Recitation of the Holy Quran with Tajweed is an essential activity as a Muslim. Reciting Quran correctly indicates the correct meaning of the words of Allah has been received from this significant resource among Muslims. That is why Muslims stress on the Quranic Education since in the early age. It is important to pronounce the letter correctly based on its characteristics as well as the articulation point of each letter. In this paper, the characteristic based on Qalqalah letters is considered to be analyzed. The audio signal from a person who is very good at Quranic recitation was taken and analyzed. We implement spectral analysis to find the features of Qalqalah letters and extract the correlation between the first formant frequency and the pharyngeal space of the signal. Spectrogram was successfully implemented and proved this relation, and it described the mechanism of Qalqalah correctly, which is unique as compared to other Quranic letters.
The word TAJWEED means "to improve" or "to make better". It is also the rules
and knowledge that help people to recite the Holy Quran as similar as what has been
recited by the Prophet Mohammed peace and blessings be upon him. Moreover, an
important part of Tajweed is to pronounce the letters from its correct articulations
(Makharij) and by giving the letters its inherent characteristics (Sifaat) and dues in
conditional characteristics. The characteristics of Arabic letters helps in differentiate
letters that have the same points of articulations. The Sifaat of the Arabic letters are
divided into two groups, characteristics with opposites and characteristics without
opposites, [
Qalqalah is one of the most known but very difficult characteristics to be
pronounced. It means bouncing or echoing the sound. The letters with this characteristic
show an echoing sound when the phone carries sukoon (°). Qalqalah letters come
from the strength category of Quranic letters. Strength means the points of articulation
of this group of letters is completely closed when the pronunciations come with
Sukoon. The consonants letters pronunciation happens by the collusion between the
parts of the articulation of these letters. While the pronunciation of vowel letters occur
by the parting between the two parts of the articulation points, [
The research on Tajweed has been conducted previously by some researchers, but
still can be considered within a small associated society. The Multilayer Perceptron
(MLP) has been used to investigate the process of detecting the right recitation of
Qalqalah Kubra of a reader. Mel-frequency Cepstrum Coefficients (MFCC) has been
implemented as a feature extraction to get the essential characteristics of the
pronunciation signals. Then the MLP was trained to distinguish between the correct and the
incorrect pronunciation. It is clear from the results that the classification accuracy was
high, and the MLP was claimed to be successful to differentiate between the correct
and the incorrect pronunciation, [
On the other hand, Spectrogram has been implemented as well in the research involving the Arabic letters pronunciation. The authors in for example have investigated the impacts of implementing vowels (fatha, kasra, and dhamma) on the dental consonants. Dental consonants are the letters that the teeth are part of its points of articulations. It is divided into three groups as follows; • Labiodental, which can occur on the lower lip and the upper teeth. • Dental, which can occur on the tongue against the upper teeth. • Interdental, which can occur at the tip of the tongue between the upper and the lower front teeth.
The experiment has been conducted among Malay children (non-native Arabic speakers). Spectrogram and formants have been used to find the influence of the vowels on the dental consonants. They have found that the effect of Kasrah and dhamma can be easily extracted from the spectrogram.
The paper is organized as follows; Section II describes the process of sound production, formant frequencies and its relation to the path of the air inside the pharynx and the spectrogram analysis. Section III is described the experiment and the spectrogram results. Section IV explained the conclusion and the recommended future works. 2
Phones are produced by three steps. First, the process of providing the energy for the
sound from the lungs, followed by converting this energy into a sound by the
vibration of the vocal folds. The last step is to translate this sound from the vocal fold into
an understandable speech. Fig. 1 shows the diagram of the speech system from the
energy generation until the filtering of the speech in the vocal tract, where the final
stage of the system is the vocal tract. The vocal tract works as a filter to shape the
signal from the vocal folds which contains many frequencies into a new signal with
formant frequencies and to be sensible to human. When the air flows from the vocal
fold through the oral cavity, the tongue, lips, teeth, and various regions of the mouth
configuration and shape help in filtering the sound into its formant frequencies, [
The spectrogram is a visual illustration to represent the frequency components of
the speech signal. It is considered one of the applications of the Fourier transform.
Fig.2. Shows the steps to calculate the spectrogram graphs [
and Window
FFT 20log10 (abs(.)) stack vector colour map spectrogram In order to calculate the spectrogram of Qalqalah letters, the first step was collecting the pronunciation data in terms of audio signals. These data has been recorded from a person who is a specialist in recitation Al-Quran with Tajweed, where the recording process was conducted inside a studio room to reduce the environmental noises. The five letters of Qalqalah which they are gathered in one Arabic sentence “ق - ط- ب– ج د” have been recorded, and the silence segments have been removed by using Matlab software. The data was then normalized such that the amplitude varies from 1 to -1. The short-time Fourier transform (STFT) has been implemented to calculate the spectrogram of the audio signals. Table 1 summarizes the configurations of the STFT that was used in this paper. • The voiced sound which is clear and in a dark color and it has formants. • The silent segments followed the main sound which represents the period when the articulation is completely closed and the air is imprisoned inside it. • The last period is representing the burst pattern that should happen to make the correct pronunciation of Qalqalah.
Formant frequencies are one of the key features that can be extracted from the
spectrograms. The first formant frequency is related to the tongue position, and the tongue
position will affect the pharyngeal space. The more pharyngeal space leads to less
first formant value. Moreover, the degree of mouth opening is related to the first
formant frequency, which more opening leads to a higher formant frequency and
viceversa [
The spectrogram has been successfully implemented to investigate the characteristics of Qalqalah sounds. This technique was an initial attempt in characterizing one of the important sifaat of Quranic letters in Tajweed. The results obtained showed the relation between the pharyngeal space and the first formant frequency, which are inversely proportional. Moreover, this technique was able to illustrate the mechanism of the pronunciation of the Qalqalah sounds, where the voiced segment of the sound, silent segment and the burst segment at the end of Qalqalah phones were shown. These features will be used to classify and distinguish the correct recitation of Qalqalah and other sifaat so that the correct pronunciation of Quranic letters will benefit the reciters more.