We have obtained new solutions and methods for the process of thermal Comptonization. We modify the solution to the kinetic equation of Sunyaev \& Titarchuk to allow its application up to mildly relativistic electron temperatures and optical depths $\gtrsim{1}$. The solution can be used for spectral fitting of X-ray spectra from astrophysical sources. We also have developed an accurate Monte Carlo method for calculating spectra and timing properties of thermal-Comptonization sources. The accuracy of our kinetic-equation solution is verified by comparison with the Monte Carlo results. We also compare our results with those of other publicly available methods. Furthermore, based on our Monte Carlo code, we present distributions of the photon emission times and the evolution of the average photon energy for both up and down-scattering.

We investigate the observed correlation between the 2--10 keV X-ray luminosity (in unit of the Eddington luminosity; $l_X \equiv L_X/L_{Edd}$) and the photon index ($\Gamma$) of the X-ray spectrum for both black hole X-ray binaries (BHBs) and active galactic nuclei (AGNs). We construct a large sample, with $10^{-9} < l_X < 10^{-1}$. We find that $\Gamma$ is positively and negatively correlated with $l_X$ when $l_X > 10^{-3}$ and $10^{-6.5} < l_X < 10^{-3}$ respectively, while $\Gamma$ is nearly a constant when $l_X < 10^{-6.5}$. We explain the above correlation in the framework of a coupled hot accretion flow -- jet model. The radio emission always come from the jet while the X-ray emission comes from the accretion flow and jet when $l_X$ is above and below $10^{-6.5}$, respectively. More specifically, we assume that with the increase of mass accretion rate, the hot accretion flow develops into a clumpy and further a disc -- corona two-phase structure because of thermal instability. We argue that such kind of two-phase accretion flow can explain the observed positive correlation.

Active galactic nuclei (AGN) and galactic black hole binaries (GBHs) represent two classes of accreting black holes. They both contain an accretion disc emitting a thermal radiation, and a non-thermal X-ray emitting corona. GBHs exhibit state transitions, and their spectral states are characterized by different luminosity levels and shapes of the spectral energy distribution (SED). If AGN transitioned in a similar way, the characteristic timescales of such transitions would exceed ~10^5 years. Thus, the probability to observe an individual AGN transiting between different spectral states is very low. In this paper we follow a spectral evolution of a GBH, GRO~J1655-40, and then apply its SED evolution pattern to a simulated population of AGN under an assumption that a large sample of AGN should contain a mixture of sources in different spectral states. We model the X-ray spectra of GRO J1655-40 with the eqpair model and then scale the best-fitting models with the black hole mass to simulate the AGN spectra. We compare the simulated and observed AGN SEDs to determine the spectral states of observed Type 1 AGN, LINER and NLS1 populations. We conclude that bright Type 1 AGN and NLS1 galaxies are in a spectral state similar to the soft spectral state of GBHs, while the spectral state of LINERs may correspond to the hard spectral state of GBHs. We find that taking into account a spread in the black hole masses over several orders of magnitude, as in the observed AGN samples, leads to a correlation between the X-ray loudness and the monochromatic luminosity at 2500A, similar to that found in observations. We predict however that the correlation changes its sign around a critical luminosity of L_crit ~ 0.01 L_E.

We study a large sample of RXTE PCA/HEXTE observations of Cyg X-1. We characterize the spectra by soft and hard X-ray colours (which define the spectral states), and fit them with a physical model of hybrid, thermal/non-thermal Comptonization. We then fit the power spectra by a sum of Lorentzians. We show the resulting correlations between the spectral colour (defining the hard, intermediate and soft spectral states), the Comptonization amplification factor, the fractional Compton reflection strength, and the peak frequency of the Lorentzians. We also calculate the fractional variability (rms) as a function of photon energy. We fit the obtained rms dependencies by physical models of varying accretion rate, a varying soft-photon input from an outer disc to a hot inner plasma, and a varying dissipation rate in a hot corona above a disc.

We consider implications of a possible presence of a Thomson-thick, low-temperature, plasma cloud surrounding the compact object in the binary system Cyg X-3. The presence of such a cloud was earlier inferred from the energy-independent orbital modulation of the X-ray flux and the lack of high frequencies in its power spectra. Here, we study the effect of Compton scattering by the cloud on the X-ray energy and power spectra, concentrating on the hard spectral state. The process reduces the energy of the high-energy break/cut-off in the energy spectra, which allows us to determine the Thomson optical depth. This, together with the observed cut-off in the power spectrum, determines the size of the plasma to be 2x10^9 cm. At this size, the cloud will be in thermal equilibrium in the photon field of the X-ray source, which yields the cloud temperature of 3 keV, which refines the determination of the Thomson optical depth to 7. At these parameters, thermal bremsstrahlung emission of the cloud becomes important as well. The physical origin of the cloud is likely to be collision of the very strong stellar wind of the companion Wolf-Rayet star with a small accretion disc formed by the wind accretion. Our model thus explains the peculiar X-ray energy and power spectra of Cyg X-3.

We draw a comparison between AGN and Galactic black hole binaries using a uniform description of spectral energy distribution of these two classes of accreting X-ray sources. We parametrize spectra of GBHs with an alpha_GBH parameter which we define as a slope of a nominal power law function between 3 and 20 keV. We show that this parameter can be treated as an equivalent of the X-ray loudness, alpha_OX, used to describe AGN spectra. We do not find linear correlation between the alpha_GBH and disc flux (similar to that between alpha_OX and optical/UV luminosity found in AGN). Instead, we show that alpha_GBH follows a well defined pattern during a GBH outburst. We find that alpha_GBH tend to cluster around 1, 1.5 and 2, which correspond to a hard, very high/intermediate and soft spectral state, respectively. We conclude that majority of the observed Type 1 radio quiet AGN are in a spectral state corresponding to a very high/intermediate state of GBHs. The same conclusion is valid for radio loud AGN. We also study variations of the spectral slopes (alpha_GBH and the X-ray photon index, Gamma) as a function of disc and Comptonization fluxes. We discuss these dependencies in the context of correlations of alpha_OX and Gamma with the optical/UV and X-ray 2 keV fluxes considered for AGN and quasars.

