Note: Descriptions are shown in the official language in which they were submitted.
CA 02090948 2001-07-12
MUSI(:AL ENTERTAINMENT SYSTEM
Field of the Invention
The present invention relates generally to entertainment systems and, in
particular, to musical entertainment systems wherein a participant sings along
with
a prerecorded song.
Background of the Invention
One of the newest forms of entertainment to become popular in Japan and
the United States is karaoke. A karaoke machine typically comprises a stereo
sound system and a large: video monitor or television screen. A videotape or
videodisc player is couplevd to the video monitor to simultaneously play a
music
video while a musical song that lacks a vocal track is played on the stereo
system.
As the music video is played on the video monitor, the words of the song are
displayed at the same time as they are to be sung. A microphone is also
coupled to
the stereo system so that a participant can sing the words of the song being
played
as the music video is shown.
Not surprisingly, the quality of such impromptu singing performances
varies greatly depending on the singing ability of the participant. As a
result,
many people are hesitant to stand up and sing in front of a crowd of friends
and/or
hecklers. 'This hesitation is usually due to a perceived lack of talent on the
part of
the "would be participant." However, some people, despite the words of
encouragement, are not blessed with the ability to remain on pitch with a
musical
accompaniment being played. Therefore, a need exists for an entertainment
system that can alter the pitch of the notes sung by a participant to
correspond to
the proper pitch of the song; being played.
-2- ~)~i.~~~~~
Prior to the present invention, inexpensive equipment has not been available
to alter the pitch of a vocal signal in a way that sounds natural. While
musical
pitch shifters that can alter the pitch of a signal produced by a musical
instrument
such as a guitar or synthesizer have been well known for many years, such
devices
do not work well on vocal sounds.
In any periodic musical signal, there is always a fundamental frequency that
determines the particular pitch of the signal as well as numerous harmonics,
which
give character to the musical note. It is the particular combination of the
harmonic
frequencies with the fundamental frequency that make, for example, a guitar
and a
violin playing the same note sound different from one another. In a musical
instrument such as a guitar, flute, saxophone or a keyboard, as the notes
played by
the instrument vary, the spectral envelope containing the fundamental
frequency
and the harmonics expands or contracts correspondingly. Therefore, fox musical
instruments one can alter the pitch of a note by sampling sound from the
instrument and playing the sampled sound back at a rate either faster or
slower,
without the pitch-shifted notes sounding artificial. Although this method
works
well to shift the pitch of a note from a musical instrument, it does not work
well
for shifting the pitch of a vocal signal or sung note.
In a vocal signal, there is typically a fundamental frequency that determines
the pitch of a note an individual is singing, as well as a set of harmonic
frequencies
that add character and timbre to the note. In contrast with a musical
instrument, as
the pitch of a vocal signal varies, the spectral.envelope of the harmonics
retains the
same shape but the individual frequency components that make up the spectral
envelope may change in magnitude. Therefore, shifting the pitch of a vocal
signal
by sampling a note as it is sung and by playing back the sampled signal at a
rate
that is either faster or slower does not sound natural, because that method
varies
the shape of the spectral envelope. In order to alter the pitch of a vocal
note in a
way that sounds natural, a method is required for varying the frequency of the
fundamental, while maintaining the overall shape of the spectral envelope.
The inventors have found that the method, as set forth in the article by
K. Lent, "An Efficient Method for Pitch Shifting Digitally Sampled Sounds,"
Computer Music Journal, Volume 13, IvTo. 4, Winter, pp. 65-71 (1989)
(hereafter
referred to as the Lent method), is particularly suited for use in shifting
the pitch of
a vocal signal because the method maintains the shape of the spectral
envelope.
However, the actual implementation of the Lent method, as set forth in the
referenced paper, is computationally complex and difficult to implement in
real
IVLL\6091AP,DOC
CA 02090948 2001-07-12
-3-
time with inexpensive computing equipment. Additionally, the Lent method
requires that the fundamental frequency of a signal be known exactly.
Unfortunately, this is a problem because vocal signals are difficult to
analyze.
More specifically, because the fundamental frequency of a given note when sung
may vary considerably, it is difficult for a pitch shifter to accurately
determine the
fundamental freduency. 'fhe Lent method does not address the problem of
accurately determining the fundamental frequency of a complex vocal signal.
Therefore, there exists a need for a method and apparatus for shifting the
pitch of a vocal signal that can operate substantially in real time and be
implemented with inexpensive computing equipment. This method and apparatus
should be able to quickly analyze an input vocal signal and compare it to a
Reference Note that corresponds to the "correct" pitch of the song being
played.
The method and apparatus should then shift the pitch of the input vocal signal
so
that it is on pitch with the F:eference Note in a way that sounds natural.
Summary of the Invention
In accordance with one aspect of the present invention, a Karaoke-type
entertainment system is provided. 'The system comprises a stereo system and a
video monitor. A video player provides a video signal to the video monitor to
play
a "music video" as a musical accompaniment signal that lacks a vocal track is
played on the stereo systerr~. Included in the video signal are the words of
the song
as they are to be sung to the accompaniment. A microphone is coupled to the
stereo system so that a participant can sing the words shown on the video
monitor
as the musical accompaniment is played on the stereo system.
In accordance with another aspect of the invention there is provided a
method for shifting a pitch of an input vocal signal sung by a user of a
Karaoke
system such that the input vocal signal is on key with a prerecorded song
played by
the Karaoke system. The method involves the steps of sampling the input vocal
signal, storing the sampled input vocal signal in a digital memory, and
analyzing
the stored input vocal signal to determine the pitch of the input vocal
signal. The
method also involves the steps of reading a code, stored with the prerecorded
song,
that defines a pitch of a Reference Note, and shifting the pitch of the input
vocal
signal to be substantially equal to the pitch of the Reference Note. The pitch
of the
CA 02090948 2001-07-12
-3 A-
Reference Note defines the pitch at which the input vocal signal should be
sung in
order to be on key with the prerecorded song. The pitch of the input vocal
signal is
shifted by scaling the stored input vocal signal by a window function and
replicating the scaled input vocal signal at a rate that is a function of a
fundamental
frequency of the Reference Note. The prerecorded song may be stored on a laser
disk or a video tape and tile step of reading a code may involve reading a
subcode
stored on the laser disk or video tape which indicates the fundamental
frequency of
the Reference Note.
