Classification of the classical male singing voice using long-term average spectrum

Area of Science:

• Vocal pedagogy
• Acoustic analysis
• Music performance

Background:

• Singing-voice classification is traditionally subjective and prone to misclassification, particularly in emerging artists.
• Objective acoustic measurements may offer a more standardized approach to voice classification.
• Long-term average spectrum (LTAS) is an objective measure that has potential for voice classification.

Purpose of the Study:

• To investigate the relationship between Long-term average spectrum (LTAS) and traditional singing-voice classification.
• To determine if LTAS can serve as an objective tool for classifying classical singers' voices.

Main Methods:

• A descriptive, between-subject study design was employed.
• Nine professional male classical singers recorded "The Star-Spangled Banner" in their preferred key.
• Long-term average spectrum (LTAS) was calculated for different song segments (phrases, remainder, entire song) and compared with singing-voice classification and vocal range.

Main Results:

• A moderate correlation was found between singing-voice classification and overall LTAS average, with the strongest correlation observed for the entire song sample.
• Singing-voice classification demonstrated a strong correlation with the singers' vocal range.
• The correlation strength varied depending on the LTAS sample length analyzed.

Conclusions:

• Long-term average spectrum (LTAS) shows potential as an objective aid for singing-voice classification.
• Further research is needed to clarify the optimal application of LTAS in voice classification, considering factors like sample length and content.
• The influence of phonetic and pitch content on LTAS requires further investigation for accurate classification.