The owl’s auditory system offers a good substrate to address the question of how plausible neural models can be. To determine what and where acoustic objects are, the brain must compare and combine noisy information coming from each ear. Comparison, combination selectivity, and noise reduction can be represented by equations. In the owl’s auditory pathway, we can find neurons that effectively perform these types of processing. In the Nucleus Laminaris neurons perform a running cross-correlation that indicates the degree of similarity between right- and left-side inputs. Neurons of the auditory midbrain become selective to combinations of spatial cues by a multiplicative-like process on convergent inputs coming from independent processing pathways. In the inferior colliculus, an average-like process produces a tuning to interaural time difference that would require at least five repetitions of the same stimulus one stage before. Thus, the owl’s auditory system appears to perform the computations that models predict, representing an interesting example of how close to the nature of neural processing mathematical operations are.