The method may also involve the step of combining the pitch shifted input
vocal signal and prerecorded song, and playing the combined pitch shifted
input
vocal signal and prerecorded song on the Karaoke system.
The step of scaling the stored input vocal signal may involve the step of
multiplying a portion of the stored input vocal signal by a smoothly varying
function, such as a piece-wise linear approximation of a Harming window.
In accordance with another aspect of the invention, there is provided an
apparatus for shifting the pitch of an input vocal signal sung by a user of
the
Karaoke machine so that the pitch of the input vocal signal is on key with a
prerecorded song played by the Karaoke machine. The apparatus includes a
microphone for creating an electrical signal representative of the input vocal
signal, an analog-to-digital converter, a digital memory, computing means,
means
for receiving a code, and a pitch shifter. The analog-to-digital converter is
connected to receive thc: electrical signal produced by the microphone for
producing a digitized input vocal signal representative of the singer's voice.
The
digital memory stores the digitized input vocal signal, and the computing
means
determines the pitch of the; digitized input vocal sigmal. The means for
receiving a
code, which may be stored in a MIDI format, indicates a pitch of a Reference
Note
at which the pitch of the input vocal signal should be sung to be on key with
a
prerecorded song played by the Karaoke machine. The pitch shifter shifts the
pitch
of the digitized input vocal signal to equal to the pitch of the Reference
Note.
The prerecorded song may be stored on a storage device, such as a laser
disk or a video tape or a read only memory (ROM) card, that includes a series
of
CA 02090948 2001-07-12
-3B-
codes that indicate a pitch of a series of Reference Notes at which the pitch
of the
input vocal signal should be sung to be on key with the prerecorded song.
The apparatus may also include a mixer for combining the pitch shifted
input vocal signal and the prerecorded song played by the Karaoke system.
The entertainment system of the present embodiment of the invention may
thus include a pitch corrector that determines the pitch of an input note sung
by a
participant and compares it with the pitch of a Reference Note received from
the
video player. If the pitch of the input note sung by the participant is not
equivalent
to the pitch of the Reference Note, the pitch corrector shifts the pitch of
the input
note so that the pitch substantially equals the pitch of the Reference Note.
The
pitch-shifted note is applied to an input of the stereo system and played with
the
musical accompaniment signal so that it sounds like the participant is singing
the
words of the song on pitch.
If desired, the musical accompaniment and the Reference Note may be
stored on a computer storal;e device such as a floppy disc. A sequencer
computer
may read the musical accompaniment signal and drive a synthesizer to play the
accompaniment. The seque,ncer computer may also read the Reference Notes from
the computer storage device and transmit them to the pitch corrector so the
pitch
corrector can adjust the pitch of the input note sung by the participant to
equal the
pitch of the Reference Notes.
With embodiments of the present inventive entertainment system, it is
possible to boost the performance level of even the most mediocre of singers.
CA 02090948 2001-07-12
_ c~ _
S E~rief Description of the Drawings
The foregoing aspects and many of the attendant advantages of this
invention will become more readily appreciated as the same becomes better
understood by reference to the following detailed description of specific
embodiments of the invention, when taken in conjunction with the accompanying
drawings, wherein:
FIGURF~ 1 is a block diagram of a typical karaoke entertainment system;
FIGURE 2 is a block diagram of a karaoke entertainment system
according to a first embodiment of the present invention;
FIGURE 3 is a block diagram of a pitch corrector according to the first
1 S embodiment of the invention;
FIGURE 4 is a flow chart illustrating the steps of a method for shifting the
pitch of an input vocal signal according to the first embodiment of the
invention;
FIGURE 5 is a flow chart showing the steps of a method for determining
if a note is beginning;
FIGURE 6 is a flow chart showing the steps of a method for determining
if a note is continuing;
FIGURE 7 is a flow chart showing the steps of a method for detecting
octave errors used in the method according to the first embodiment of the
invention;
FIGURE 8 is a diagram showing how the pitch of vocal signal is changed
according to the first embodiment of the invention;
FIGURE 9 shoves the steps used to generate a piecewise linear
approximation of a Harming window according to first embodiment of the
invention;
FIGURE 10 is a block diagram of a signal processor chip that is included
in the pitch corrector in accordance with the first embodiment of the
invention;
FIGURE 11 is a block diagram of a pitch shifter included within the signal
processor chip;
FIGURE 12 is a graph of an input vocal signal that is representative of a
sibilant sound; and
FIGURE 13 is a block diagram of a karaoke entertainment system
according to a second embodiment of the present invention.
-5- ,
~~~~~~,~~li~
Detailed Description of the Preferred Embodiment
To illustrate the environment in which the present invention is used, a block
diagram of a typical karaoke machine is shown in FIGURE 1. The ka.raoke
system 1 includes a video player 2, a video monitor 4, a stereo system 6 and a
microphone 30. The video player has two outputs leads. The first lead carnes a
video signal from the video player 2 to the video monitor 4, while the second
lead
carries an audio signal from the video player 2 to the stereo system 6. The
microphone 30 is coupled to an input of the stereo system 6.
As the karaoke system is used, a participant or disk jockey selects a music
video of a song to be played and inserts the video in the video player 2. As
the
music video is shown on the video monitor, the words of the song are displayed
for a participant to sing. The participant is given the microphone 30, and his
or
her singing is combined with the audio signal (i.e., the background music of
the
song) and played by the stereo system through a set of speakers 8. As
described
above, the quality of the performance given by the participant is largely
dependent
on the singing ability of the participant. The present invention seeks to
adjust the
pitch of the notes sung by the participant so that the participant sings on
pitch with
the song being played.
FIGURE 2 is a block diagram of a karaoke system 5 according to the
present invention. The system 5 is configured in the same way as the system
shown in FIGURE 1 with the addition of a pitch corrector 10. The pitch
corrector 10 is disposed between the microphone 30 and the stereo system 6.
