The deconvolution, or cleaning, of radio interferometric images often involves computing model visibilities from a list of clean components, in order that the contribution from the model can be subtracted from the observed visibilities. This step is normally performed using a forward fast Fourier transform (FFT), followed by a 'degridding' step that interpolates over the uv plane to construct the model visibilities. An alternative approach is to calculate the model visibilities directly by summing over all the members of the clean component list, which is a more accurate method that can also be much slower. However, if the clean components are used to construct a model image on the surface of the celestial sphere then the model visibilities can be generated directly from the wavelet coefficients, and the sparsity of the model means that most of these coefficients are zero, and can be ignored. We have constructed a prototype imager that uses a spherical-wavelet representation of the model image to generate model visibilities during each major cycle, and find empirically that the execution time scales with the wavelet resolution level, J, as O(1.07 J), and with the number of distinct clean components, N_C, as O(N_C). The prototype organises the wavelet coefficients into a tree structure, and does not store or process the zero wavelet coefficients.

Wide-field imaging has become a major challenge for modern radio astronomy, which uses high sensitivity acquisition systems that deal with huge amounts of data. In this paper we investigate a fast wide-field imaging solution based on the w-projection algorithm, which is intended for modern astronomy systems. The core idea of the proposed method is to reduce the computational complexity of the convolution kernel generation step, specifically by replacing the standard two-dimensional FFT by the one-dimensional Hankel transform. Experimental results show that the optimised w-projection proposed here produces equivalent dirty image results in a circular image region, at a significantly lower computational cost than standard $w$-projection. One of the main advantages of the proposed solution is its slow scaling with the number of w-planes, thus enabling more accurate output results at a lower computational cost.

Prompted by a recent claim by Barrows et al. that X-ray AGN are often found significantly offset from the centres of their host galaxies, we have looked for examples of compact radio sources which are offset from the optical centroids of nearby (z < 0.2) galaxies. We have selected a sample of 345 galaxies from the Sloan Digital Sky Survey (SDSS) galaxy catalog which have nearby compact radio sources listed in the Cosmic-Lens All Sky Survey (CLASS) catalog. We find only three matches (0.87 per cent of the sample) with offsets greater than 600 milliarcsec (mas), which is considerably fewer than we would have expected from the Barrows et al. X-ray survey. We fit our histogram of offsets with a Rayleigh distribution with {\sigma} = 60.5 mas, but find that there is an excess of objects with separations greater than approximately 150 mas. Assuming that this excess represents AGN with real offsets, we place an upper limit of approximately 17 per cent on the fraction of offset AGN in our radio-selected sample. We select 38 objects with offsets greater than 150 mas, and find they have some diverse properties: some are well known, such as Mrk 273 and Arp 220, and others have dust lanes which may have affected the optical astrometry, while a few are strong new candidates for offset AGN.

We present X-ray spectral variability of 24 local active galactic nuclei (AGN) from the Palomar sample of nearby galaxies, as observed mainly by Swift. From hardness ratio measurements, we find that 18 AGN with low accretion rates show hardening with increasing count rate, converse to the softer-when-brighter behaviour normally observed in AGN with higher accretion rates. Two AGN show softening with increasing count rate, two show more complex behaviour, and two do not show any simple relationship. Sufficient data were available for the spectra of 13 AGN to be summed in flux-bins. In 9 of these sources, correlated luminosity-dependent changes in the photon index ($\Gamma$) of a power-law component are found to be the main cause of hardness variability. For 6 objects, with a low accretion rate as a fraction of the Eddington rate (\.m$\mathrm{_{Edd}}$), $\Gamma$ is anticorrelated with \.m$\mathrm{_{Edd}}$, i.e. `harder-when-brighter' behaviour is observed. The 3 higher-\.m$\mathrm{_{Edd}}$-rate objects show a positive correlation between $\Gamma$ and \.m$\mathrm{_{Edd}}$. This transition from harder-when-brighter at low \.m$\mathrm{_{Edd}}$s to softer-when-brighter at high \.m$\mathrm{_{Edd}}$s can be explained by a change in the dominant source of seed-photons for X-ray emission from cyclo-synchrotron emission from the Comptonising corona itself to thermal seed-photons from the accretion disc. This transition is also seen in the `hard state' of black hole X-ray binaries (BHXRBs). The results support the idea that LINERs are analogues of BHXRBs in the hard state and that Seyferts are analogues of BHXRBs in either the high-accretion-rate end of the hard state or in the hard-intermediate state.

Below an accretion rate of approximately a few per cent of the Eddington accretion rate, X-ray binary systems are not usually found in the soft spectral state. However, at accretion rates a factor of a few lower still, in the hard state, there is another spectral transition which is well observed but not well understood. Below ~0.5-1 per cent of the Eddington accretion rate (m_crit), the spectral index hardens with increasing accretion rate, but above m_crit, although still in the hard state, the spectral index softens with increasing accretion rate. Here we use a combination of X-ray spectral fitting and a study of short time-scale spectral variability to examine the behaviour of three well-known X-ray binaries: Cygnus X-1, GX 339-4 and XTE J1118+480. In Cygnus X-1 we find separate hard and soft continuum components, and show using root-mean-square (rms) spectra that the soft component dominates the variability. The spectral transition at m_crit is clearly present in the hard-state hardness-intensity diagrams of Cygnus X-1. Above m_crit, GX 339-4 shows similar softer-when-brighter behaviour at both long and short time-scales. Similarly, XTE J1118+480, which remains well below m_crit, is harder-when-brighter behaviour on all time-scales. We interpret these results in terms of two continuum components: a hard power-law which dominates the spectra when the accretion rate is low, probably arising from Comptonisation of cyclo-synchrotron photons from the corona, and a soft power-law which dominates at higher accretion rates, arising from Comptonisation of seed photons from the accretion disc.

The way in which the X-ray photon index, {\Gamma}, varies as a function of count rate is a strong diagnostic of the emission processes and emission geometry around accreting compact objects. Here we present the results from a study using a new, and simple, method designed to improve sensitivity to the measurement of the variability of {\Gamma} on very short time-scales. We have measured {\Gamma} in ~2 million spectra, extracted from observations with a variety of different accretion rates and spectral states, on time-scales as short as 16 ms for the high mass X-ray binary Cygnus X-1, and have cross-correlated these measurements with the source count rate. In the soft-state cross-correlation functions (CCFs) we find a positive peak at zero lag, stronger and narrower in the softer observations. Assuming that the X-rays are produced by Compton scattering of soft seed photons by high energy electrons in a corona, these results are consistent with Compton cooling of the corona by seed photons from the inner edge of the accretion disc, the truncation radius of which increases with increasing hardness ratio. The CCFs produced from the hard-state observations, however, show an anti-correlation which is most easily explained by variation in the energy of the electrons in the corona rather than in variation of the seed photon flux. The hard-state CCFs can be decomposed into a narrow anti-correlation at zero lag, which we tentatively associate with the effects of self-Comptonisation of cyclo-synchrotron seed photons in either a hot, optically thin accretion flow or the base of the jet, and a second, asymmetric component which we suggest is produced as a consequence of a lag between the soft and hard X-ray emission. The lag may be caused by a radial temperature/energy gradient in the Comptonising electrons combined with the inward propagation of accretion rate perturbations.