JPH07109554B2 - Performance data processor - Google Patents

Description

TECHNICAL FIELD The present invention relates to a performance data processing device for recording / reproducing performance data of an electronic musical instrument or the like.

(B) Conventional Technology At present, a sequencer, which is a recording / reproducing device for performance data, has been put into practical use. The sequencer can record performance data when an electronic musical instrument or the like is actually played in real time, and can play back the recorded performance data with the electronic musical instrument or the like. When playing back recorded performance data, if you want to play only a part, especially if you want to repeat that part, it is possible to specify a specific section and repeat only that section from the past. A sequencer has been proposed (for example, JP-A-58-42).
No. 093).

(C) Problem to be Solved by the Invention However, in the conventional sequencer, the repetitive section cannot be specified during the reproduction of the performance data, and the data of the section to be repeated is specified by the address or the measure number. Since it had to be done, there was a drawback in that it was necessary to confirm the measure number of the repeated section in the score of this song. Also, since the fineness that can be specified is in units of measures, there is a drawback that the melody cannot be repeated as it is for songs with syncopation.

In view of the above-mentioned drawbacks of the prior art, it is an object of the present invention to provide a performance data processing device capable of setting a repeat section while reproducing performance data (song) and repeating at the set timing itself.

(D) Means for Solving the Problem The present invention is a performance data processing apparatus having a reproducing means for sequentially reading and reproducing prerecorded performance data in accordance with a clock generated at a predetermined interval. Timing input means for receiving input of at least three or more phrase timings during reproduction, and phrase timing management means for managing each input phrase timing by using a clock number which is the count number of the clocks, Designating means for arbitrarily designating the start timing and end timing of the section from three or more phrase timings, and the timing management means for determining the number of clocks for the start timing and end timing of the section designated by the designating means, Read the performance data during this period Characterized by providing a regeneration control means for controlling the reproducing device to emit.

Further, the present invention is characterized in that the designating means is the at least 3
A means for designating and inputting one phrase timing from one or more phrase timings as a section start timing, and a means for determining a phrase timing immediately after the designated and input phrase timing as a section end timing. Is characterized by.

Further, according to the present invention, the reproducing means reproduces the performance data a plurality of times so that the phrase timing can be additionally input, and the phrase timing managing means manages the phrase timings in ascending order of timing, and a new phrase timing. When is input, the phrase timing is inserted into an appropriate order of the phrase timings already managed.

(E) Action The performance data processing device of the present invention accepts input of at least three or more phrase timings during reproduction of prerecorded performance data. During reproduction, a musical piece based on performance data recorded from an electronic musical instrument or the like is being sounded, so the user may perform timing input (pressing a switch or the like) while listening to the musical piece. The input timing is managed by the timing management means based on the number of clocks. When reproducing a part of the section, if the start timing and the end timing of the section are designated from the input phrase timing, only the performance data in that section is reproduced. Since both the start timing and end timing are specified by the number of clocks, it is possible to specify the timing with an accuracy of several tenths of a measure, and it is possible to specify a fine section even in a syncopated rhythm or in the middle of a fine passage. It will be possible. Also, since you can enter three or more phrase timings and specify the start timing and end timing of the section from this, you can divide the phrase in advance and listen to each section later. It will be possible.

(F) Embodiment FIG. 12 is a block diagram of the control unit of the sequencer according to the embodiment of the present invention (FIG. 12A) and a schematic view of the operation panel (FIG. 12B).

In FIG. 1A, a CPU that controls the operation of this device is connected to each operation unit via a bus 2. A program memory 3 register group 4, an automatic performance memory 5, a tempo clock oscillator 6, an operation panel 7, and an input / output circuit 8 are connected to the bus 2. The program memory 3 is composed of ROM,
The register group 4 is composed of RAM. The automatic performance memory 5 is composed of a battery-backed RAM or a floppy disk device. Input / output circuit 8 is MI
It is a DI interface and takes in performance data as MIDI code data from a performance device such as a keyboard.
The recorded performance data (song data) is output to the tone generator.

In FIG. 1B, the operation panel 7 of this device is provided with various switch groups and indicators. The switches provided are a start / stop switch 11, a pause switch 12, a rewind switch 13, a fast-forward switch 14, a record switch (REC switch) 15, and a phrase switch 1.
6, punch-in / out switch 17, song switch 18, tempo switch 19, function switch 20, tempo up switch 21, tempo down switch 22, track number switch group 23, and numeric keypad 24. In addition, song data during RUN (recording / playback) (1 consisting of multiple tracks
Display for displaying the number of measures and beats of the performance data for a song 25
Equipped with a punch in / out switch
An LED 26 is provided in association with 17.