The
pitch corrector receives an input vocal signal sung by the participant from
the
microphone 30 and determines the pitch of the input vocal signal. The pitch
corrector then compares the pitch of the input vocal signal to the pitch of a
Reference Note received on a lead 7 that extends from the video player 2 or
some
other source to an input of the pitch corrector. Preferably, the Reference
Notes are
stored as a subcode on a laser disk or a videotape in a MIDI format. It is to
be
understood that the present invention is not intended to be limited to a
karaoke
entertainment system that uses a video player as the source of the Reference
Notes;
other types of entertainment systems can also benefit from the use of a pitch
corrector of the type contemplated by the invention. In this regard, any
source of
digital information such as a MIDI-compatible keyboard, guitar synthesizer, or
ROM card can be used to provide Reference Notes to the pitch corrector.
The pitch corrector 20 compares the pitch of the input vocal signal received
from the microphone 30 with the pitch of the Reference Notes and shifts the
pitch
rvuon9~nP.noc
-6- ~ s :. ; ~, [a t >
of the input vocal signal so that it is "on pitch" with the Reference Note.
The
pitch-shifted vocal signal is applied to an input of the stereo system 6 on a
lead 9.
Therefore, the resultant sound produced by the stereo system 6 is the
accompaniment signal and a pitch-shifted input vocal signal that is "on pitch"
with
the accompaniment.
FIGURE 3 is a block diagram of a pitch corrector 10 according to the
present invention. The pitch corrector 10 receives an input vocal signal 20
and
produces a pitch-shifted output vocal signa122 on the lead 9. The pitch
corrector 10 receives the input vocal signal 20 from a microphone 30 or from
another source, such as a tape recorder, which produces an electrical signal
representative of an input vocal signal. The input vocal signal is first
applied to an
input filter 32 on a lead 34. The filter 32 preferably comprises an anti-
aliasing
filter that reduces the magnitude of any high-frequency noise signals picked
up by
the microphone 30. After being filtered by the filter 32, the input vocal
signal 20
is converted from an analog format to a digital format by an analog-to-digital
(A/D) converter 36, which is coupled to the output of the filter 32 by a lead
38.
The output of the A/D converter 36 is coupled to a signal processor 50 by a
lead 42. The signal processor block 50 receives the digitized input vocal
signal on
a lead 42 and stores it in a circular array included within a random access
memory
(RAM) 44. The RAM 44 and a read-only memory (ROM) 48 are coupled to the
signal processor block 50 by a bus 46.
The signal processor block 50 shifts the pitch of the input vocal signal by
extracting a portion of the input vocal signal 20 stored in the RAM 44 and by
replicating the extracted portion at a rate substantially equal to the
fundamental
frequency of the Reference Note, as will be described below. It should be
noted
that the term "pitch" and "fundamental frequency" of a note, as used in this
specification, are synonymous. Similarly, the period of a note is simply the
inverse of the fundamental frequency or pitch as is well known to those
skilled in
the art of musical electronics.
A bus S2 couples the signal processor 50 to a microprocessor 40 so that the
microprocessor can supply a set of parameters used by the signal processor 50
to
shift the pitch of the input vocal signal. The microprocessor 40 preferably is
an
eight-bit architecture-type chip, Model No. 80C31, made by Intel Corporation.
Coupled to the microprocessor 40 by a bus 41 are an external random-access
memory (RAM) 40a and an external read-only memory (ROM) 40b. The signal
processor 50 transfers data stored in the RAM 44 to the microprocessor 40
IVLL\d0?tAP.DOC
-
~,~ ~ 'r~ ~ E~: ~~
according to a variety of methods as will be readily apparent to those skilled
in the
art.
The output of the signal processor 50 is coupled to a digital-to-analog
(D~/A) converter 54 by a lead 56. The D/A converter 54 converts the pitch-
shifted
vocal signal from a digital format to an analog format. The output signal of
the
D/A converter 54 is in turn coupled by a lead 62 to a reconstruction filter
60. The
reconstruction filter removes any high-frequency noise signals that may have
been
added to the pitch-shifted vocal signal by the signal processor 50. The
filtered,
pitch-shifted output vocal signal is output from the pitch corrector 10 on the
lead 9.
FIGURE 4 illustrates the steps of a method, shown generally at 100, for
analyzing an input vocal signal and for shifting the pitch of the input vocal
signal
according to the present invention. The method begins at a start block 105 and
proceeds to block 110, wherein the input vocal signal is sampled and stored in
the
circular array contained within RAM 44 shown in Figure 3. Operating "in
parallel" with and independently of block 110 are two subroutines shown in
blocks 111 and 112. In block 112 an estimation is made of the fundamental
frequency of the input vocal signal, the level of the input vocal signal, and
whether
the input vocal signal is periodic. If the input signal is not periodic, block
112
returns an indication that the input vocal signal is nonperiodic as well as an
indication of whether the input vocal signal is representative of a sibilant
sound.
Sibilant sounds are sounds like "sh," "ch," "s," etc. For a pitch-shifted
vocal
signal to sound natural, the pitch of these types of sounds should not be
shifted.
Therefore, it is necessary to detect them and bypass the pitch-shifting
algorithm, as
will be described below. The operation of block 112, i.e., how the estimate of
the
fundamental frequency and the estimate of the level of the input vocal signal
are
made, is fully described in commonly assigned U.S. Patent No.4,688,464.
Briefly, block 112 determines the fundamental frequency of the input vocal
signal
based upon the time the input vocal signal takes to cross a set of alternate
positive
and negative thresholds. How the present invention detects the presence of a
sibilant sound is fully described below.
The block 111, which also operates "in parallel" with block 110, calls "an
octave error" subroutine 400. As will also be further described below, the
octave
error subroutine 400 determines if the fundamental frequency of the input
vocal
signal, determined by block 112, is an octave lower than the actual
fundamental
frequency of the input vocal signal. While the Ixnt method works well for
shifting
the pitch of a vocal signal, it is particularly sensitive to octave errors
wherein a
IVLL\61191AP.DOC
_8_
..
.:: ,.1 R3 i ':> ..~. ;,'
wrong determination is made of what octave a particular note is being sung.
Therefore, additional checks are made to ensure that a correct octave
determination
ha s been made. Blocks 11 l and 112 are routines that continually run during
the
implementation of the method 100.
