phase, the behavior of the jet, revealed by infrared and radio observations, also begins to change. The infrared emission drops almost as soon as the state transition begins (14 ), indicating a change in the jet properties(density andmagneticfield) close to the black hole. The radio emission begins to vary more dramatically, showing oscilla- tions and flare events superposed on an overall decline (8 , 15 ). At a cer- tain point, there are one or more large radio flares, which can be two or more orders of magnitude more luminous than the previous existing, steadier jet in the hard state. In several notable cases, high-resolution radio observa- tions after such flares have directly resolved radio- or even x-ray–emitting blobs moving away from the central black hole (16 , 17 ), which can be kinematically traced back to the time hard transition, although also sho ing a range of luminosities at which can occur (even in the same sourc generally occurs at a luminosity o few percent of the Eddington lum nosity (24 ). In fact, the soft state h never been convincingly observ in any BHXRB at luminosities b low 1% Eddington. By the time t source reaches the canonical ha state again, with almost exactly t same spectral and timing charact istics as the initial hard state, the has reappeared, and the accreti disc wind is gone. Once in the ha state, the source decline continu typically below the detection lev of all-sky or regular x-ray monitorin and are observed only occasiona until their next outburst. These qu phases are not without interest, ho ever, for it is during these perio that—without the glare of the brig X-ray spectrum X-ray luminosity SOFT HARD A B C D E F Black Holes Fender & Belloni 2012 Science X-ray luminosity During this phase, the behavior of the jet, revealed by infrared and radio observations, also begins to change. hard transition, although also show- ing a range of luminosities at which it can occur (even in the same source), X-ray spectrum SOFT HARD B C Black Holes soft X-ray spectrum hard During this phase, the behavior of the jet, revealed by infrared and radio observations, also begins to change. The infrared emission drops almost as soon as the state transition begins (14 ), indicating a change in the jet properties(density andmagneticfield) close to the black hole. The radio emission begins to vary more dramatically, showing oscilla- tions and flare events superposed on an overall decline (8 , 15 ). At a cer- tain point, there are one or more large radio flares, which can be two or more orders of magnitude more luminous than the previous existing, steadier jet in the hard state. In several notable cases, high-resolution radio observa- tions after such flares have directly resolved radio- or even x-ray–emitting blobs moving away from the central black hole (16 , 17 ), which can be kinematically traced back to the time of the state transition. It has been re- cently shown that in some cases, the ejection is coincident in time with the appearance of the strong QPOs (15 ). The soft state (D → E). As the spectral transition continues, these strong QPOs disappear, and the over- all level of x-ray variability drops as hard transition, althou ing a range of luminos can occur (even in the generally occurs at a few percent of the E nosity (24 ). In fact, th never been convinci in any BHXRB at lu low 1% Eddington. B source reaches the c state again, with alm same spectral and tim istics as the initial ha has reappeared, and disc wind is gone. O state, the source dec typically below the d of all-sky or regular x- and are observed on until their next outbur phases are not withou ever, for it is during that—without the gla accretion disc—resea to accurately measure tions of the compan optical telescopes an mate the mass of the self (25 , 26 ). These cycles of clear changes in the w X-ray spectrum X-ray luminosity SOFT HARD A B C D E F Black Holes GRS 1915+105 low/hard XRB low-luminosity AGNs, Sgr A* N+06; N+14 Open question: Nature of state transitions? Jet-disk connection