FIG. 13 is a diagram for explaining the structure of the automatic performance memory 5. The automatic performance memory 5 can store 99 pieces of song data (song numbers 1 to 99). Each song data storage area has an initial tempo value storage area and 10 tracks (part data storage area: track
No. 1 to 10) are provided. Performance data for each independent part can be recorded on each track.
At N time, they proceed simultaneously. Each track is composed of a performance data storage area, an empty area for storing additional data, and a punch area. The punch area is an area for temporarily buffering new performance data written by the punch-in record. Thereafter, in the present embodiment, each track is represented by a storage area group identified by PAT (i) (i = 1 to 10), and a pointer for pointing the performance data to be read out according to the progress of the song is designated as PNT (i) ( i
= 1 to 10).

14 (A) to (G) are diagrams showing the format of performance data recorded on the track. The same figure (A),
(B), (E), and (F) are formats of event data of key-on, key-off, tempo up, and tempo down, respectively, and (C) and (D) of FIG. Number) data (duration data: clock data). The long duration data of FIG. 9D is used when storing an interval exceeding one measure. Normally, event data and duration data are recorded alternately, but in the case of simultaneous pressing of chords on the keyboard, coincident event data is recorded without the duration data sandwiched. The tempo up / down tempo event data can be recorded only on the 10th track. That is, the 10th track is generally used for the rhythm part, and the rhythm tempo is controlled by the performance data of this track.

Here, the function of this sequencer will be described together with the function of each switch of the operation panel 8.

The start / stop switch 11 is a switch for RUN, that is, for starting / stopping writing / reading of the automatic performance memory 5.

When RUN starts, the song data of the selected song number is read. Of this song data, performance data is written to the track (i) for which the record mode (REC (i) = 1: described later) has been set, and the play mode (PLY (i) = 1: described later) is set. The recorded performance data is read from the recorded track (i).

The pause switch 12 is a switch for pausing during the reading of song data, or for releasing the pausing and restarting the reading.

The rewind switch 13 and the fast-forward switch 14 are used when fast-forwarding and rewinding the song data reading. Actually, it is done by increasing the frequency of the tempo clock.

Any of the pause switch 12, rewind switch 13, and fast-forward switch 14 is writing to any track (REC (i) = 1
It does not work when and PNT (i) ≠ −1: to be described later.

The record switch 15 is pressed at the same time as the track number switch described later, and the record mode (REC
Used when (i) = 1: set later.

The phrase switch 16 is a switch for setting a phrase point in the middle of a song and starting reading from the set phrase point.

If you press the phrase switch 16 and the function switch at the same time while reading the song data (automatically playing the song), the phrase point will be set for the timing, and the phrase switch → numeric keypad (1-9) → pause key before reading the song data. When the keys are pressed in the order of, song data from the phrase point indicated by the ten-key pad to the next phrase point is repeatedly read.

The punch-in / out switch 17 is a switch used when a part of the performance data that has already been recorded is replaced with new performance data (punch-in record mode), and when this switch is turned on while the song data is being read, The punch-in timing is set, and then the punch-out timing is set by turning it on again while reading the song data. The punch-in timing is between the punch-in timing and the punch-out timing.
After setting the punch-in section in this way, set the record mode to the track whose performance data you want to replace and set the RU
If set to N, the track in the record mode is being written only in this section while reading the song data.

The LED 26 provided in association with this switch is a display for displaying each stage of the punch-in record mode. It is normally off, but it flashes when the punch-in timing is set, and lights when the punch-out timing is set. After this, the song starts and flashes again at punch-in record. When the punch-in section is released (released when the punch-in / out switch is turned on for the third time), it goes out.

The song switch 18 is a switch for selecting a song number. After turning on this switch, use the numeric keypad to 2
If you enter a digit, it means that you have selected the song with that song number.

The tempo switch 19 is a switch for inputting tempo data from the ten keys. After turning on this switch, enter a numerical value (20 to 300) from the numeric keypad to set the tempo of this numerical value.

When the tempo up switch 21 or tempo down switch 22 is turned on, the tempo can be set to a desired value based on the initially set tempo.