After block 110, the method proceeds to a block 114, which calls a "note
beginning" subroutine 200. The note beginning subroutine 200 determines if the
input vocal signal sampled in block 110 marks the beginning of a new note sung
by
the participant. The results of the subroutine 200 are tested in decision
block 115.
If the answer to decision block 115 is no, meaning that a new note is not
beginning, the method proceeds to block 118, where a note "off' counter is
incremented and a note "on" counter is cleared. The note "off" counter keeps
track of the length of time since the last note was sung into the pitch
corrector.
Similarly, the note "on" counter keeps track of the length of time a Current
Note
has been sung by the participant. These counters help in determining what note
a
participant is singing as will be further described below. After block 118,
the
method loops back to block 114 until the answer from decision block 115 is
yes.
Once it is determined, by decision block 115, that a note is beginning, the
method proceeds to block 119 wherein a variable, Current Note, is assigned to
correspond to the pitch of the input vocal signal. For example, if the input
vocal
signal had a fundamental frequency of approximately 440 Hertz, the method
would
assign note A to the variable Current Note. The pitch of the Current Note is
then
used for comparison against the pitch of a Reference Note supplied by the
video
player (not shown).
To determine which musical note is assigned to the variable, Current Note,
a look-up table stored in the external ROM 40b shown in FIGURE 3 is used.
Contained within the look-up table are the notes of an equal tempered scale
stored
as ranges of fundamental frequencies. Therefore, for any given input signal,
there
will be a corresponding note from the table that will be assigned to the
variable
Current Note. In the preferred embodiment, the range of frequencies that
corresponds to a given note extends t 50 cents (hundredths of a semitone) on
either side of the fundamental frequency to allow for slight variations in the
fundamental frequency of the input vocal signal when assigning the Current
Note.
For example, if the participant were singing flat, such that the input vocal
signal
had a fundamental frequency of 435 Hertz, the method would still assign note A
to
the variable Current Note.
IVLL16091AP.DOC
After block 119, the method proceeds to block I20, wherein the Reference
Note is read. As described above, the Reference Note is received by the
microprocessor from the video player on a lead 7 shown in FIGURE 3. However,
other sources could be used to supply the Reference Notes such as a 1~IIDI-
compatible sequencPr, etc. After reading the Reference Note, the method
proceeds
to a block 123 wherein the pitch of the stored input vocal signal is shifted
to the
pitch of the Reference Note. The operation of block 124 is described in
further
detail below.
After block 124, the method proceeds to block 126, wherein an acceptable
range of frequencies for the next note is determined. In the preferred
embodiment,
once the variable Current Note is assigned to correspond to the fundamental
frequency of the input vocal signal in block 119, the acceptable range of
fundamental frequencies is initially set to be the fundamental frequency of
the
Current Note ~ 25 percent. By assigning an acceptable range of frequencies for
a
next note, a more educated assignment can be made each time for the Current
Note. This logic is based upon the assumption that a human voice is capable of
changing notes only at a limited rate. Therefore, if the fundamental frequency
as
determined by the block 112 falls outside of the acceptable range of
frequencies by
t 25 percent, the method assumes that the fundamental frequency reading from
2,0 block 112 is in error.
After block 126, the method proceeds to block 127 that calls a "note
continuing" subroutine 300, which determines if the Current Note is continuing
to
be sung by the participant or has ended. The operation of subroutine 300 is
fully
described below. Upon returning from subroutine 300, a decision block 128
tests
the results of subroutine 300. If the answer to decision block 128 is yes, the
method proceeds to block 130, which increments the note "on" counter. After
block 130, the method loops back to block 119, and reassigns the variable
Current
Note to be the fundamental frequency of the input vocal signal. If the answer
to
decision block 128 is no, the method proceeds to block 132, wherein the note
"on"
counter is cleared, and the note "off' counter is set to one. After block 132,
the
method proceeds to a block 134 in which a pitch shifter (not shown) is
disabled.
After block 134, the method loops back to block 114 in order to begin looking
for
a new note in the input vocal signal. The method 100 continues looking for a
new
note to begin in the input vocal signal, assigning a value to the Current
Note,
reading the Reference Note, comparing the pitch of the Current Note to the
pitch
IVLL\6091AP.DOC
-1~- ~~~~~~F~~c3
of the Reference Note, and shifting the pitch of the Current Note to equal the
pitch
of the Reference Note as long as the song that the participant is singing
continues.
FIGURE 5 is a flow chart of the "note beginning" subroutine 200 (shown in
block 114 in FIGURE 4), which determines if the participant is singing a new
note. Subroutine 20(? begins at block 205 and proceeds to block 210, wherein
the
fundamental frequency and level of the input vocal signal are read from block
112
(also shown in FIGURE 4). After block 210, the subroutine proceeds to decision
block 212, which determines if the level of the input vocal signal is above a
predetermined threshold. The threshold value is preferably set to be greater
than
the level of background noise that enters the microphone 30 (shown in
FIGURE 3). If the level of the input vocal signal is not above the threshold,
subroutine 200 proceeds to return block 214, which indicates that a new note
is not
beginning. As a result, the note "off' counter is incremented and the note
"on"
counter is cleared as shown in block 118 of FIGURE 4. If the level of the
input
vocal signal is above the predetermined threshold, subroutine 200 proceeds to
decision block 216, which determines if the input vocal signal is
representative of a
sibilant sound. The operation of block 216 is more fully described below. If
the
vocal signal is representative of a sibilant sound, the subroutine proceeds to
return
block 214.
If the input vocal signal is not a sibilant sound, the subroutine proceeds to
decision block 218, which determines if the input vocal signal is periodic.
The
answer to decision block 218 is also provided by the block 112 (shown in
FIGURE 4). If the input vocal signal is not periodic, the subroutine proceeds
to
return block 214, which indicates that a new note is not beginning. If the
input
signal is periodic, subroutine 200 proceeds to block 219 and determines if the
fundamental frequency of the input vocal signal exceeds the range capable of
being
sung by a human voice. Specifically, if the fundamental frequency exceeds
approximately 1000 Hertz, then the subroutine returns at block 214.
