Analysis of the Steelpan Notes using FFT Algorithm
Davinder Pal Sharma1, Arvindra Sampath2 and Deepchand Gangasingh3
1,2,3 Department of Physics, University of the West Indies, St. Augustine, Trinidad & Tobago
Email: 1Davinder.Sharma@sta.uwi.edu, 2arvindrasampath@yahoo.com,
3deepchandgangasingh@hotmail.com
Abstract— Steelpan is a percussion musical instrument which evolved in Trinidad
& Tobago during the 1930s. It is the national instrument of Trinidad & Tobago
and is very modern instrument as compared to piano, violin etc., which have been
around for many centuries. For better understanding and in-depth analysis of the
notes generated by the Steelpan, particularly leading type of Steelpan called
Tenor-pan, a very popular digital signal processing (DSP) tool known as Fast
Fourier Transform (FFT) has been used in this study. Recording of Tenor-pan
notes was done in a noise free environment and they were processed through FFT
using Radix-2, Decimation-in-Time (DIT) algorithm. Processing has been done
using software (Matlab) as well as hardware (TMS320VC5416 digital signal
processor). FFT simulation of the Steelpan audio files has been done using
Simulink and DSP System toolboxes of Matlab. Code Composer Studio (CCS) has
also been used to simulate and implement the FFT algorithm on DSP chip. The
results from both simulation and implementation have been presented here.
Present study will assist in developing applications / software for steel-pan trainers
throughout Trinidad and Tobago and furthermore the world.
I. INTRODUCTION
The world is now rapidly moving from analog to digital counterpart due to its lower cost, higher speed, lownoise, less power consumption, higher flexibility and reliability. Digital signal processing is the technology that is the key to the growth of digital world. In the earliest days its applications were limited to sonar and seismology. As digital processing circuits and analog to digital converters (ADC) have become faster, the number of applications for DSP has exploded. Due to Digital signal processor gaining many advances in both research and many applications, over the past five decades the field of DSP has seen tremendous growth.Digital signal processor is now used in many areas of technology, from mobile phones to modems, home appliances, medical equipment, image and audio processing.
Today musical instruments makes up a US $17billion business since musicians and producers are turning to electronics to increase their skills and experiment with new sounds .This large increase in the electronics has seen the digital signal processing market growing at a faster rate than the rest of the semiconductor industry.
Present study has been carried out to assist in developing steel-pan based software / applications and help trainers in the training process. In-depth analysis of each Steelpan note is essential for such tasks so we have analyzed each steel-pan notes using popular and powerful DSP technique known as FFT. We have performed simulations for each notes and verified the results with digital signal processor based implementation.
The cleanest samples were imported into Simulink to obtain the FFT of samples. These samples were also used in Matlab to obtain the data values for implementation. The data values were then sent to Code Composer Studio (CCS) and the DSP chip after conversion to Q15 format using Q15 converter.
The Tenor-pan, a special type of Steel-pan, was used in the present study. It has three layers which include an inner (_in), middle (_mid) and outer (_out) layer consisting of twenty-nine different musical notes. Twenty samples of each note were played on the Steel-pan, which were recorded by a microphone connected to a laptop. Examples of these notes are: D_in, D_mid and D_out, which represents the inner, middle and outer D notes of the Steel-pan respectively. Microsoft Sound Recorder was used to record and store the media files at a sampling frequency of 44.1 kHz.
A sound, editing software called Audacity was utilized to cut out or obtain the cleanest, single sample, from the twenty samples for each of the Steel-pan notes. Audacity was then used to remove the background noise for each of the twenty-nine extracted sample notes. The purpose of this noise reduction was to avoid the hissing sound that is heard when the microphone is ON and no samples were being recorded. Also, noise reduction was useful in avoiding the echo of previous notes to be synchronized with the extracted sample.
These extracted media files were then reduced to a time period of 0.1 second, to ensure that only the main sound of each note can be simulated and not any unnecessary echo of the note. These newly saved samples of 0.1 second were used in both simulation and implementation processes.
.
Replies
Claude hope this info helps...
Sound radiation from Caribbean steelpans using nearfield acoustical holography
Brian Russel Copeland
The sound-field of a Caribbean low tenor steelpan is examined using nearfield acoustical holography (NAH). The particular instrument, a Bertrand Kelman low tenor, has been previously examined in a single plane using a four microphone sound intensity technique. This paper will examine the sound radiation in a cuboid space enclosing the instrument at three frequencies. Presented will be the results of sound pressure level (SPL), active acoustic intensity (AI), and reactive acoustic intensity (RI) obtained through application of the NAH technique. The limitations of NAH when applied to an irregularly shaped object such as a tenor steelpan will also be presented.
https://www.researchgate.net/publication/221849542_Sound_radiation_...
He currently co-ordinates the Steelpan Initiatives Project (SIP), which saw the development and patenting of the G-Pan, a re-engineered form of the traditional steelpan, and the Percussive Harmonic instrument (P.H.I.), an electronic form of the traditional steelpan. Professor Copeland is also the Convener of the Steelpan Research Centre, UWI, and is a former Member of the Board of Directors of CARIRI.
Check out his profile…https://whensteeltalks.ning.com/forum/topics/what-pan-waiting-for-n...
"Today musical instruments makes up a US $17billion business since musicians and producers are turning to electronics to increase their skills and experiment with new sounds"
Andy Narell is supposed to be coming out with his SAMPLES pretty soon!!!
But I still see the need find a good sound engineer to HYBRIDIZE THE NOTES and create the TONE that I have been waiting for ALL THESE YEARS.
THERE is NO LIMIT to the BEAUTIFUL DIGITAL PAN MUSIC that could be made and marketed globally. Some ah dem young 5000 players who spend all that time on ONE PANORAMA SONG need to REDIRECT their energy and get into to some MUSIC STUDIOS (to increase their skills and experiment with new sounds) or else they are going to end up LISTENING TO FOREIGNERS supplying them with WORLD CLASS PAN MUSIC!!!