Rectifying overflow and underflow in equalized audio waveforms

Details Rectifying overflow and underflow in equalized audio waveforms Rectifying overflow and underflow in equalized audio waveforms

2000-03-01

2004-09-21

Ramakrishnaiah, Melur (Department: 2643)

Data processing: generic control systems or specific application

Specific application, apparatus or process

Digital audio data processing system

C381S102000, C381S104000, C381S094800

Reexamination Certificate

active

06795740

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates generally to the field of audio signal processing. More specifically, the invention relates to an improved method, apparatus or computer program product for rectifying overflow/underflow in audio signal processing applications.
BACKGROUND OF THE INVENTION
Sound editing and signal processing applications are well known in the art and there are a number of sound editing/signal processing software packages available in the market, including SoundEdit 16 from Macromedia, Peak from Bias, or WavesConvert Pro from KS Waves. In conventional sound editing/signal processing applications, it is often desirable to increase the amplitude, power, or signal level of either the overall audio signal, or certain portions thereof. In a typical sound editing application, an equalizer is used to boost the amplitude or signal level of the entire audio signal, or portions of the audio signal which fall within specific frequency bands. However, when the amplitude or signal level of the overall audio signal, or portions of the audio signal falling within specific frequency bands is increased, the resulting amplitude or power level of the audio signal may exceed a maximum allowable amplitude or power level in certain portions of the audio signal. Alternatively, the audio signal may fall below a minimum amplitude or power level in certain portions of the resulting audio signal. Under these circumstances, the equalized audio signal is said to contain overflow and/or underflow.
FIG. 1
illustrates the concept of overflow and underflow. As shown in
FIG. 1
an audio signal exceeds an acceptable maximum positive audio level
100
in those areas
101
and
102
which are designated as overflow. Moreover, the audio exceeds an acceptable maximum negative audio level
150
in that area
103
designated as underflow. It is understood that overflow/underflow may be caused by other sound-editing applications, such as amplification, emphasis, or normalization. Overflow/underflow is a very undesirable phenomenon in any type of audio signal environment since it creates significant signal distortion and generates an annoying buzz/cracking sound. Accordingly, it is desirable to minimize, or to reduce the effect of overflow/underflow in audio signal applications.
Prior art sound editing/signal processing software packages, such as the SoundEdit
16
software package from Macromedia, reduce the effect of overflow/underflow by simply clipping audio levels where overflow/underflow occurs to the maximum allowable positive or negative signal levels.
FIG. 2
illustrates the concept of clipping the audio level of an audio signal to maximum allowable positive or negative signal levels. As shown in
FIG. 2
, an equalized audio signal is clipped to a maximum allowable positive audio level at sections
201
and
202
(where the dashed lines indicate the unclipped original signal level and the solid lines represent the clipped signal level). The audio signal in
FIG. 2
is also clipped to a maximum allowable negative audio level at section
203
(where the dashed line indicates the unclipped signal level and the solid line represents the clipped signal level).
The process of clipping the audio level of an audio signal after boosting the audio signal is a very simple approach and is used by most conventional sound-editing applications. However, the resulting or clipped audio signal suffers non-smooth changes in audio levels over time, along with significant signal loss, which typically creates an obvious and unpleasant rattling sound in the resulting audio signal.
Accordingly, what is needed is a more efficient technique for eliminating overflow/underflow in audio signal processing without clipping the audio signal in order to maintain the original audio signal characteristics and general shape of the waveform as much as possible.
SUMMARY OF THE INVENTION
The invention is a method, apparatus and computer readable program for eliminating overflow and underflow by re-adjusting or rectifying the audio levels of an audio signal in those portions where overflow/underflow occurs. Preferably, this is done after the signal has been bocisted by some other audio signal processing application, such as an equalizer.
In a first preferred embodiment a single stage scale down solution is provided which can handle overflow/underflow up to a predetermined value or degree. In this embodiment, the starting and end points of the overflow/underflow areas in the audio signal are detected, and a number of discrete sampling points along the audio signal curve between these two points are adjusted by a predetermined multiplier in order to reduce the positive level (in the case of overflow) or negative level (in the case of underflow) of each discrete sampling point. Preferably, the single stage scaling solution is used to correct overflow/underflow of up to 10 dB, although it is understood that it may be used to correct overflow/underflow greater than 10 db. Using the single stage scale down solution of the present invention, a new audio signal is created having no overflow/underflow, yet maintaining the former generally smooth shape of the audio signal waveform and signal characteristics, without completely clipping the audio levels.
In a second preferred embodiment a multi-stage scale down solution is provided which can handle overflow/underflow greater than a predetermined value. Preferably the multi-stage scale down solution is used to rectify audio signals having overflow/underflow greater than 10 db, although it is understood that the multi-stage scaling solution may be used to correct overflow/underflow less than 10 db.
In a preferred embodiment of the multi-stage scale down solution, the starting and ending points of the overflow/underflow areas are detected. The audio signal waveform located between the starting and ending points of the overflow/underflow is then divided into a number of sections. A number of discrete sampling points along the signal curve in each of the sections are then adjusted by a select multiplier in order to reduce the positive level (in the case of overflow) or negative level (in the case of underflow) of each sampling point. Each section may preferably have a different select multiplier, such that the discrete sampling points in one section may be adjusted by a first multiplier in order to reduce their audio levels; while discrete sampling points in another section may be adjusted by a second multiplier (which is preferably different than the first multiplier) in order to reduce their audio levels. Accordingly, a new audio signal is constructed with reduced overflow/underflow, yet maintaining a generally smooth signal curve and characteristics, without completely clipping the audio levels.


REFERENCES:
patent: 3631365 (1971-12-01), Dolby
patent: 3846719 (1974-11-01), Dolby
patent: 4249042 (1981-02-01), Orban
patent: 4398061 (1983-08-01), McMann, Jr.
patent: 4498060 (1985-02-01), Dolby
patent: 4507791 (1985-03-01), Gundry
patent: 4757540 (1988-07-01), Davis
patent: 4843626 (1989-06-01), Werrbach
patent: 4922535 (1990-05-01), Dolby
patent: 5168526 (1992-12-01), Orban
patent: 5509080 (1996-04-01), Roberts
patent: 5581480 (1996-12-01), Olson et al.
patent: 5714918 (1998-02-01), Menkhoff
patent: 5850453 (1998-12-01), Klayman et al.
patent: 5881157 (1999-03-01), Lee et al.
patent: 5892830 (1999-04-01), Klayman
patent: 5892832 (1999-04-01), Beyer et al.
patent: 5892833 (1999-04-01), Maag et al.
patent: 4110684904 (1999-03-01), None

Affiliated with

Also associated with

Rating

Rectifying overflow and underflow in equalized audio waveforms does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Rectifying overflow and underflow in equalized audio waveforms, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Rectifying overflow and underflow in equalized audio waveforms will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3249996