Having found that fundamental frequency is in the range of a human voice,
subroutine 200 proceeds from the decision block 219 and reads the note "off"
counter, as shown in block 220. After block 220, subroutine 200 proceeds to
decision block 224, which determines if the previous note has been "off' for a
time less than or equal to 100 milliseconds. If the previous note did not end
less
than 100 milliseconds ago, subroutine 200 proceeds to return block 226, which
indicates that a new note is being sung by the participant. As a result, the
Current
Note is assigned to correspond to the input vocal signal as shown in block 119
IVLL~G091AP.DOC
-11=
(FaGURE 4) and described above. If the answer to decision block 224 is yes,
meaning that the previous note did end less than or equal to 100 milliseconds
ago,
the; subroutine 200 proceeds to decision block 225. Decision block 225
determines
if there has been a large increase in the level of the input vocal signal
since the last
time subroutine 200 was called. If the level of the input vocal signal
increases
by 2, i.e., doubles, subroutine 200 proceeds to block 227, which reduces the
range
of acceptable frequencies as determined by block 126 in FIGURE 2. In the
preferred embodiment, the acceptable range is reduced from the fundamental
frequency of the previous note, t 25 percent, to the fundamental frequency of
the
previous note, ~ 12.5 percent. The present method operates under the
assumption
that a large increase in the input vocal signal precedes a point at which it
is
difficult to determine the fundamental frequency. By reducing the range of
acceptable frequencies, subroutine 200 avoids a "lock on" to a frequency that
is not
the fundamental frequency, but is instead a harmonic of the input vocal
signal.
If the answer to decision block 225 is "no," or after reducing the
acceptable range of frequencies in block 227, subroutine 200 proceeds to
decision
block 228, which determines if the fundamental frequency of the input signal
is
within the acceptable range (as calculated in block 126 of FIGURE 4 or as
reduced
in block 22?). If the answer to decision block 228 is "yes," subroutine 200
proceeds to return block 226 because a new note is beginning.
If the answer to decision block 228 is "no," meaning that the fundamental
frequency is not within the acceptable range, subroutine 200 proceeds to
decision
block 230, which determines if integer multiples (2x, 3x, 4x) or fractions
(1/2,
113, 1/4) of the fundamental frequency are within the acceptable range. If the
answer to decision block 230 is no, subroutine 200 proceeds to return block
214
because a new note is not beginning. If the answer to decision block 230 is
"yes,"
meaning that an integer multiple or fraction of the fundamental frequency lies
within the acceptable range, subroutine 200 proceeds to block 232, which
divides
or multiplies the fundamental frequency so that the result is within the
acceptable
range. For example, if the fundamental frequency is 1/3 of the expected
frequency
t25 percent, then the fundamental frequency is multiplied by 3, etc. After
block 232, subroutine 200 proceeds to return block 226 because that a new note
is
being sung by the musician.
FIGURE 6 is a detailed flow chart of "note continuing" subroutine 300
called at block 127 (shown in FIGURE 4). The purpose of subroutine 300 is to
determine whether the Current Note being sung by the participant is continuing
or
IVLL\609IAP,DOC
_12_ ~; ~,, .<y ~ ~ f~ t;
3N ~..~ Y J a ~ w ' L~
whether it has ended. Subroutine 300 begins at block 310 and proceeds to
block 312, which reads the fundamental frequency and level of the input vocal
signal as determined by block 112 (shown in FIGURE 4). After block 312,
subroutine 300 proceeds to decision block 314, which because determines if the
level of the input signal exceeds the ,predetermined threshold. If the answer
to
block 314 is "no," the subroutine 300 proceeds to return block 317 because the
Current Note is not continuing. As a result, note "on" counter is cleared and
the
note "off" counter is set to "on" as shown in block 132 of FIGURE 4. If the
level
is above the threshold, subroutine 300 proceeds to decision block 316, which
determines if the input vocal signal is representative of a sibilant sound. If
the
answer to decision block 316 is "yes," the subroutine 300 proceeds to xeturn
block 317. If the answer to decision block 316 is "no," subroutine 300
proceeds to
decision block 318, which determines if the input vocal signal is periodic, by
checking the results of block 112. If the answer to decision block 318 is
"no,"
subroutine 300 proceeds to return block 317. If the answer to decision block
318
is "yes," subroutine 300 proceeds to decision block 319, which determines if
the
fundamental frequency of the input vocal sound is within the range of a human
voice. Block 319 operates in the same way as block 219 (shown in FIGURE 5).
If the answer to decision block 319 is "no," subroutine 300 proceeds to return
block 317. If the answer to decision block 319 is "yes," subroutine 300
proceeds
to decision block 320.
Decision block 320 operates in the same way as block 225 (shown in
FIGURE 5) to determine if there is a large increase in the level of the input
vocal
signal. If the answer to block 320 is "yes," the range of acceptable
frequencies is
reduced in block 322. If either the answer to decision block 320 is "no" or
after
the range of acceptable frequencies has been reduced in block 322, subroutine
300
proceeds to decision block 324 that determines if the fundamental frequency of
the
input signal is within the acceptable range, as determined by block 126 (in
FIGURE 4) or as reduced in block 322. If the answer to decision block 324 is
"yes," subroutine 300 proceeds to return block 326, which indicates that the
note is
continuing. As a result, the note "on" counter is incremented. See block 130,
FIGURE 4 and the preceding description. If the answer to decision block 324 is
no, meaning that the fundamental frequency is not within the acceptable range,
subroutine 300 proceeds to decision block 328, which determines if integer
multiples (2x, 3x, 4x) or fractions (1/2, 1/3, 1/4) of the fundamental
frequency are
within the acceptable range. If the answer to decision block 328 is "no," the
NLL\6091AP.DOC
_13_ . :... ~, .<.,~ t., ~ y>
subroutine 300 proceeds to return block 317 because the note is not
continuing. If
the answer to decision block 328 is "yes," subroutine 300 proceeds to block
329,
which determines if there has been a jump in the octave of the input signal
and
updates octave up and octave down counters. An "octave up" jump is detected by
S a doubling of the fundamental frequency, while an "octave down" jump is
detected
by a halving of the fundamental frequency. A pair of counter variables, Octave
Up and Octave Down, keep track of the number of times the input vocal signal
jumps an octave up and down, respectively. These variables are updated in the
block 329, before the subroutine proceeds to decision block 330.