We build a simple model of the optical/UV emission from irradiation of the outer disc by the inner disc and coronal emission in black hole binaries. We apply this to the broadband Swift data from the outburst of the black hole binary XTE J1817--330 to confirm previous results that the optical/UV emission in the soft state is consistent with a reprocessing a constant fraction of the bolometric X-ray luminosity. However, this is very surprising as the disc temperature drops by more than a factor 3 in the soft state, which should produce a marked change in the reprocessing efficiency. The easiest way to match the observed constant reprocessed fraction is for the disc skin to be highly ionized (as suggested 30 years ago by van Paradijs), so that the bulk of the disc flux is reflected and only the hardest X-rays heat the disc. The constant reprocessed fraction also favours direct illumination of the disc over a scattering origin as the optical depth/solid angle of any scattering material (wind/corona) over the disc should decrease as the source luminosity declines. By contrast, the reprocessed fraction increases very significantly (by a factor ~6) as the source enters the hard state. This dramatic change is not evident from X-ray/UV flux correlations as it is masked by bandpass effects. However, it does not necessarily signal a change in emission e.g. the emergence of the jet dominating the optical/UV flux as the reflection albedo must change with the dramatic change in spectral shape.

We review X-ray flux modulation from X-ray binaries on time scales corresponding to the orbital period and those at longer time scales (so called superorbital). Those modulations provide a powerful tool to constrain geometry of the accretion flow. The most common cause of the superorbital variability appears to be precession. We then discuss two specific examples of discoveries of a coupling between the two types of variability and their physical interpretation. One is Cyg X-1, a black-hole binary with a high-mass companion, in which case we find the presence of an accretion bulge formed by collision of the stellar wind with the outer edge of the precessing accretion disc. The other is 4U 1820-303, a neutron star accreting from a low-mass white dwarf, in which case we interpret the superorbital variability as accretion rate modulation induced by interactions in a triple stellar system. Then, the varying accretion rate leads to changes of the size of the accretion bulge in that system, obscuring the centrally-emitted X-rays.

The X-ray quasi-periodic oscillation (QPO) seen in RE J1034+396 is so far unique amongst AGN. Here we look at the another unique feature of RE J1034+396, namely its huge soft X-ray excess, to see if this is related in any way to the detection of the QPO. We show that all potential models considered for the soft energy excess can fit the 0.3-10 keV X-ray spectrum, but that the energy dependence of the rapid variability (which is dominated by the QPO) strongly supports a spectral decomposition where the soft excess is from low temperature Comptonization of the disc emission and remains mostly constant, while the rapid variability is produced by the power law tail changing in normalization. The presence of the QPO in the tail rather than in the disc is a common feature in black hole binaries, but low temperature Comptonization of the disc spectrum is not generally seen in these systems. The main exception to this is GRS 1915+105, the only black hole binary which routinely shows super-Eddington luminosities. We speculate that super-Eddington accretion rates lead to a change in disc structure, and that this also triggers the X-ray QPO.

Active galactic nuclei and quasars are thought to be scaled up versions of Galactic black hole binaries, powered by accretion onto supermassive black holes with masses of 10^6-10^9 M_Sun, as opposed to the ~10 M_Sun in binaries. One example of the similarities between these two types of systems is the characteristic rapid X-ray variability seen from the accretion flow. The power spectrum of this variability in black hole binaries consists of a broad, band-limited noise with multiple quasi-periodic oscillations superimposed, where power is concentrated over a narrow range of frequencies. Although the broad noise component has been observed in many active galactic nuclei, there are no significant detections of quasi-periodic oscillations. Here we report the discovery of a ~1h X-ray periodicity in a bright active galaxy RE J1034+396. The signal is highly statistically significant (at the 5.6 sigma level) and very coherent, with quality factor Q > 16. This reinforces the link between stellar and supermassive black holes, emphasizing the universal properties of accretion onto objects with very different masses. The X-ray modulation arises from the direct vicinity of the black hole, so this provides a new tool for studying active galactic nuclei.

The key aspect of the very successful truncated disc model for the low/hard X-ray spectral state in black hole binaries is that the geometrically thin disc recedes back from the last stable orbit at the transition to this state. This has recently been challenged by direct observations of the low/hard state disc from CCD data. We reanalyze the Swift and RXTE campaign covering the 2006 outburst of XTE J1817-330 and show that these data actually strongly support the truncated disc model as the transition spectra unambiguously show that the disc begins to recede as the source leaves the disc dominated soft state. The disc radius inferred for the proper low/hard state is less clear-cut, but we show that the effect of irradiation from the energetically dominant hot plasma leads to an underestimate of the disc radius by a factor of 2-3 in this state. This may also produce the soft excess reported in some hard-state spectra. The inferred radius becomes still larger when the potential difference in stress at the inner boundary, increased colour temperature correction from incomplete thermalization of the irradiation, and loss of observable disc photons from Comptonization in the hot plasma are taken into account. We conclude that the inner disc radius in XTE J1817-330 in the low/hard spectral state is at least 6-8 times that seen in the disc dominated high/soft state, and that recession of the inner disc is the trigger for the soft--hard state transition, as predicted by the truncated disc models.

There is increasing evidence that supermassive black holes in active galactic nuclei (AGN) are scaled-up versions of Galactic black holes. We show that the amplitude of high-frequency X-ray variability in the hard spectral state is inversely proportional to the black hole mass over eight orders of magnitude. We have analyzed all available hard-state data from RXTE of seven Galactic black holes. Their power density spectra change dramatically from observation to observation, except for the high-frequency (>10 Hz) tail, which seems to have a universal shape, roughly represented by a power law of index -2. The amplitude of the tail, C_M (extrapolated to 1 Hz), remains approximately constant for a given source, regardless of the luminosity, unlike the break or QPO frequencies, which are usually strongly correlated with luminosity. Comparison with a moderate-luminosity sample of AGN shows that the amplitude of the tail is a simple function of black hole mass, C_M = C/M, where C = 1.25 M_Sol / Hz. This makes C_M a robust estimator of the black hole mass which is easy to apply to low- to moderate-luminosity supermassive black holes. The high-frequency tail with its universal shape is an invariant feature of a black hole and, possibly, an imprint of the last stable orbit.

(abridged) The soft excess seen in the X-ray spectra of many high mass accretion rate AGN can be well modelled by reflection from a partially ionised accretion disc. However, the derived parameters are often extreme, both in terms of the underlying spacetime and the reflection geometry, and these models require that the disc is far from hydrostatic equilibrium. An alternative model uses similarly partially ionised, velocity smeared material but seen in absorption, though again the derived velocities are extreme, requiring magnetic driving (in the jet?) rather than a simple line driven disc wind. We find that while both models give comparably good fits to XMM--Newton data, we favour the absorption model as, unlike reflection, all the derived spectral indices are soft. This is as expected by analogy with the correspondingly high mass accretion rate stellar mass black holes. Furthermore, these X-ray spectra are consistent with a one--to--one mapping between AGN type and spectral state, with NLS1's having softer spectra corresponding to the very high state, while the broad line AGN have Gamma~2 as expected for the high/soft state. We also use the simultaneous OM data to derive the ratio of disc to total accretion power which is another tracer of spectral state in X-ray binaries. This does not always show that the disc in NLS1's contributes less than 80 per cent of the total power, as expected for a very high state. We suggest that this is an artifact of the standard disc models used to determine the disc luminosity in our fits. The disc seen in the very high state of black hole binaries is often observed to be distorted from the standard shape, and a similar effect in NLS1's could recover the correspondance between black hole binary spectral state and AGN type.

