b$~F-test probability that the fit improvement achieved by the inclusion of a hard tail in addition to a cutoff powerlaw was by chance, i.e. that the hard tail is not real. Note that since ISGRI has lower sensitivity above 500~keV and none above 700~keV, we can in most cases not test for the presence of the additional component but only check the compatibility with the ISGRI~data. \\ $^c$~Parameters of alternative fit models are given in Sect.~\ref{secresults}. \\ $^d$~Spectrum very probably flawed by source contamination from IGR~J16358-472. \end{table}