The present method of analyzing input vocal signals operates by keeping
track of the number of times the fundamental frequency determined by block 112
jumps an octave. For example, if the participant begins to sing a word that
begins
with a "W" at A-440 Hertz, the fundamental frequency may begin at A-220 Hertz,
jump to A-440 Hertz, back to A-220 Hertz, up to A-880 Hertz, etc. The two
variables, Octave Up and Octave Down, keep track of the number of times the
fundamental frequency jumps an octave from A-440 Hertz. Because the present '
method has no way of knowing which of the octaves A-220 Hertz, A-440 Hertz, or
A-880 Hertz is the correct frequency being sung by the participant, an initial
estimate is made. The initial estimate is assumed to be correct but is allowed
to
change either up or down for the first six times through subroutine 300. After
the
note has been "on" for between 100-200 milliseconds, it is necessary for the
method to "lock on" or choose one of the octaves. However, after
about 200 milliseconds, if the ratio of the number of times the fundamental
frequency drops an octave, as compared to the length of time the note has been
on,
exceeds 50 percent, then the method needs to determine whether an octave error
has been made and, thus, that the wrong choice for the octave was made
initially.
Decision block 330 determines if the Current Note has been on for a time
greater than or equal to 200 milliseconds, as determined by the note "on"
counter.
If the answer to decision block 330 is "no," then subroutine 300 proceeds to
return
block 326 because the Current Note is continuing. Upon returning to block 119
(shown in FIGURE 4), the variable Current Note is updated to reflect the new
fundamental frequency. If the answer to decision block 330 is yes, subroutine
300
proceeds to decision block 334, which determines a ratio of the count in the
Octave
Down counter to the time the Current Note has been on. If this ratio
exceeds 50 percent, subroutine 300 proceeds to block 336, which reads the
results
of the octave error subroutine 400 called for in block 111 in FIGURE 4.
rvu.rco9inr.noc
~... y ~. ,: ~ ~ f~ o
' 14' ~r ~:~ e~ ~ u' ~~ ~)
If the answer to decision block 334 is no, subroutine 300 proceeds to
block 335 which calculates a ratio of the count in the Octave Up counter to
the
time Current Note has been on. If this ratio does not exceed 50 percent, then
subroutine 300 proceeds to block 332, which corrects the fundamental
frequency.
For example, if the six readings had indicated that the fundamental frequency
was 440 Hertz and then the fundamental frequency was determined to
be 880 Hertz, the ratio of the Octave Up counter to the note "on" counter
would
not exceed 50 percent and the 880 Hertz reading would be divided by two. After
block 332 the subroutine proceeds to return block 326. If the answer to
decision
block 335 is "yes," then it is assumed that the fundamental frequency is the
correct
fundamental frequency and an error was made initially when the Current Note
was
assigned a value. Therefore, the subroutine 300 proceeds to block 337 that
clears
the note "on" and octave counters before proceeding to return block 326. Upon
returning, the Current Note will be updated to reflect the new higher octave.
If the answer to decision block 334 is "yes," then subroutine 300 proceeds
to block 336, which reads the result of the octave error subroutine. The
results of
the octave error subroutine are tested in decision block 338. If there is not
an
octave error (i.e., initial estimate of the octave of the input vocal signal
was
correct), then the fundamental frequency just determined is an octave lower
than
the actual fundamental frequency of the input vocal signal. Therefore, the
frequency is multiplied by two in block 332. If there is an octave error, then
it is
assumed that the fundamental frequency just determined is the correct
fundamental
frequency and the subroutine proceeds to return block 326 and the initial
estimate
of the octave that the participant was singing was incorrect. Therefore, the
note
"on" counter and octave counters are cleared in block 337 before returning to
block 326 so that the new fundamental frequency will now be assigned to the
variable Current Note.
Turning now to FIGURE 7, a detailed flow chart showing the operation of
the octave error subroutine 400 (referenced in FIGURE 2) is shown.
Subroutine 400 begins at start block 410 and proceeds to block 412, which
calculates the 0th lag autocorrelation (Rx(0)) of the input vocal signal for a
period
of L samples. In the preferred embodiment, L is set equal to 256. The 0th lag
autocorrelation is determined using the formula given in Equation 1:
IVW.~6091AP.DOC
-15- ~~~~~'~~~
L -1
Rx (0) _ ~x(n) ~ x(n)
n =0 (1)
where x(n) is the input vocal signal stored in the circular array within the
RAM 44 (shown in FIGURE 3). After block 412, subroutine 400 proceeds to
block 414 wherein the P/2th lag autocorrelation (Rx(P/2)) is calculated
according
to Equation 2:
L -1
Rx (P / 2) _ ~x(n) ~ x(n -P / 2)
n =0 (2)
wherein P is the period of the fundamental frequency of the input vocal
signal. If the ratio of the 0th autocorrelation to the P/2th lag
autocorrelation
exceeds 0.10 as determined by a decision block 41b, subroutine 400 proceeds to
decision block 418 that determines if the fundamental frequency is half of the
acceptable range, i.e., an octave lower than expected. If the answer to
decision
block 418 is yes, subroutine 400 proceeds to block 420, which declares an
octave
error. If the answer to either decision blocks 416 or 418 is no, subroutine
400
proceeds directly to return block 422. Subroutine 400, in effect, compares the
magnitude of the fundamental frequency of the input vocal signal to the
magnitude
of the even harmonics. Because an octave error is typically indicated by a
large
value of the even harmonics, as compared to the fundamental frequency, the
ratiometric determination can be made, and the initial estimate of fundamental
frequency then corrected to reflect the actual fundamental frequency of the
input
vocal signal.
FIGURE 8 is a diagram showing how the method of the present invention
creates a pitch-shifted vocal signal. The input vocal signal 500 is shown
having a
period Tf. A portion of the input vocal signal is extracted by multiplying the
signal
by a window 502 having a duration preferably equal to twice the period r f, .
In the
preferred embodiment, the window is shaped to be an approximation of a Hanning
window in order to reduce high-frequency noise in the pitch-shifted output
vocal
signal. However, other smoothly varying functions may be employed. The result
of multiplying the input vocal signal 500 by the window 502 is shown as a
scaled
input vocal signal504. As can be seen, the scaled input vocal signal is
substantially zero everywhere except under the bell-shaped portion of window
502.