(abridged) We review how the recent increase in X-ray and radio data from black hole and neutron star binaries can be merged together with theoretical advances to give a coherent picture of the physics of the accretion flow in strong gravity. Both long term X-ray light curves, X-ray spectra, the rapid X-ray variability and the radio jet behaviour are consistent with a model where a standard outer accretion disc is truncated at low luminosities, being replaced by a hot, inner flow which also acts as the launching site of the jet. Decreasing the disc truncation radius leads to softer spectra, as well as higher frequencies (including QPO's) in the power spectra, and a faster jet. The collapse of the hot flow when the disc reaches the last stable orbit triggers the dramatic decrease in radio flux, as well as giving a qualitative (and often quantitative) explanation for the major hard--soft spectral transition seen in black holes and neutron stars. After collapse of the hot inner flow, the spectrum in black hole systems can be dominated by the disc emission. Its behaviour is consistent with the existence of a last stable orbit, and such data can be used to estimate the black hole spin. These systems can also show very different spectra at these high luminosities, in which the disc spectrum is strongly distorted by Comptonization. The structure of the accretion flow becomes increasingly uncertain as the luminosity approaches (and exceeds) the Eddington luminosity, though there is growing evidence that winds play an important role. We stress that these high Eddington fraction flows are key to understanding many disparate and currently very active fields such as ULX, Narrow Line Seyfert 1's, and the growth of the first black holes in the Early Universe.

We discuss various ideas for the origin of the soft X-ray excess seen in AGN. There are clear advantages to models where this arises from atomic processes in partially ionised rather than where it is a true continuum component. However, current data cannot distinguish between models where this material is seen in reflection or absorption. While higher energy data may break the degeneracies, we also suggest that strong outflows are extremely likely to be present, lending more physical plausibility to an absorption origin. This more messy picture of NLS1's means that they are probably not good places to test GR, but they do give insight into the spectra expected from the first QSO's in the early Universe.

We systematically analyze all the available X-ray spectra of disc accreting neutron stars (atolls and millisecond pulsars) from the RXTE database. We show that while these all have similar spectral evolution as a function of mass accretion rate, there are also subtle differences. There are two different types of hard/soft transition, those where the spectrum softens at all energies, leading to a diagonal track on a colour-colour diagram, and those where only the higher energy spectrum softens, giving a vertical track. The luminosity at which the transition occurs is correlated with this spectral behaviour, with the vertical transition at L/L_Edd ~0.02 while the diagonal one is at ~0.1. Superimposed on this is the well known hysteresis effect, but we show that classic, large scale hysteresis occurs only in the outbursting sources, indicating that its origin is in the dramatic rate of change of mass accretion rate during the disc instability. We show that the long term mass accretion rate correlates with the transition behaviour, and speculate that this is due to the magnetic field being able to emerge from the neutron star surface for low average mass accretion rates. While this is not strong enough to collimate the flow except in the millisecond pulsars, its presence may affect the inner accretion flow by changing the properties of the jet.

We report the discovery, using RXTE data, of a dependence of the X-ray orbital modulation depth on the X-ray spectral state in the ultracompact atoll binary \source. This state (measured by us by the position on the X-ray colour-colour diagram) is tightly coupled to the accretion rate, which, in turn, is coupled to the phase of the 170-d superorbital cycle of this source. The modulation depth is much stronger in the high-luminosity, so-called banana, state than in the low-luminosity, island, state. We find the X-ray modulation is independent of energy, which rules out bound-free X-ray absorption in an optically thin medium as the cause of the modulation. We also find a significant dependence of the offset phase of the orbital modulation on the spectral state, which favours the model in which the modulation is caused by scattering in hot gas around a bulge at the disc edge, which both size and the position vary with the accretion rate. Estimates of the source inclination appear to rule out a model in which the bulge itself occults a part of an accretion disc corona. We calculate the average flux of this source over the course of its superorbital variability (which has the period of 170 d), and find it to be fully compatible with the model of accretion due to the angular momentum loss via emission of gravitational radiation. Also, we compare the dates of all X-ray bursts observed from this source by BeppoSAX and RXTE with the RXTE light curve, and find all of them to coincide with deep minima of the flux, confirming previous results based on smaller samples.

We perform a comprehensive analysis of the superorbital modulation in the ultracompact X-ray source 4U 1820-303, consisting of a white dwarf accreting onto a neutron star. Based on RXTE data, we measure the fractional amplitude of the source superorbital variability (with a 170-d quasi-period) in the folded and averaged light curves, and find it to be by a factor of about 2. As proposed before, the superorbital variability can be explained by oscillations of the binary eccentricity. We now present detailed calculations of the eccentricity-dependent flow through the inner Lagrangian point, and find a maximum of the eccentricity of about 0.004 is sufficient to explain the observed fractional amplitude. We then study hierarchical triple models yielding the required quasi-periodic eccentricity oscillations through the Kozai process. We find the resulting theoretical light curves to match well the observed ones. We constrain the ratio of the semimajor axes of the outer and inner systems, the component masses, and the inclination angle between the inner and outer orbits. Last but not least, we discover a remarkable and puzzling synchronization between the observed period of the superorbital variability (equal to the period of the eccentricity oscillations in our model) and the period of the general-relativistic periastron precession of the binary.

Narrow Line Seyfert 1 galaxies are generally accreting at high fractions of the Eddington limit. They can show complex X-ray spectra, with a strong `soft excess' below 2 keV and a sharp drop at ~7 keV. There is strong evidence linking the soft excess to either reflection or absorption from relativistic, partially ionized material close to the black hole. The reflection models can also simultaneously produce the 7 keV feature from fluorescent iron Ka line emission from the disc. Here we show that absorption can also produce a sharp feature at 7 keV from the P Cygni profile which results from absorption/scattering/emission of He- and H-like iron Ka resonance lines in the wind. We demonstrate this explicitly by fitting the iron feature seen in XMM-Newton data from 1H 0707-495 to a P Cygni profile. The resulting column and ionization required to produce this feature are probably larger than those needed to produce the soft excess. Nonetheless, the absorbing material could still be a single structure with stratified ionization such as that produced by the ionization instability.

