Raga identification in Carnatic music is challenging due to its complex tonal structures, ornamentations (gamakas), and pitch variations. Unlike Western music, Carnatic music relies on melodic nuances and microtonal shifts, making conventional Music Information Retrieval (MIR) techniques insufficient (Natesan and Beigi, Carnatic Raga identification system using rigorous time-delay neural network, 2024). Current MIR algorithms, designed mainly for Western music, struggle to capture these intricacies, leading to inaccurate raga recognition. This study addresses this gap by comparing two advanced approaches: a tonal frequency-mapping algorithm and deep learning-based methods. The first approach isolates specific swara frequencies and maps them to raga structures, offering a rule-based solution tailored to Carnatic music. The second utilizes long short-term memory (LSTM) and time delay neural (TDNN) networks, which adapt to the dynamic patterns of Carnatic music, enabling nuanced recognition. By evaluating these methods, this research enhances automatic raga recognition and demonstrates how technology can support the preservation and appreciation of Carnatic music’s cultural heritage.