English language learners have difficulties in English in pronouncing fricative consonants, voiced and voiceless TH, plosive consonants and vowel sounds. Serious speech mistakes can lead to miscommunication since poor pronunciation abilities make it difficult for other people to hear what you are saying. Additionally, students may miss their lecturers' direct explanations of proper pronunciation when taking online English lessons with inconsistent Internet access. Proper pronunciation can help students participate in conversations, form connections, and improve graduates' employability abilities. Furthermore, pronunciation skills are learnt independently at higher education institutions. Before designing a web-based tool for self-directed learning of English pronunciation, the designers must get feedback from the students about their self-directed learning, self-evaluation of English language proficiency and English learning experiences through Online Distance Learning (ODL). This survey is a preliminary survey towards the development of a heutagogical English pronunciation application. An online questionnaire was disseminated to 424 respondents from various higher education institutions in Malaysia.  In this survey, firstly we aimed to get feedback from the students about their self-directed learning, self-evaluation of English language proficiency and English learning experiences through Online Distance Learning (ODL). The survey consisted of demographic data of respondents, standardized test of English language proficiency, self-evaluation of English language proficiency and English language learning experiences during COVID-19. The findings suggested that although language learners have developed their self-directed learning skills, they still need to improve these skills in learning the language using ODL. Learners consider themselves as having average English language proficiency. They were responsible for improving their pronunciation skills and needed assistant tools to improve their pronunciation skills. It is recommended that an application that uses a heutagogical approach such as Speech-to-Text and Text-to Speech need to be developed.