Microquasars are galactic black hole binary systems with radio jets which can sometimes be spatially resolved to show superluminal motion. The first and best known of this class of objects is GRS 1915+105, the brightest accreting source in our Galaxy. There is persistent speculation that strong jet emission could be linked to black hole spin. If so, the high spin should also be evident in accretion disc spectra. We search the RXTE archive to find disc-dominated X-ray spectra from this object, as these are the only ones which can give reliable spin determinations by this method. Finding these is complicated by the rapid, unique limit cycle variability, but we are able to identify such spectra by going to the shortest possible time resolution (16 s). We fit them with a simple multicolour disc blackbody diskbb, and with the best current model which include full radiative transfer as well as relativistic effects bhspec. Both these models show that the spin is intermediate, neither zero nor maximal. bhspec, the most physical model, gives a value for the dimensionless spin of a* \~ 0.7 for a distance of 12.5 kpc and inclination of 66 deg. This, together with the range of spins 0.1 < a* < 0.8 derived using this method for other black holes, suggests that jet emission is probably fundamentally powered by gravity rather than spin, and implies that high-to-maximal spin is not a pre-requisite for powerful relativistic jets.

Values of black hole masses are frequently determined with the help of the reverberation method. This method requires a specific geometrical factor related to the distribution of the orbits of the Broad Line Region clouds. Onken et al. determined the value f^2= 1.37+/-0.45 from the black hole mass - dispersion relation. In this paper we determine this factor using an independent mass determination from the X-ray variance method for a number of Seyfert 1 galaxies and comparing them with the reverberation results by Peterson et al. We obtain mean value f^2 = 1.12 +/- 0.54, consistent with Onken et al. Both values are larger than the value 0.75 corresponding to a spherical geometry. It indicates that most probably all values of the black hole masses obtained with the use of the Kaspi et al. formulae should be multiplied by a factor of \sim 1.7. This also shows that the Broad Line Region is rather flat, and hints for a dependence of the factor f^2 on a source inclination seem to be present in the data.

The origin of the soft excess seen in many AGN below ~1 keV is still an unsolved problem. It is unlikely to represent a true continuum component as its characteristic 'temperature' shows a remarkable constancy over a wide range of AGN luminosity and black hole mass. This instead favours an association with atomic processes, in particular with the increase in opacity between 0.7-2 keV associated with partially ionized O and Fe. The opacity jump can give rise to a soft excess either through reflection or transmission, and both scenarios can fit the spectra equally well as long as there is strong velocity shear to smear out the characteristic narrow atomic features. Here we use orthogonal constraints from the energy-dependent variability. The rms spectra seen in XMM-Newton AGN data often show a broad peak between 0.7-2 keV. We show that the absorption model can explain the data well if the ionization state of the smeared absorption responds to luminosity changes in the continuum.

We have analysed X-ray outbursts from several Galactic black hole (GBH) transients, as seen by the ASM on board RXTE. We have used the best estimates of distance and black hole mass to find their luminosity (scaled to the Eddington limit), which allowed for direct comparison of many sources. We have found two distinct hard-to-soft state transitions in the initial part of the outburst. The distinction is made on the basis of the transition luminosity, its duration, the shape of the track in the hardness-luminosity diagram, and evolution of the hardness ratio. The bright/slow transition occurs at ~30 per cent of Eddington (estimated bolometric) luminosity and takes >~30 days, during which the source quickly reaches the intermediate/very high state and then proceeds to the soft state at much slower pace. The dark/slow transition is less luminous (<~10 per cent of Eddington), shorter (<~15 days) and the source does not slow its transition rate before reaching the soft state. We speculate that the distinction is due to irradiation and evaporation of the disc, which sustains the Comptonizing corona in the bright intemediate/very high state.

There is growing evidence from both spectral and timing properties that there is a truncated inner accretion disc in low mass accretion rate Galactic black hole systems. The detection of extremely smeared relativistic iron lines in some of these systems is the only current piece of evidence which conflicts with this geometrical interpretation of the low/hard state. Here we show that the line width in the BeppoSAX data of a bright low/hard state of the transient black hole XTE J1650-500 is indeed consistent with extreme relativistic effects. However, the relativistic smearing can be significantly reduced if there is also resonance iron K line absorption from an outflowing disc wind. The iron line smearing is then completely compatible with a truncated disc, so gives no information on the black hole spin.

The galactic black hole binary systems give an observational template showing how the accretion flow changes as a function of increasing mass accretion rate, or L/L_Edd. These data can be synthetised with theoretical models of the accretion flow to give a coherent picture of accretion in strong gravity, in which the major hard-soft spectral transition is triggered by a change in the nature and geometry of the inner accretion flow from a hot, optically thin plasma to a cool, optically thick accretion disc. However, a straightforward application of these models to AGN gives clear discrepancies in overall spectral shape. Either the underlying accretion model is wrong, despite its success in describing the Galactic systems and/or there is additional physics which breaks the simple scaling from stellar to supermassive black holes.

We investigate the correspondence between the variability seen in the stellar and supermassive black holes. Galactic Black Hole (GBH) power density spectra (PDS) are generally complex, and dependent on spectral state. In the low/hard state the high-frequency rollover in the PDS moves in a way which is not simply related to luminosity. Hence this feature can only be used as an approximate indicator rather than as an accurate tracer of black hole mass in AGN. The X-ray spectrum in the high/soft state is dominated by the disc in GBH, which is rather stable. We show that the PDS of the Comptonized tail in GBH can be much more variable, and that it is this which should be compared to AGN due to their much lower disc temperature. This bandpass effect removes a problem in interpreting the (often highly variable) Narrow Line Seyfert 1 (NLS1) galaxies as the counterparts of the high mass accretion rate GBH. Additionally, we speculate that some NLS1 (e.g. Akn 564) are counterparts of the very high state. The Comptonized tail in this state is also highly variable, but with PDS which can be roughly described as band-limited noise. This shape is similar to that seen in the low/hard state, so merely seeing such band-limited noise in the power spectrum of an AGN does not necessarily imply low luminosity. We also stress that Cyg X-1, often used for comparison with AGN, is not a typical black hole system due to its persistent nature. In particular, the shape of its power spectrum in the high/soft state is markedly different from that of other (transient) GBH systems in this state. The fact that the NLS1s NGC 4051 and MCG-6-30-15 do appear to show a power spectrum similar to that of the high/soft state of Cyg X-1 may lend observational support to theoretical speculation that the Hydrogen ionization disc instability does not operate in AGN.

