Multi-Pitch Estimation of Inharmonic Signals
(2013) FMS820 20131Mathematical Statistics
- Abstract (Swedish)
- A signal with harmonic structure is often characterized by its fundamental frequency,
or pitch. This single parameter contain vital information of a range of
applications such as musical transcription, tuning of stringed instruments, speech
processing and more. Some signal sources, for instance a sti vibrating string, exhibit
waveforms with slightly deviating harmonic structure. This phenomenon is
known as inharmonicity and it complicates the matter of estimating the fundamental
frequency. If several signals with this inharmonic structure are added together,
for instance for a musical chord from a stringed instrument, we arrive at the problem
formulation of this work. How does one estimate the fundamental frequency of
a... (More) - A signal with harmonic structure is often characterized by its fundamental frequency,
or pitch. This single parameter contain vital information of a range of
applications such as musical transcription, tuning of stringed instruments, speech
processing and more. Some signal sources, for instance a sti vibrating string, exhibit
waveforms with slightly deviating harmonic structure. This phenomenon is
known as inharmonicity and it complicates the matter of estimating the fundamental
frequency. If several signals with this inharmonic structure are added together,
for instance for a musical chord from a stringed instrument, we arrive at the problem
formulation of this work. How does one estimate the fundamental frequency of
a multi-pitch signal containing inharmonicities?
In this work we present a multi-pitch estimator able to handle inharmonic signals.
This algorithm uses a source separating technique in order to apply a single
pitch estimator to the resulting sub-problems. The performance of the estimator is
evaluated and compared to other multi-pitch estimators with good results. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/3629131
- author
- Nilsson, Tommy
- supervisor
- organization
- course
- FMS820 20131
- year
- 2013
- type
- H2 - Master's Degree (Two Years)
- subject
- language
- English
- id
- 3629131
- date added to LUP
- 2013-04-02 14:56:15
- date last changed
- 2013-04-02 14:56:15
@misc{3629131, abstract = {{A signal with harmonic structure is often characterized by its fundamental frequency, or pitch. This single parameter contain vital information of a range of applications such as musical transcription, tuning of stringed instruments, speech processing and more. Some signal sources, for instance a sti vibrating string, exhibit waveforms with slightly deviating harmonic structure. This phenomenon is known as inharmonicity and it complicates the matter of estimating the fundamental frequency. If several signals with this inharmonic structure are added together, for instance for a musical chord from a stringed instrument, we arrive at the problem formulation of this work. How does one estimate the fundamental frequency of a multi-pitch signal containing inharmonicities? In this work we present a multi-pitch estimator able to handle inharmonic signals. This algorithm uses a source separating technique in order to apply a single pitch estimator to the resulting sub-problems. The performance of the estimator is evaluated and compared to other multi-pitch estimators with good results.}}, author = {{Nilsson, Tommy}}, language = {{eng}}, note = {{Student Paper}}, title = {{Multi-Pitch Estimation of Inharmonic Signals}}, year = {{2013}}, }