Local Field Potentials (LFP) and Multiple Unit Activity (MUA) are indicators of perisynaptic and spiking activity respectively. An analysis of changes in dependence between these two types of signals may help us understand computations at the level of small networks and characterize the type of processing carried out by a cortical area. The goal of our research is to develop a principled and theoretically sound approach to measuring the information content of electrical signals in the brain (LFPs and spikes), as it relates to stimulus. Traditionally, the LFP has been broken up into a series of well established bands from the EEG literature (delta, alpha, beta), however it is not clear that this decomposition is maximally informative in relating the stimulus and response. To better understand this issue we analyze multiple electrode recordings from the macaque primary visual cortex (V1), under stimulation by continuous natural movies. We first decompose the information about the stimulus, contained in the LFPs and spikes, into different frequency bands, so as to determine which parts of the signal carry the most stimulus content. We also determine the interaction between different LFP and spike frequencies, and test the extent to which these interactions are stimulus driven. In this way, we are able to determine whether bands are synergistic (contain more information jointly than if viewed separately), independent, or redundant. This analysis leads to a frequency decomposition methodology for the LFP that follows the signal structure, rather than prior beliefs; clarifies the dependence between different LFP frequencies and spikes; and gives insight into the neurophysiological mechanisms for stimulus encoding. We further measure the effect of temporal resolution on these information theoretic quantities, to determine the timescale over which information transfer occurs. It is also possible to extract component features from the input stimulus, using image processing methods: examples include flow fields, frame-to-frame per-pixel differences, contrast, luminance, and low and high frequency power in each frame. We investigate the extent to which neural activity is driven by particular stimulus features, and find the fraction of total information transmission associated with each of these. Finally, we investigate the information between neural signals as a function of separation between recording sites. In particular, it can be determined, as a function of signal frequency, the distance at which no significant dependence of any kind exists between signals; this may then be compared to known anatomical features (e.g column width) of the cortex.