The track number switch 23 is composed of ten key switches with numerical values of 1 to 10 and is used to set / cancel the play mode / record mode for each track of the selected song data. . As described above, when the record switch 15 is pressed at the same time, the record mode is set on the track, and when only this switch is pressed alone, the play mode set / reset inversion operation is performed.

Further, the display 25 is used to display the number of bars and the number of clocks during playback / recording.

FIG. 15 is a diagram showing registers provided in the register group 14. These registers are used for the following purposes. 16 to 18 schematically show what value is stored in each register.

SONG: Song No. register = selected song No. (= 1
~ 99) is memorized.

RUN: RUN flag = Set when song data is being read / written.

PDE: Pause flag = Set when the clock is paused during RUN (when song data read / write is paused).

STP: Progress direction register = "1" is set when reading is performed in the forward direction, and "-1" is set when reading is performed in the reverse direction (rewinding). A general clock counter GCLK (described later) is counted based on this numerical value.

FR: Fast-forward flag = This flag is set when the fast-forward switch 14 or the rewind switch 13 is fast forwarding in the forward or reverse direction.

TMP: Tempo value register = a register for storing the tempo during RUN. The tempo clock oscillator is controlled based on this data.

TMPI: Initial tempo value register = When a song is selected, the initial tempo value data stored at the beginning of the song data is assigned and stored.

DTMP: Tempo increase / decrease rate register = a register for storing the increase / decrease rate with respect to the initial tempo value calculated when a new tempo value is set in song data. This tempo increase / decrease rate is recorded on the 10th track as tempo up event data and tempo down event data.

PRS: Phrase mode flag = flag that is set when in phrase mode (mode that repeatedly plays a specified phrase section) PNC: Punch mode flag = three-state flag that takes a value of 0/1 / -1. It indicates that the punch-in section is not set and is in a normal state. When 1 indicates that the punch-in section has been set but is not in the punch-in record, and when -1, it indicates that the punch-in record is in progress.

GCLK: General tempo clock counter = Counter that accumulates tempo clock from the beginning of song data (16th
PICLK, POCLK: Punch-in timing clock register,
Punch-out timing clock counter = a register that stores clock values for punch-in timing and punch-out timing (see FIG. 17).

PNTPI (i), PNTPO (i) (i = 1 to 10): punch-in timing pointer, punch-out timing pointer = punch-in timing, the position (address) of pointer PNT (i) (see FIG. 13) at punch-out timing Remember. Since the pointer PNT (i) always points to duration data, this PNTPI (i),
PNTPO (i) indicates the duration data to which the punch-in timing and punch-out timing clock values belong (see FIG. 17).

TMPI (i): Punch-in event deviation register = register that stores the number of clocks from the event immediately before the punch-in timing (see Fig. 17) TMPO (i): Punch-out event deviation register = clock until the event immediately after the punch-out timing A register for storing the number (see FIG. 17) These TMPI (i) and TMPO (i) are used to judge the punch-in section in the punch-in record mode.

EVCLK (i): Event clock register = register that stores the clock value of the next event when RUN. This value is GC
When it matches LK, the next event is executed (see Figure 16).

REC (i), PLY (i): Record mode register, play mode register = REC when song data is running
= 1 (set) tracks are written (PNC
= 1), and reading is performed from the track of PLY = 1. By combining this mode,
It is also possible to play the keyboard or the like while playing back the performance data of a certain track and write the performance data to another track (multi-recording).

PRSBUF (n) (n = 1 to 9): Phrase register = a register for storing clock values (GCLK values) corresponding to phrase Nos. 1 to 9. When the phrase switch 16 is turned on during RUN, the clock value corresponding to the timing is stored in this register as a phrase point. The phrase points can be stored at nine positions 1 to 9, and after storage, they are sorted in the order of 1 to 9 from the smallest clock value.

PRSST, PRSEND: Phrase start clock value register, phrase end clock value register = Stores the start clock value and end clock value of the specified phrase when phrase mode is set. The start clock value is the clock value stored in PRSBUF (n) of the specified phrase n, and the end clock value is the clock value stored in PRSBUF (n + 1) (see FIG. 18).

STPNT (n, i): phrase start pointer = stores the position (address) of the pointer PNT (i) (see FIG. 13) at the phrase point (see FIG. 18).

STCLK (n, i): Phrase start event shift register = register that stores the number of clocks from the event immediately before the phrase point (see Fig. 18) STPNT (n, i), STCLK (n, i): It is set for all tracks for each phrase point.

The operation of this sequencer will be described with reference to the flow charts of FIGS.

