Three calibration campaigns of the spectrometer SPI have been performed before launch in order to determine the instrument characteristics, such as the effective detection area, the spectral resolution and the angular resolution. Absolute determination of the effective area has been obtained from simulations and measurements. At 1 MeV, the effective area is 65 cm^2 for a point source on the optical axis, the spectral resolution ~2.3 keV. The angular resolution is better than 2.5 deg and the source separation capability about 1 deg. Some temperature dependant parameters will require permanent in-flight calibration.

Supernova 1998bu in the galaxy M96 was observed by COMPTEL for a total of 88 days starting 17 days after the explosion. We searched for a signal in the 847 keV and 1238 keV lines of radioactive 56Co from this type Ia supernova. Using several different analysis methods, we did not detect SN1998bu. Our measurements should have been sensitive enough to detect 60Co gamma-rays as predicted from supernova models. Our 2-sigma flux limit is 2.3 10^{-5} photons cm^{-2} s^{-1}; this would correspond to 0.35 solar mass of ejected 56Ni, if SN1998bu were at a distance of 11.3 Mpc and transparent to MeV gamma rays for the period of our measurements. We discuss our measurements in the context of common supernova models, and conclude disfavoring a supernova event with large mixing and major parts of the freshly-generated radioactivity in outer layers.

The selection of the GLAST burst monitor (GBM) by NASA will allow the investigation of the relation between the keV and the MeV-GeV emission from gamma-ray bursts. The GBM consists of 12 NaI and 2 BGO crystals allowing a continuous measurement of the energy spectra of gamma-ray bursts from ~5 keV to \~30 MeV. One feature of the GBM is its high time resolution for time-resolved gamma-ray spectroscopy. Moreover the arrangement of the NaI crystals allows a rapid on-board location (<15 degrees) of a gamma-ray burst within a FoV of ~8.6 sr. This position will be communicated to the main instrument of GLAST making follow-up observations at high energies possible.

The type Ia supernova SN1998bu in M96 was observed by COMPTEL for a total of 88 days starting 17 days after the first detection of the supernova. The accumulated effective observation time was 4.14 Msec. The COMPTEL observations were performed in a special instrument mode improving the low-energy sensitivity. We generated images in the 847 keV and 1238 keV lines of 56Co, using improved point spread functions for the low-energy mode. We do not detect SN1998bu. A spectral analysis of our data also confirms the non-detection of the supernova. We discuss the event for which our upper limits constrain the standard supernova models.

The shield of the INTEGRAL spectrometer provides a large detection area with a wide field-of-view. Calculations have been performed to check whether the temporal analysis of the counting rate of the SPI anticoincidence allows the detection of explosions of novae. The background rate of the shield as well as its response to gamma-ray have been modelled with monte-carlo simulations. Accounting for uncertainties in the rate of novae, their distribution in the Galaxy and their light curves in hard X-ray domain, the number of nova explosions detectable with this method during the INTEGRAL mission, is estimated. Such observationnal mode will allow to improve our knowledge on nuclear runaway in novae. Since the maximum of magnitude in the visible happens later than in gamma-ray, SPI will provide alert for optical observations.