FILE CONVERSION AFTERMATH: ANALYSIS OF AUDIO FILE STRUCTURE FORMAT Abstract JENNIFER L. SANTOS 1 JASMIN D. NIGUIDULA Technological innovation has brought a massive leap in data processing. As information turns out to be broadly accessible, various tools have been produced to create more types of format on existing data. This format is then manipulated to fit in different processing structure to generate needed results. In information technology, audio is one of the most flexible types of data that could be manipulated into different forms. In this light, this paper evaluates the conversion of multiple audio files in waveform audio file (.wav) formats to.mp3, wma, and. Aac formats using standard parameters. The evaluation has demonstrated noteworthy changes as compared to the original files and as supported by interesting fact, this study further explained the structures behind the converted output. Keywords: Audio Compression, Audio Format, Audio Analysis 1. INTRODUCTION The sounds that human hear from nature are analog, and these are processed by our auditory senses in their original format, but if theses sounds would be stored, they would have to be converted in digital form. Audio file is created through the PC either from a recording or produced naturally by a synthesizer. There are many applications available nowadays where people can just select, listen, and sometimes download songs they love. Music lovers can download the audio files in various formats. Even journal articles can be converted to audio files particularly in mp3 format so people with low-vision or no vision at all can have access to the journals. Downloading data from the internet may affect the size and quality of the data. For easy uploading and downloading in the internet, data compression is used. Data compression is a process where allowable number of bits is reduced for easy storage and transmission of data. Ordinary music listeners may not be able to distinguish compressed music from uncompressed music most especially if these are digital audio files. Audio files in compact discs (CD s) are uncompressed, and needs 1 Technological Institute of the Philippines (T.I.P.)- Manila Email: 1 jlsantos@ceu.edu.ph, 2 jasniguidula@yahoo.com 51
to be converted to compressed formats to be read on a computer or portable device. The quality of the audio file is dependent on the bit rate as better quality of the audio is achieved when it is higher. This paper aims to analyze the factors that affect the audio quality upon audio conversion. Comparison of original data and converted data will be presented and evaluated in terms of audio file size, audio duration, audio stream size, audio sampling rate, and audio overall bit rate. 2. RELATED WORKS There are various distinctive sorts of Audio records. The most well-known are Wave documents (wav) and MPEG Layer-3 records (mp3). The way the sound is packed and put away is called the codec which decides how little the document size is. Some document sorts dependably utilize a specific codec. Waveform Audio (wav) File Format is a standard sound record arrange utilized for the most part in Windows PCs. file MPEG-1 Audio Layer-3 (mp3) is a standard format for downloading and storing around one-twelfth of the size of the original audio files while maintaining the quality of the sound. Free Lossless Audio Codec (flac) on the other hand is a lossless compression codec, compression and decompression of this file type does not affect the original file. Windows Media Audio (wma) is an audio format developed by Microsoft, and it can encode digital audio like that of mp3 but with a higher rate. Advanced Audio Coding format (aac) is used in compressing lossy digital audio files. AAC and MP3 has similar rates but AAC produces better quality of sound. [6] Data compression may be classified as lossless or lossy. In lossy compression, the size of the compressed file is very small as compared to the original data because some signals were removed. [7]Compression of audio, video, and image use lossy methods. [8] Additionally, data compression is a process that reduces specific files. It discards redundancy and inaudible data to lessen the file size and transmission time of the data while maintaining the supposed audio quality. The term audio compression is used when audio signals are used as data. Original signal is filtered while removing unwanted signals so that only true digital signal is produced. MP3 uses variable-length Huffman codes and tis results to a more effective data compression. The sampling rate can have values of be 32, 44.1, or 48 khz. It is very well appropriate for audio transmission over the internet. Different algorithm is used in converting audio data such as Run-length encoding (RLE), Burrows-wheeler transform (BWT), Move to front transform (MTF), and Arithmetic coding (ARI). Other techniques are Shannon- Fano, Huffman Coding. The result of data compression is dependent on the data source while the characteristic of the output relies whether the converted format is lossy or lossless. 52
3. METHODOLOGY a. Audio Conversion and Audio Analyzer Tool The Waveform Audio File (.wav) format audio files used were extracted using the online website www.onlinevideoconverter.com. The videos were randomly selected from the nursery rhymes in YouTube.com. The URL of the videos were copied and pasted into the online converter to convert the videos into audio files. This research utilizes an audio conversion tool known as format factory. The tools feature includes conversion of video, audio and picture files. Specifically, it converts audio data to MPEG-1 Audio Layer-3 (.mp3), Windows Media Audio (.wma), and Advanced Audio Coding (.aac) format. As the research assess multiple audio files converted to four audio formats, another tool is then used for the analysis of the results. The data results of the conversion are then evaluated using an open source application. The Media Info is a tool that accesses to technical information such that of reviewing different audio format, customization of format, exporting of information, providing a graphical user interface, and integration of data in a shell. b. Formulation Experimental method was used in this research as it manipulates multiple data, and controls the rest of the data. In this case, three nursery songs (audio files) are used entitled Baa Baa Black Sheep, Head Shoulders Knees Toes, and Hey Diddle Diddle. The three original data used are in Waveform Audio File (.wav) format and are subject for conversion using Format Factory into MPEG-1 Audio Layer-3 (.mp3), Windows Media Audio (.wma), Free Lossless Audio Codec (.flac), and Advanced Audio Coding (.aac) formats. After the conversion, the results are then analyzed in the MediaInfo application utilizing the accompanying parameters: Table 1. Parameters for Identifying Audio Conversion Findings Parameters for Audio Conversion Analysis Audio File Size Audio Duration Overall bit rate Sampling rate Stream Size 53
