Muriqui has a set of 14 discrete elements identifiable in their vocalizations [1] divided in 4 short elements and 10 long elements. Vocalizations composed only of short elements (staccatos) are preferentially used for short distance communication and those containing long elements (neighs), for long distance communication. The paper demonstrates that Muriqui vocalizations can be defined by 3 source features: whistled [W], noisy [N] and trilled [T]. In addition [W] & [N] can be combined and superimposed which is not the case for [Tr] that only appears alone. [W] can be realized with the secondary features high [H] or low [L] and/or short [S] or long [Lg]. The feature [H] is defined by a fundamental frequency above 2000 Hz for the lowest spectrum component identified on a (FFT) spectrum. The [L] feature is defined by a fundamental frequency below 2000 Hz for the lowest spectrum component identified on a (FFT) spectrum. Duration features [Lg] and [S] are defined as durations above and below 200 ms. These secondary features are estimated values obtained from a representative database of muriqui calls [1]. [N] can be realized with the secondary features strong [S] or weak [We] and/or short [S] or long [Lg]. The [S] feature is defined by a noise component above 2000 Hz on an FFT spectrum. The [We] feature is defined by a noise component below 2000 Hz on a FFT spectrum. Duration features [Lg ] and [S] are defined as durations above and below 200 ms. These features values are rounded values obtained from a representative database of muriqui calls [1]. [Tr] is realized with short transients (bursts) < 5/10 ms that are repeated several times. Muriquis use these features to form the 14 discrete elements, and combine elements into strings to form acoustically variable vocalizations.