Planet detection through microlensing is usually limited by a well-known degeneracy in the Einstein timescale \(t_E\), which prevents mass and distance of the lens to be univocally determined. It has been shown that a satellite in geosynchronous orbit could provide masses and distances for most standard planetary events (\(t_E \approx 20\) days) via a microlens parallax measurement. This paper extends the analysis to shorter Einstein timescales, \(t_E \approx 1\) day, when dealing with the case of Jupiter-mass lenses. We then study the capabilities of a low Earth orbit satellite on even shorter timescales, \(t_E \approx 0.1\) days. A Fisher matrix analysis is employed to predict how the 1-\(\sigma\) error on parallax depends on \(t_E\) and the peak magnification of the microlensing event. It is shown that a geosynchronous satellite could detect parallaxes for Jupiter-mass free floaters and discover planetary systems around very low-mass brown dwarfs. Moreover, a low Earth orbit satellite could lead to the discovery of Earth-mass free-floating planets. Limitations to these results can be the strong requirements on the photometry, the effects of blending, and in the case of the low orbit, the Earth's umbra.