>The SABRE Engine is a hybrid air breathing rocket engine. In the past, attempts to design single stage to orbit rockets have been unsuccessful largely due to the weight of oxidiser, such as liquid oxygen.<p>Carrying your oxidizer with you allows you to avoid losses related to the accelerating (which is slowing down in the plane's frame or reference) of the unnecessary 80% of the incoming air, ie. nitrogen. Decompressing that nitrogen back in the working chamber (which also decreases the efficiency of the chamber) allows to recuperate back some of that loss. At slow speeds (like typical jet) - most of that loss is recuperated back. The higher the speed the bigger the share of that unrecuperated loss, due to higher temperatures, etc... Granted, thrust-wise having larger working mass moving slowly would usually/theoretically be welcome (larger prop effect, one example in application to rockets - air-augmented rockets like <a href="http://www.astronautix.com/g/gnom.html" rel="nofollow">http://www.astronautix.com/g/gnom.html</a>), yet at higher speeds other things like that mentioned loss, efficiency, complexity/cost, etc. start to become major factors.<p>Additionally the potential amount of oxidizer saved while reaching the orbit is really not that much. Quick mind napkin calculation shows that on the order of several tons per engine which is close to a wash given the additional hardware (thus weight) necessary for such scheme to work.<p>If anything, 2 stage system with returnable 1st stage should work much better/cheaper/robust as we can already see from SpaceX and SpaceShipOne. Developing such first stage - basically a plane for quick acceleration to M5-M8 and returning back to the airport should be pretty simple relatively to any other orbital/suborbital or long-range supersonic effort.