We study fractional variability as a function of energy from black-hole X-ray binaries on timescales from milliseconds to hundreds of seconds. We build a theoretical model of energy-dependent variability in which the X-ray energy spectrum varies in response to a changing physical parameter. We compare these models to rms spectra obtained from RXTE PCA observations of black-hole binaries XTE J1550-564 and XTE J1650-500. We show that two main variability models are consistent with the data: variable seed photon input in the hard state and variable power in the Comptonized component in the soft and very high states. The lack of clear reflection features in the rms spectra implies that the reflection and the X-ray continuum, when integrated over Fourier frequencies, are correlated and vary with similar fractional amplitudes. Our models predict two important features of rms spectra, not possible to be clearly seen by the PCA due sensitivity limits. At soft X-rays, >~3 keV, we predict the presence of a break in the rms spectrum at energy directly related to the seed photon temperature. At higher energies, ~20--30 keV, we predict a peak in the rms spectrum originating from the variability of the spectrum produced by a hybrid thermal/non-thermal electron distribution. If these features are confirmed by broad-band observations, they will impose important constraints on the origin of the seed photons for Comptonization and the electron distribution in the hot plasma.

We present an exhaustive analysis of five broad-band observations of GRS 1915+105 in two variability states, chi and omega, observed simultaneously by the PCA, HEXTE and OSSE. We find all the spectra well fitted by Comptonization of disc blackbody photons, with very strong evidence for the presence of a nonthermal electron component in the Comptonizing plasma. Both the energy and the power spectra in the chi state are typical to the very high/intermediate state of black-hole binaries. The spectrum of the omega state is characterized by a strong blackbody component Comptonized by thermal electrons and a weak nonthermal tail. We then calculate rms spectra (fractional variability as functions of energy) for the PCA data. We accurately model the rms spectra by coherent superposition of variability in the components implied by the spectral fits, namely a less variable blackbody and more variable Comptonization. The latter dominates at high energies, resulting in a flattening of the rms at high energies in most of the data. This is also the case for the spectra of the QPOs present in the chi state. Then, some of our data require a radial dependence of the rms of the disc blackbody. We also study the distance to the source, and find 11 kpc as the most likely value, contrary to a recent claim of a much lower value.

We have undertaken an extensive study of X-ray data from the accreting millisecond pulsar XTE J1751-305 observed by RXTE and XMM-Newton during its 2002 outburst. In all aspects this source is similar to a prototypical millisecond pulsar SAX J1808.4-3658, except for the higher peak luminosity of 13 per cent of Eddington, and the optical depth of the hard X-ray source larger by factor ~2. Its broad-band X-ray spectrum can be modelled by three components. We interpret the two soft components as thermal emission from a colder (kT ~ 0.6 keV) accretion disc and a hotter (~1 keV) spot on the neutron star surface. We interpret the hard component as thermal Comptonization in plasma of temperature ~40 keV and optical depth of ~1.5 in a slab geometry. The plasma is heated by the accretion shock as the material collimated by the magnetic field impacts on to the surface. The seed photons for Comptonization are provided by the hotspot, not by the disc. The Compton reflection is weak and the disc is probably truncated into an optically thin flow above the magnetospheric radius. Rotation of the emission region with the star creates an almost sinusoidal pulse profile with rms amplitude of 3.3 per cent. The energy-dependent soft phase lags can be modelled by two pulsating components shifted in phase, which is naturally explained by a different character of emission of the optically thick spot and optically thin shock combined with the action of the Doppler boosting.

We review radiative processes responsible for X-ray emission in hard (low) and soft (high) spectral states of black-hole binaries. The main process in the hard state appears to be scattering of blackbody photons from a cold disk by thermal electrons in a hot inner flow, and any contribution from nonthermal synchrotron emission is at most small. In the soft states, blackbody disk emission dominates energetically, and its high-energy tail is due to scattering by hybrid, thermal/nonthermal electrons, probably in active regions above the disk surface. State transitions appear to correspond to a variable inner radius of the cold disk driven by changes of the accretion rate. The existence of two accretion solutions, hot and cold, in a range of the accretion rate leads to hysteresis in low-mass X-ray binaries.

We systematically analyzed about a terabyte of data from Galactic binary systems to characterize how these spectra change as a function of mass accretion rate. With this volume of data we are able to identify the differences in behaviour of black holes and neutron stars, implying that black holes really do have an event horizon. The spectral evolution seen as a function of L/L_Edd can be linked to current state-of-the-art numerical simulations of accretion flows, and we develop qualitative and quantitative models of the spectral changes in all types of binary systems. The key ingredient of this picture is that the major hard-soft spectral transition is driven by a changing inner radius of the accretion disc. This also qualitatively explains the changes seen in the variability power spectra of these objects. This lays the foundation for a self-consistent, theoretically motivated framework in which to explain both the spectral and timing properties of accreting compact objects.

We study X-ray and variability and distance of GX 339-4. We derive d>7 kpc, based on recent determination of the binary parameters. We study data from the Ginga/ASM, the CGRO/BATSE, and the RXTE/ASM, PCA and HEXTE. From 1987 to 2004, GX 339-4 underwent 15 outbursts and went through all known states of black-hole binaries. We also present the PCA data from the initial stage of the 2004 outburst. We then study colour-colour and colour-flux correlations. In the hard state, there is a strong anticorrelation between the 1.5-5 and 3-12 keV spectral slopes, which we explain by thermal Comptonization of disc photons. There is also a softening of the spectrum above 3 keV with the increasing flux that becomes stronger with increasing energy up to 200 keV. This indicates an anticorrelation between the electron temperature and luminosity, explained by hot accretion models. In addition, we see a variable broad-band slope with a pivot at 200 keV. We confirm the presence of pronounced hysteresis, with the hard-to-soft state transitions occurring at much higher (and variable) luminosities than the soft-to-hard transitions. We fit the ASM data with a model consisting of an outer accretion disc and a hot inner flow. State transitions are associated then with variations in the disc truncation radius, which we fit as 6GM/c^2 in the soft state and several times that in the hard state. The disappearence of the inner disc takes place at a lower accretion rate than its initial appearance due to the dependence of the transitions on the source history. We provide further evidence against the X-ray emission in the hard state being nonthermal synchrotron, and explain the observed radio-X-ray correlation by the jet power being correlated with the accretion power.

We systematically analyse all publicly available XMM-Newton spectra of radio-quiet PG quasars. The soft X-ray excess in these objects is well modelled by an additional, cool, Compton scattering region. However, the remarkably constant temperature derived for this component over the whole sample requires a puzzling fine tuning of the parameters. Instead, we propose that the soft excess is an artifact of strong, relativistically smeared, partially ionized absorption. The strong jump in opacity at ~0.7 keV from OVII, OVIII and iron can lead to an apparent soft excess below this energy, which is trivially constant since it depends on atomic processes. This can have a dramatic effect on the derived spectrum, which has implications for fitting the relativistic smearing of the reflected iron line emission from the disc.

We compare the evolution of spectral shape with luminosity in GRS 1915+105 with that of 'normal' black holes. The pathological time variability of GRS 1915+105 does not indicate that it belongs in a different class to all the other objects. At comparable fraction of Eddington luminosity its spectra and (more importantly) time variability behaviour are similar to that seen in the 'normal' black holes. Its unique limit-cycle variability only appears when it radiates at uniquely high (super-Eddington) luminosities.

