Traditionally, translation and interpreting are differentiated in terms of speaking and writing (PAGURA, 2015; MUNDAY, 2016), which excludes video translations into Brazilian Sign Language (Libras). This research aims to elucidate the differences between translation and interpretating, without relying solely on the record of the final product or the source text. Considering discussions about support (RODRIGUES, 2018; SOUZA, 2021), the concept of writing and speaking (SOUZA, 2021), Kade s definitions (apud POCHHACKER, 2016, 2018) for translation and interpretation, and researches that focus on translations into Libras on video (SILVÉRIO et. al., 2012; CARNEIRO, VITAL and SOUZA, 2020; PYFERS, 1999; GALASSO et. al., 2018), the following defining criteria for translation processes (translation or interpretating) were selected, which include both sign languages and oral languages: time for textual planning, presentation of the source text, work rhythm, external support, work method, refinement of the target text, necessary skills and abilities, guide material used, unintentional disfluencies, multimodal resources as translation solutions and cognitive efforts triggered at the time of filming. Bibliometric analysis was carried out on different databases, as well as the analysis of 7 videos, available on the ViaLibras website. The guide material used and the unintentional disfluencies point to different cognitive efforts that are activated simultaneously in typical translation processes. So that is a fundamental criterion in the differentiation of processes. A guide material that recovers all the translation decisions made before filming was of particular importance in defining typical sign language translation processes. It is hoped that the work can contribute to the training of Libras translators and interpreters.