Typographic, layout and logical elements constitute visual text-signals of a document that carry semantic information over and above its content. Although they are important to the reader, most of the current Text-to-Speech (TtS) systems do not support them. As there is a lack of studies on how blind perceive them and aiming to incorporate them efficiently in advanced TtS systems, we investigate in a systematic way the perception of the main typographic text-signals by 73 blind and sighted students. The results show that both groups of the participants perceive that font-styles are used largely to better locate, recognize or distinguish the topics or specific information in a document. Almost half of the sighted argue that they are useful for the comprehension of the content, but only 4 % of the blind students perceive the same. Most of the sighted participants (68 %) consider that bold is used to indicate an important word or phrase in the text that needs more attention by the reader, but only 23 % of them perceive the same for the italics. 27 % of the blind participants and 23 % of the sighted perceive that the role of font-size is to provide emphasis. Moreover, only 9 % of the sighted students grasp that bold is used for emphasis and 13 % of them that italics is used for light emphasis. Half of the blind participants consider that font-size plays an important role in separating the basic elements of a text (e.g. titles, footnotes), but only 13 % of the sighted believe the same. Finally, the sighted and blind students recognize the titles of a text mainly using non-identical criteria.