Disc spectra from highly luminous black hole binaries are observed to be rather simple, despite theoretical predictions to the contrary. We collate the disc-dominated spectra from multiple observations of 10 separate sources, and show that these overwhelmingly follow a multi-temperature disc blackbody spectrum with luminosity proportional to T^4, and that such discs are stable. These results are in conflict with standard Shakura-Sunyaev alpha-discs predictions, including proposed modifications such as additional energy loss in a jet, wind or corona. Accretion disc spectra can be useful, reliable guides to the disc structure, and should be used to test the next generation of accretion disc models, in which the disc viscosity is calculated self-consistently from the magnetically generated turbulent dynamo.

We have found a large number of very strong flares in the available XTE PCA data of Cyg X-1 (also seen in available HEXTE and BATSE data) with 13 flares satisfying our chosen threshold criterion, occuring both in the hard and the soft states. We analyze here in detail two of them. The strongest one took place in the soft state, with the 3-30 keV energy flux increasing 30 times with respect to the preceding 16-s average. The e-folding time is ~ 7 ms for the main flare and ~ 1 ms for its precursor. The spectrum strongly hardens during the flare. On the other hand, flares in the hard state have generally lower amplitudes and longer e-folding times, and their spectra soften during the flare, with the hardness of the spectrum at the flare peak similar for both types of the flares. The presence of the flares shows unusually dramatic events taking place in the accretion flow of Cyg X-1. On the other hand, the rate of occurence of hard-state flares shows they may represent a high-flux end of the distribution of shots present in usual lightcurves of Cyg X-1.

The key difference between neutron stars and black holes is the presence/absence of a solid surface. Recent attempts to detect this difference have concentrated on the quiescent luminosity, but here these sources are faint and difficult to observe. Instead we look at these sources when they are bright, and show that there is a clear difference between black holes and neutron stars in the evolution of their X-ray spectra which is due to the presence of a surface in the neutron stars. We also show that there is a type of X-ray spectrum which is only seen from black holes, making it a good diagnostic for the nature of new transient sources.

We fit the broad-band X/gamma-ray spectrum (0.8-1000 keV) of the accreting black hole XTE J1550-564 in the very high state. The quasi-simultaneous data from ASCA, RXTE and OSSE show that the disc is strongly Comptonized, with a high energy tail extending out to several hundreds keV. However, inverse Compton scattering by a purely thermal or purely power-law electron distribution cannot explain the observed spectrum. Instead the data require a hybrid distribution, with both thermal and non-thermal electrons scattering the disc photons. This is very similar to the electron distribution inferred for other high and very high state black hole binaries, showing that it is a generic feature of high mass accretion rate black holes.

The pulse profiles of the accreting X-ray millisecond pulsar SAX J1808.4-3658 at different energies are studied. The two main emission component, the black body and the Comptonized tail that are clearly identified in the time-averaged spectrum, show strong variability with the first component lagging the second one. The observed variability can be explained if the emission is produced by Comptonization in a hot slab (radiative shock) of Thomson optical depth ~0.3-1 at the neutron star surface. The emission patterns of the black body and the Comptonized radiation are different: a "knife"- and a "fan"-like, respectively. We construct a detailed model of the X-ray production accounting for the Doppler boosting, relativistic aberration and gravitational light bending in the Schwarzschild spacetime. We present also accurate analytical formulae for computations of the light curves from rapidly rotating neutron stars using formalism recently developed by Beloborodov (2002). Our model reproduces well the pulse profiles at different energies simultaneously, corresponding phase lags, as well as the time-averaged spectrum. We constrain the compact star mass to be bounded between 1.2 and 1.6 solar masses. By fitting the observed profiles, we determine the radius of the compact object to be R~11 km if M=1.6 M_sun, while for M=1.2 M_sun the best-fitting radius is ~6.5 km, indicating that the compact object in SAX J1808.4-3658 can be a strange star. We obtain a lower limit on the inclination of the system of 65 degrees.

We follow the timing properties of the neutron star low-mass X-ray binary system 4U 1705-44 in different spectral states, as monitored by the Rossi X-ray Timing Explorer over about a month. We fit the power density spectra using multiple Lorentzians. We show that the characteristic frequencies of these Lorentzians, when properly identified, fit within the correlations previously reported. The time evolution of these frequencies and their relation with the parameters of the energy spectra reported in Barret & Olive (2002) are used to constrain the accretion geometry changes. The spectral data were fitted by the sum of a blackbody and a Comptonized component and were interpreted in the framework of a truncated accretion disk geometry, with a varying truncation radius. If one assumes that the characteristic frequencies of the Lorentzians are some measure of this truncation radius, as in most theoretical models, then the timing data presented here strengthen the above interpretation. The soft to hard and hard to soft transitions are clearly associated with the disk receding from and approaching the neutron star respectively. During the transitions, correlations are found between the Lorentzian frequencies and the flux and temperature of the blackbody, which is thus likely to be coming from the disk. On the other hand, in the hard state, the characteristic Lorentzians frequencies which are at the lowest, remained nearly constant despite significant evolution of the spectra parameters. The disk no longer contributes to the X-ray emission, and the blackbody is now likely to be emitted by the neutron star surface which is providing the seed photons for the Comptonization.

The transient atoll source 4U 1608-52 had been extensively observed by the Rossi X-ray Timing Explorer RXTE during its 1998 outburst. We analyse its X-ray spectra as a function of inferred accretion rate from both the 1998 outburst and from the non-outburst 1996 and 1998 data. We can fit all the spectra by a model in which seed photons from the neutron star surface are Comptonized in a boundary layer. The Comptonized emission illuminates the accretion disc surface, producing an ionized, relativistically broadened reflection signature, while the direct emission from the accretion disc can also be seen. The evolution of the source can be explained if the main parameter driving the spectral evolution is the average mass accretion rate, which determines the truncation radius of the inner accretion disc. At low mass accretion rates, in the island state, the disc truncates before reaching the neutron star surface and the inner accretion flow/boundary layer is mostly optically thin. The disc emission is at too low a temperature to be observed in the RXTE spectra, but some of the seed photons from the neutron star can be seen directly through the mostly optically thin boundary layer. At higher mass accretion rates, in the banana state, the disc moves in and the boundary layer becomes much more optically thick so its temperature drops. The disc can then be seen directly, but the seed photons from the neutron star surface cannot as they are buried beneath the increasingly optically thick boundary layer.