Waveform Audio File (.wav) format have uncompressed audio in Pulse-Code Modulation (PCM) format. In, PCM the signals can only have two values, 1 or 0. There are three steps involved in pulse code modulation as displayed in figure 1. The first step involves conversion of continuous amplitude signal into discrete-timecontinuous signal. Followed by quantization where the excessive and redundant bits are reduced and compressed. And finally the analog signals are digitized in encoding, and therefore the bandwidth used by the signal is reduced. Sampling Quantizing Encoding Figure 1. Block diagram of Pulse Code Modulation Modified Discrete Cosine Transform (MDCT) is used in most lossy formats such as MP3, AAC, and WMA. MDCT is a Fourier related change in light of type-iv DCT and has an extra property of being "lapped, making it very useful in quantization. The following equation is used in MDCT. 4. RESULTS AND DISCUSSION This part of the study shows the extraction results based on the different parameters from the audio conversion tool. The audio file size, and the audio stream size are both measured in megabytes (MB), the duration of the audio files are presented in seconds (sec), overall bit rate of the audio files is presented in bits per second (kbps), and the audio sampling rate is measured in kilohertz (khz). Figure 2. Result of Audio Conversion of Baa Baa Black Sheep using Format factory as analyzed by Media Info 54
Since the audio file has been converted to formats wherein the file size has been reduced, it can be said that the audio file had undergone compression. Figure 2 shows the data extracted from the Media Info in analyzing Baa Baa Black Sheep. It can be noted that the.wav format has the biggest file size and stream size, and the.aac format has the smallest file size and stream size. Duration and sampling rate did not change after the conversions of.wav file to other audio formats used in the study. Figure 3. Result of Audio Conversion of Head Shoulder Knees and Toe using Format factory as analyzed by Media Info The figure above presents the analysis of Head Shoulder Knees and Toes. All the audio formats had 130 seconds duration and 44.1 khz sampling rate. The file size and stream size were greatly reduced from.wav to.aac format, with original value of 21.9 MB and new value of 4.93MB upon conversion. Figure 4. Result of Audio Conversion of Hey Diddle Diddle using Format factory as analyzed by MediaInfo 55
Figure 4 displays data for Hey Diddle Diddle. Like the first two figures, it is noticeable that the.wav format has the highest values in all parameters used in the study, while the lowest values are associated with the.aac format. The overall bit rate of the original files has a very large difference with all of the converted files. The audio overall bit rate have dropped by 77.32% when converted from Waveform Audio File (.wav) format into Advanced Audio Coding (.aac) and 60.95% when converted from Waveform Audio File (.wav) format into MPEG-1 Audio Layer-3 (.mp3), Furthermore, all the figures show that the largest drop of file size is the conversion of the audio file from WAV format to MP3 and AAC format, and this is also true with regards to the audio stream size. The audio files and the stream size have reduced by almost 78.05% when converted from Waveform Audio File (.wav) format into Advanced Audio Coding (.aac) and 77.33% when converted from Waveform Audio File (.wav) formatinto MPEG-1 Audio Layer-3 (.mp3). 5. CONCLUSION From the given results and analysis on this study, it can be perceived that among the audio file format evaluated, Waveform Audio File (.wav) format has the biggest file size, and stream size being the original file contrary to the result of Advanced Audio Coding (.aac) format. Furthermore, the audio duration and the audio sampling rate were not affected by the conversion of audio files from Waveform Audio File (.wav) format to.mp3,.wma, and.aac formats. The audio overall bit rate changed upon the conversion applied to the audio files resulting to a 77.32% reduction. The highest value of overall bit rate is noted in the Waveform Audio File (.wav) format and the lowest value is on the MPEG-1 Audio Layer-3 (.mp3) format. Overall bit rate of Advanced Audio Coding (.aac) format is variable and so it was not measured exactly by Media Info. REFERENCES A. Kaur, N. S. Sethi at H. Singh, A Review on Data Compression Techniques, International Journal of Advanced Research in Computer Science and Software Engineering, pp. 769-773, 2015. N. Amal, Light and easy with multimedia components, Sunday Mail, p. 23, 29 August 1999. C. Breen at J. Seff, Rip, Store, Organize, Macworld, pp. 60-63, 2004. 20 October 2011. [Online]. Available: https://librarynews.marygrove.edu/2011/10/20/did-you-know-you-canconvert-journal-articles-to-audio-files/. [Na-access 14 March 2017]. K. Mc Elhearn, Everything you need to knopw about digital audio files, Macworld, pp. 128-129, 2016. 56
http://www.nch.com.au/acm/formats.html, [Online]. Available: http://www.nch.com.au/acm/formats.html. [Naaccess 14 March 2017]. S. R.Mahalakshmi, Data Compression in Multimedia (Text,Image,Audio and Video), nternational Journal of Engineering Sciences & Research Technology, pp. 376-382, October 2014. A. J. Maan, Analysis and Comparison of Algorithms for Lossless Data Compression, International Journal of Information and Computation Technology, vol. III, blg. 3, pp. 139-146, 2013. K. Dangarwala at J. Shah, C Implementation & comparison of companding & silence audio compression techniques, International Journal of Computer Science Issues, vol. 7, blg. 2, pp. 26-30, 2010. D. Sinha at A. H. Tewfik, Low Bit Rate Transparent Audio Compression using Adapted Wavelets, IEEE TRANSACTIONS ON SIGNAL PROCESSING, vol. 41, blg. 12, pp. 3463-3479, 1993. 57