-16- x'; "" ,1 r r ~C~3
'~L~ ~.J '~ ~::.i ~:~ 1,~
Therefore, what has been extracted from input vocal signal 500 is a portion
having
a dluration of twice the period Tf.
A pitch-shifted vocal signal 506 having an increased pitch is produced by
replicating the scaled input vocal signal 504 at a rate of fundamental
frequency of
Reference Note. By adjusting the rate at which the scaled input vocal signal
504 is
replicated, the pitch of the input vocal signal can be varied without altering
the
shape of the spectral envelope of the input vocal signal, as discussed above.
Because a Harming window 502 shown in FIGURE 8 is computationally
difficult to compute in real time with a simple microprocessor, the present
method
approximates a Hanning window using a piecewise linear approximation.
FIGURE 9 shows how the approximation of the window function 520 is computed.
For purposes of illustration, it is assumed that the period r f of the
fundamental
frequency of the input vocal signal is 63. This number is obtained from the
block 112 shown in FIGURE 4, according to the method disclosed in U.S. Patent
No.4,688,464 as described earlier. The piecewise linear approximation is
generated using two lines 522 and 524, each having a different slope and a
different duration. The line 522 is broken into two segments 522a and 522b,
with
the second line 524 disposed between them. The slope of line 522 is designated
as
Slopel, while the slope of line 524 is designated as Slope2. The calculations
of the
slopes and durations are given by Equations 3-6:
Slopel =Int (Peak / r f) (3)
Slope2 =Slopel +1 (4)
duration of Slope2 =Peak -(r f ~ slopel) (5)
duration of Slopel =rf -duration of Slopel (6)
The variable Peak is a predefined variable and in the preferred embodiment
equals 128. Applying these equations to the piecewise linear approximation 520
(shown in FIGURE 9) results in the slope of 2 for line 522 and a slope of 3
for
line 524. The duration of the segment 522a is 30, the duration of segment 522b
is 31, and the duration of line 524 is 2. Any odd durations are always added
to
line 522b. The second half of the pieeewise linear approximation 520 is made
by
providing a mirror image of the left half, having the same durations, but with
NLL\6091AP.DOC
17 ~~rt~~~~~1~
negative slopes. By using only slopes having integer values, the
multiplication
operations needed to extract a portion of the waveforms are simpler and, thus,
enable the present method to operate substantially in real time, with an
inexpensive
microprocessor. Furthermore, noninteger slope values would introduce unwanted
hi~;h-frequency modulations to the pitch-shifted vocal signal.
FIGURE 10 shows a block diagram of the signal processor block 50 as
(shown in FIGURE 3). Signal processor block 50 produces the pitch-shifted
vocal
signal, having a pitch equal to the pitch of the Reference Note. A pitch
shifter 550
is used to replicate the scaled input vocal signals at a rates equal to the
fundamental
frequency of the Reference Note. The pitch shifter 550 receives the period of
the
Reference Note from the microprocessor on a lead 552. Also supplied to the
pitch
shifter 550 on lead 556 from the microprocessor is a mathematical description
of
the piecewise Linear approximation of the Harming window. The period, T f, of
the
fundamental frequency of the input vocal signal is applied to a fundamental
timer 602 on lead 612. The lead 612 is also coupled to the microprocessor 40.
The
fundamental timer 602 is set to time a predetermined interval by Loading it
with an
appropriate number.
By loading the fundamental timer 602 with the period Tf of the fundamental
frequency of the input vocal signal, the fundamental timer 602 times an
interval
having the same duration as the period of the fundamental frequency of the
input
signal. Each time the fundamental timer times its interval, a start pointer
604 is
loaded with the start address in RAM 44 from where the portion of the input
vocal
signal is to be retrieved.
As described above, RAM 44 is configured as a circular array in which the
input vocal data are stored. A write pointer 45 is always updated to indicate
the
next available location in memory in which input vocal data can be stored. The
present method assumes that the pitch detection subroutine (shown as block 112
in
FIGURE 4) takes about 20 milliseconds to complete its determination of the
fundamental frequency of the input signal. Therefore, the point within the
circular
array from which the input vocal signal is to be retrieved can be determined
by
subtracting the number of samples of the input vocal signal taken in 20
milliseconds from the address of the write pointer 45. Thus, the fundamental
timer 602 and the start pointer 604 operate together to determine the start
address
in RAM 44 from which input vocal signal is to be extracted. Each time the
fundamental timer 602 times an interval equal to the period rf , the start
NLL\60S11AP.DOC
-18- :'f ~~, ;, ~
pointer 604 is updated to be the address at the write pointer 45 less 20
milliseconds
multiplied by the rate at which the input vocal signal is sampled.
The pitch shifter 550 multiplies the input vocal data stored in RAM 44 by
the window function. The pitch shifter 550 receives the sampled input vocal
data
on lead 614 (connected to the lead 46) and outputs the result on a leads 616.
A
switch 620 connects the output of signal processor block 50 to a lead 56 The
switch 620 is controlled by a bypass signal transmitted on lead 624 from the
microprocessor. If a note is not detected (due to sibilance, low level, etc.),
the
lead 56 receives the sampled input vocal signal from lead 614 directly, and
the
pitch shifter 550 is bypassed. As stated above, in order to make the pitch-
shifted
vocal signal sound natural, the pitch of a sibilant sound should not be
shifted.
FIGURE 11 shows a detailed block diagram of the shifter 550, as shown in
FIGURE 10. As stated above, and shown in FIGURE 8, the pitch of the input
vocal signal is shifted by replicating the scaled input vocal signal at a rate
equal to
the fundamental frequency of the Reference Note. Included within the pitch
shifter 550 is a timer 558, which is loaded with the period of the Reference
Note.
The timer 558 times an interval equal to the period of the Reference Note. As
the
timer 558 times an interval equal to the period of the Reference Note, rg, a
signal
is sent on lead 560 to fader allocation block 566. The fader allocation block
566
triggers one of four faders 568, 570, 572, and 574 to begin generating a
portion of
pitch-shifted output signal by multiplying the sampled input vocal signal by
the
window function. The fader allocation block 566 is coupled to the faders by a
set
of leads 566a, 566b, 566c, and 566d.