We analyse five observations of the X-ray binary GX 339-4 by the soft gamma-ray OSSE detector on board CGRO simultaneous with either Ginga or RXTE observations. The source was bright during four of them, with the luminosity of L ~ 10^{37} erg/s and the spectrum typical for hard states of accreting black holes, and it was in an off state during the fifth one, with L ~ 10^{35} erg/s. Our broad-band spectral fits show the mean electron energy of electrons in the Comptonizing plasma decreasing with increasing luminosity within the hard (bright) state. For the observation with the best statistics at soft gamma-rays, approximately 1/4 of energy in the Comptonizing plasma is probably carried by non-thermal electrons. Then, considering the efficiency of Comptonized hybrid synchrotron emission allows us to obtain an upper limit on the strength of the magnetic field in the X-ray source. Furthermore, this synchrotron emission is capable of producing the optical spectrum observed in an optically-high state of GX 339-4. In the off state, the hard X-ray spectrum is consistent with being dominated by bremsstrahlung. The unusually strong Fe K alpha line observed by the PCA during that state is found not to be intrinsic to the source but to originate mostly in the Galactic diffuse emission.

The pulse profiles of the accreting X-ray millisecond pulsar SAX J1808.4-3658 at different energies are studied. The two main emission component, a black body and a power-law tail, clearly identified in the time-averaged spectrum, do not vary in phase. We show that the observed variability can be easily explained if the emission patterns of the black body and the Comptonized radiation are different: a "knife" and a "fan"-like, respectively. We suggest that Comptonization in a hot slab (radiative shock) of Thomson optical depth \~0.3 at the surface of the neutron star may be responsible for the emission. We construct a detailed model of the X-ray production accounting for the Doppler boosting, relativistic aberration and gravitational light bending. The model reproduces well the pulse profiles at different energies simultaneously, corresponding phase lags, as well as the time-averaged spectrum. By fitting the observed pulse profiles we obtain constraints on the neutron star radius (R=7.5+-1.0 km), the inclination of the system i>60 deg, and the angle between the magnetic dipole and the rotational axis delta=20deg+-5deg.

Disc accreting neutron stars come in two distinct varieties, atolls and Z sources, named after their differently shaped tracks on a colour-colour diagram as the source luminosity changes. Here we present analysis of three transient atoll sources showing that there is an additional branch in the colour-colour diagram of atoll sources which appears at very low luminosities. This new branch connects to the top of previously known C-shaped (atoll) path, forming a horizontal track where the average source flux decrease from right to left. This turns the C-shape into a Z. Thus both atolls and Z sources share the same topology on the colour-colour diagram, and evolve in similar way as a function of increasing averaged mass accretion rate. This strongly favours models in which the underlying geometry of these sources changes in similar ways. A possible scenario is one where the truncated disc approaches the neutron star when the accretion rate increases, but that in the atolls the disc is truncated by evaporation (similarly to black holes) whereas in the Z sources it is truncated by the magnetic field.

We present new results based on RXTE observations of SAX J1808.4-3658. The X-ray spectrum can be fitted by a two-component model. We interpret the soft component as blackbody emission from a heated spot on the neutron star, and the hard component as coming from Comptonization in plasma heated by the accretion shock as the material collimated by the magnetic field impacts onto the neutron star surface. The hotspot is probably the source of seed photons for Comptonization. The hard component illuminates the disc, giving rise to a reflected spectrum. The inferred evolution of the inner radius is not consistent with the magnetic field truncating the disc. Instead it seems more likely that the inner disc radius is set by some much longer time-scale process, most probably connected to the overall evolution of the accretion disc. This disc truncation mechanism would then have to be generic in all low mass accretion rate flows both in disc accreting neutron stars and black hole systems. The phase resolved spectra show clearly that the blackbody and hard Comptonized spectra pulse independently. This obviously gives an energy dependent phase lag. Full general relativistic effects are not required to explain this. We do not detect Compton reflection from the neutron star surface, though we predict that it should be present in the X-ray spectrum. This would give an unambiguous observational measure of M/R if there is any iron on the neutron star surface which is not completely ionized.

We analyze the RXTE observations of the April 1998 outburst of SAX J1808.4-3658. We show that the soft blackbody component of the X-ray spectrum is from the hotspot on the neutron star surface, while the hard component is from Comptonization of these blackbody photons in the plasma heated by the accretion shock. The disc cannot provide enough seed photons for Comptonization. It is truncated into an optically thin flow at ~20-40 R_g by some generic mechanisms common for black holes and neutron stars. Only below the truncation radius the flow is disrupted by the magnetic field and collimated onto the surface. The pulse profiles of both spectral components imply several constraints on the geometry of the system: the inclination angle, location and size of the hotspot and the shock.

We present a general relativistic accretion disc model and its application to the soft-state X-ray spectra of black hole binaries. The model assumes a flat, optically thick disc around a rotating Kerr black hole. The disc locally radiates away the dissipated energy as a blackbody. Special and general relativistic effects influencing photons emitted by the disc are taken into account. The emerging spectrum, as seen by a distant observer, is parametrized by the black hole mass and spin, the accretion rate, the disc inclination angle and the inner disc radius. We fit the ASCA soft state X-ray spectra of LMC X-1 and GRO J1655-40 by this model. We find that having additional limits on the black hole mass and inclination angle from optical/UV observations, we can constrain the black hole spin from X-ray data. In LMC X-1 the constrain is weak, we can only rule out the maximally rotating black hole. In GRO J1655-40 we can limit the spin much better, and we find 0.68 < a < 0.88. Accretion discs in both sources are radiation pressure dominated. We don't find Compton reflection features in the spectra of any of these objects.

We present X-ray/gamma-ray spectra of Cyg X-1 observed during the transition from the hard to the soft state and in the soft state by ASCA, RXTE and OSSE in 1996 May and June. The spectra consist of a dominant soft component below ~2 keV and a power-law-like continuum extending to at least ~800 keV. We interpret them as emission from an optically-thick, cold accretion disc and from an optically-thin, non-thermal corona above the disc. A fraction f ~ 0.6 of total available power is dissipated in the corona. We model the soft component by multi-colour blackbody disc emission taking into account the torque-free inner-boundary condition. If the disc extends down to the minimum stable orbit, the ASCA/RXTE data yield the most probable black hole mass of about 10 solar masses and an accretion rate about 0.5 L_E/c^2, locating Cyg X-1 in the soft state in the upper part of the stable, gas-pressure dominated, accretion-disc solution branch. The spectrum of the corona is well modelled by repeated Compton scattering of seed photons from the disc off electrons with a hybrid, thermal/non-thermal distribution. The electron distribution can be characterized by a Maxwellian with an equilibrium temperature of kT ~ 30--50 keV and a Thomson optical depth of ~0.3 and a quasi-power-law tail. The compactness of the corona is between 2 and 7, and a presence of a significant population of electron-positron pairs is ruled out. We find strong signatures of Compton reflection from a cold and ionized medium, presumably an accretion disc, with an apparent reflector solid angle ~0.5--0.7. The reflected continuum is accompanied by a broad iron K-alpha line.