Included within each of the faders 568, 570, 572, and 574, respectively, is
a read pointer 568a, 570a, 572a, and 574a and a window pointer 568b, 570b,
572b, and 574b. Each time a fader is requested, the current value of the start
pointer 604 is loaded into the read pointer of the triggered fader to indicate
the
start address in RAM 44 from where the sampled input vocal signal is to be
read.
The window pointers 568b, 570b, 572b, and 574b keep track of the part of the
piecewise linear approximation of the window function that is to be multiplied
by
the input vocal data. The pitch shifter 550 includes a window table 578 that
contains a mathematical description of the piecewise linear approximation of
the
window. The window table 578 is coupled to each of the faders by lead 580.
Each fader included within the pitch shifter operates in the same manner.
Therefore, the following description of fader 568 applies equally to the other
faders.
rvvLVCO9uP.noc
-19-
~i~~~~~!.~~
Assume for example that the Reference Note has a fundamental frequency
of 440 Hz and that the input vocal signal has a fundamental frequency of 420
Hz.
Therefore, the participant is singing flat compared to the Reference Note. The
period of the fundamental frequency of the Reference Note TR equals 2.27
S milliseconds while the period of the fundamental frequency of the input
vocal
signal rf equals 2.38 milliseconds. The fundamental timer 602 is set to time
intervals of 2.38 milliseconds. Therefore, the start point is continually
updated to
be the current address of the write pointer 45 - (2.38 milliseconds * the
sampling
rate of the A/D converter 36 shown in FIGURE 3). The Reference Note timer is
set to time an interval equal to 2.27 milliseconds. Therefore, every 2.27
milliseconds an available fader begins multiplying a portion of the stored
input
vocal signal by the window function. The results of the multiplication are
output
from the four faders to summer 582, where the signals are combined to create a
pitch-shifted vocal signal. The faders read the stored input vocal signal at a
rate
equal to the sampling rate of the A/D converter 36. If the pitch of the
Reference
Note is higher than the pitch of the input vocal signal, then parts of the
scaled
input vocal signal will overlap. Similarly, if the pitch of the Reference Note
is
lower than the pitch of the input vocal signal, the signal on lead 616 will
include
some "dead space." In either case, a pitch-shifted output signal sounds
natural.
Because the window function is chosen to have a duration equal to twice the
fundamental frequency of the input vocal signal, two faders are required to
reproduce the input vocal signal with no shift in pitch. Only one fader is
required
to produce an output signal having a pitch that in an octave below the pitch
of the
input vocal signal, while four faders are required to produce an output vocal
signal
having a pitch that in an octave above the pitch of the input vocal signal. It
is
possible to alter the window function to have a duration less than two periods
of
the input vocal signal in order to reduce the number of faders required;
however,
such a reduction in the window duration results in a corresponding decrease in
audio quality. The operation of multiplying a signal by a Harming window to
create a pitch-shifted signal is fully described in the Lent paper referenced
above.
FIGURE 12 shows a graph of an input vocal signal 500 crossing a series of
predefined thresholds used by subroutine l I2 to detect a sibilant sound. As
stated
above, sibilant sounds are recognizable in the input vocal signal by the
presence of
large-amplitude, high-frequency variations. The method of pitch detection
disclosed in U.S. Patent No. 4,688,464 is altered in the present invention.
Two
thresholds at 50 percent of the positive peak value and 50 percent of the
negative
rvwnnP.ocx
-20-
~~~'~'~~l'-~~
peak value are determined. The prior method is also altered so that a record
is
made each time the input vocal signal completes the following sequence:
crossing
the high threshold, the threshold at 50 percent of the peak value, and
recrossing the
high threshold. The method by which the threshold values are determined is
fully
described in the '464 patent. In FIGURE 12, this sequence is shown completed
at
points A and C. Similarly, the method also records each time the input vocal
signal completes the sequence of crossing the low threshold, the threshold at
50
percent of the negative peak, and recrossing the low threshold. Completions of
this sequence are shown as points B and D. If 16-160 of these occurrences are
detected in less than 8 milliseconds, the method assumes that a sibilant sound
has
been detected, so that the bypass line to the pitch shifter is enabled,
thereby
bypassing the pitch shifter as described above. In the preferred embodiment of
the
pitch corrector, the number of sequences required to signal a sibilant sound
is
adjustable.
Turning now to FIGURE 13, an alternate embodiment of an entertainment
system 650 is shown. The entertainment system includes a sequences
computer 654, a video display controller 660 and a synthesizer 670. In this
embodiment a computer starage disk, R4M card or other source of digital data
652
stores the words of a particular song to be played in a computer readable form
such
as ASCII as well as the accompaniment stored in a digital format. The
sequences
computer includes a disk drive, a microprocessor and memory knot shown). The
sequences computer has three output leads; a first lead 658 is connected to an
input
of the video display controller 660. The sequences computer reads the words of
the song from the computer storage disk and transfers them in ASCII format to
the
video display controller 660. The video display controller drives the video
monitor 4 to display the words of the song as they are to be sung. A second
lead 656 of the sequences computer is connected to the synthesizer 670. The
accompaniment signal is transmitted in a suitable digital format to the
synthesizer,
causing the synthesizer to play the accompaniment as is well known to those
skilled
in the musical electronics art. Finally, the sequences computer is connected
to the
pitch corrector 10 by a lead 7. The sequences computer reads a melody track on
the computer storage device 652. The melody track contains the stored
Reference
Notes that indicate the proper pitch of the notes as they are to be sung in
the song.
The sequences computer reads the melody track and transfers the Reference
Notes
to the pitch corrector 10 so that the pitch corrector can shift the pitch of
the input
NLL\6091AP.DOC
21
signal to the pitch of the Reference Notes according to the method described
°
above.
While the preferred embodiment of the invention has been illustrated and
de scribed, it will be appreciated that various changes can be made therein
without
departing from the spirit and scope of the invention. For example, the
sequences
computer 654, video display controller 660, synthesizer 670 and pitch
corrector 10
may be separate units or may be combined as a single computer or video game
system that accepts a cartridge containing the accompaniment, lyrics and
Reference
Notes of one or more songs to be played. Therefore, it is intended that the
scope
of the invention be determined from the following claims.
rvu.uo9ine.ooc