We analyze the simultaneous ASCA/RXTE observation of Cyg X-1 in the soft state on 1996 May 30. We apply a multi-color disk model with a torque-free boundary condition in the pseudo-Newtonian potential, and point out the main discrepancies between the commonly-used simplified `diskbb' model and our one. We estimate the black hole mass and the accretion rate, and show that during this observation the disk probably did not reach the last stable orbit and the transition from the hard to the soft state was not completed yet.

We present X-ray/gamma-ray spectra of the binary GX 339-4 observed in the hard state simultaneously by Ginga and CGRO OSSE during an outburst in 1991 September. The Ginga spectra are well represented by a power law with a photon spectral index of 1.75 and a moderately-strong Compton reflection component with a fluorescent Fe K alpha line. The OSSE data require a sharp high-energy cutoff in the power-law spectrum. The broad-band spectra are very well modelled by repeated Compton scattering in a thermal plasma with tau=1 and kT=50 keV. We also find the distance to the system to be > 3 kpc, ruling out earlier determinations of 1.3 kpc. Using this limit, the observed reddening and the orbital period, we find the allowed range of the mass of the primary is consistent with it being a black hole. The data are inconsistent with models of either homogenous or patchy coronae above the surface of an accretion disc. Rather, they are consistent with the presence of a hot inner hot disc accreting at a rate close to the maximum set by advection and surrounded by a cold outer disc. The seed photons for Comptonization are supplied by the outer cold disc and/or cold clouds within the hot disc. Pair production is negligible if electrons are thermal. The hot disc model, which scaled parameters are independent of the black-hole mass, is supported by the similarity of the spectrum of GX 339-4 to those of other black-hole binaries and Seyfert 1s. On the other hand, their spectra in the soft gamma-ray regime are significantly harder than those of weakly-magnetized neutron stars. Based on this difference, we propose that the presence of broad-band spectra corresponding to thermal Comptonization with kT of 50 keV or more represents a black-hole signature.

We analyze X-ray/gamma-ray observations of Cyg X-1 in the soft state. We find the hybrid thermal/non-thermal model fits the data very well. The significant contribution from the non-thermal electrons is required to account for the observed power law extending without a break to at least 800 keV. The presence of electron-positron pairs in the hot source is unlikely. We propose the geometry in which a cold accretion disk, surrounded by a hot corona, extends down to the last stable orbit. At the observed accretion rate the corona ensures the stability of the cold disk.

We present X-ray/gamma-ray observations of Cyg X-1 in the soft state during 1996 May-June. We analyze ASCA, RXTE and OSSE data. The spectrum consists of soft X-ray blackbody emission of an optically thick accretion disk in the vicinity of a black hole and a power law with an energy index alpha around 1.2-1.5 extending to at least several hundred keV. In the spectra, we find the presence of strong Compton reflection, which probably comes from the disk.

We present four X-ray/gamma-ray spectra of Cyg X-1 observed in the hard ('low') state simultaneously by Ginga and GRO/OSSE on 1991 July 6. The 3-30 keV Ginga spectra are well represented by power laws with an energy spectral index of alpha~0.6 and a Compton reflection component including a fluorescent Fe K-alpha corresponding to the solid angle of the reflector of ~0.3 times 2 Pi. The overall Ginga/OSSE spectra can be modelled by repeated Compton scattering in a mildly-relativistic, tau ~1, plasma. However, the high-energy cutoff is steeper than that due to single-temperature thermal Comptonisation. It can be described by a superposition of dominant optically-thin, thermal emission at kT~140 keV and a Wien-like component from an optically-thick plasma at kT~50 keV. The X-ray spectra do not show the presence of an anisotropy break required if thermal Compton scattering takes place in a corona above a cold disc. Also, the flat spectral index shows that the plasma is soft-photon starved, i.e., the luminosity in incident soft X-ray seed photons is very much less than that in the hard X-rays. Furthermore, the observed solid angle of the reflector is significantly less than 2 pi. These facts taken together strongly rule out a disc-corona geometry. Rather, the observed spectra are consistent with a geometry in which the cold accretion disc (which both supplies the seed soft X-rays and reflects hard X-rays) only exists at large radii, while the Comptonising hot plasma is located in an inner region with no cold disc. This hot plasma consists of either pure electron-positron pairs if the source size is ~5 Schwarzschild radii or it contains also protons if the size is larger.

We review observations and theoretical models of X-ray/gamma-ray spectra of radio-quiet Seyfert galaxies and of Galactic black-hole candidates (in the hard spectral state). The observed spectra share all their basic components: an underlying power law, a Compton-reflection component with an Fe K-alpha line, low-energy absorption by intervening cold matter, and a high-energy cutoff above ~ 200 keV. The X-ray energy spectral index, alpha, is typically in the range ~ 0.8-1 in Seyfert spectra from Ginga, EXOSAT and OSSE. The hard-state spectra of black-hole candidates Cyg X-1 and GX 339-4 from simultaneous Ginga/OSSE observations have alpha ~ 0.6-0.8. The Compton-reflection component corresponds to cold matter (e.g., inner or outer parts of an accretion disk) covering a solid angle of ~ (0.4-1) times 2 pi as seen from the X-ray source. The spectra are cut off in soft gamma-rays above ~ 200 keV. The broad-band spectra of both Seyferts and black-hole sources are well fitted by Compton upscattering of soft photons in thermal plasmas. Our fits yield the thermal plasma temperature of ~ 100 keV and the Thomson optical depth of tau 1. A fraction of the luminosity emitted nonthermally appears to be small and it can be constrained to < 15% in the Seyfert galaxy NGC 4151. The spectra are cut off before 511 keV, which is strongly suggestive of a thermostatic role of e+e- pair production in constraining the temperature and optical depth of the sources. The source geometry is compatible with a patchy corona above a cold disk in Seyferts, but not in Cyg X-1. In the latter, the relative weakness of reflection is compatible with reflection of emission of a hot inner disk from outer disk regions.

We present the results of 4 simultaneous observations of Cygnus X-1 by Ginga and OSSE. The X-ray/gamma-ray spectra can be described by an intrinsic continuum and a component due to Compton reflection including an iron K-alpha line. The intrinsic spectrum at X-ray energies is a power-law with a photon spectral index of Gamma=1.6. The intrinsic gamma-ray spectrum can be phenomenologically described by either a power-law without cutoff up to 150 keV, and an exponential cutoff above that energy, or by an exponentially cutoff power law and a second hard component.