FIG. 1 shows the main processing operation. When the power of this sequencer is turned on, the initial setting operation is first performed in n1. In this initial setting operation, all the flags in each register group are set to 0. In this state, the on / off event of each switch group is determined (n2 to n11), and when the on event of any switch (on / off event of the fast-forward switch / rewind switch only) is determined in this operation, the corresponding subroutine is executed. . That is, when the start / stop switch ON event is judged (n2), the start / stop subroutine (n1
2: (see FIG. 4) is performed, and when there is a key input from the input / output circuit 8, a (n3) key input subroutine (n13: see FIG. 9) is performed. When the RUN flag is 0, that is, when song data is not written / read (n4), the song switch on event (n5)
Alternatively, when there is a track number switch-on event (n6), a song select subroutine (n14: see FIG. 2) and a track select subroutine (n15: see FIG. 3)
To execute.

If all tracks are not currently being written,
That is, REC (i) = 1 and PNC for any i
When it is not 1 (n7), the pause switch ON event (n8) or the ON / OFF event (n9) of the fast-forward switch and rewind switch is determined. When the pause switch on event is judged, the pause subroutine (n16: see Fig. 5) is executed, and when it is judged that the fast-forward switch and rewind switch are on / off, the fast-forward /
Rewind (n15: see FIG. 8) is executed. By the way, RE
C (i) = 1 and PNC ≠ 1 means that the punch-in record mode is outside the punch-in section, and PNC is 1
Other than that, I try not to fast forward or pause because I am writing.

On the other hand, the phrase switch on event is accepted only when REC (i) of all tracks is reset (n10,
n11). When the on event of the phrase switch is accepted, the phrase subroutine (n18: see FIG. 7) is executed.

After that, in all cases, the tempo setting subroutine (n19: see FIG. 10) is executed to perform other processing such as main volume control, and then (n20) n2 is returned to and the same operation is repeatedly executed.

FIG. 2 is a song select subroutine. In this operation, first enter the 2-digit song No.
Store in G register (n21). Next, the song data specified by this song number is read from the automatic performance memory 5, and the initial tempo value stored at the beginning of that data is TM
The tempo clock oscillator 6 is stored in the P register and controlled at that tempo (n22).

FIG. 3 shows a track select subroutine. First,
The turned-on track number switch (track number) is detected and stored in the track pointer i (n23). Next, it is judged whether or not this track No. switch is turned on at the same time as the record switch (n24), and if so, the record flag REC (i) is set to put this track into the record mode ( n25). On the other hand, track No.
This track when the switch is turned on alone
The REC (i) flag of No. and the play flag PLY (i) are judged, and if either is set, both are reset (n27) and the process returns. If both are reset, set PLY (i) (n28) and return. That is, if only the track number switch is turned on repeatedly, the play / non-designated mode is repeatedly set.

FIG. 4 shows a start / stop subroutine. When the start / stop switch is turned on, the RUN flag is first inverted (n30). As a result of this inversion, RUN is set (=
If it is 1), the operation of n32 or less is executed, and if the RUN flag is reset (= 0), the operation of n37 or less is executed (n3
1).

The fact that the RUN flag is set to 1 as a result of the inversion means that writing / reading has started, so first clear the general tempo clock counter GCLK (n3
2) Substitute the start address of each track i of the song number specified in the pointer PNT (i) (i = 1 to 10) (n33). Next, it is judged whether or not REC (10) is set (n34), and if it is set, the tempo change setting is made, so the contents of the TMP register (at this time, the initial tempo value is stored. .) At the beginning of this track and is assigned to the initial tempo value register ITMP (n3
Five). After that, the phrase mode flag PRS is reset (n35) and the process returns. When REC (10) is reset at n34, the process proceeds from n34 directly to n36.

On the other hand, if the RUN flag is reset as a result of the inversion at n30, the process proceeds from n31 to n37, and the operation end process is performed. In n37, the end of the track (i) where REC (i) = 1 and PNC (i) = 0, that is, PAT (i) (PNT (i)) (track = part data storage area PAT (i), The end code FF _H is written in the address area indicated by the pointer PNT (i). Next, the track whose punch-in record has ended is searched (n38, n39, n41, n42), and if there is a track whose punch-in record has ended, it is linked to the previously recorded performance data (n40). The punch-in record is completed means the punch-in record mode (REC (i) = 1 and PNC (i) = 1), and the count value of the general clock counter has passed the punchout timing clock value (POCLK ≦ GCLK). In addition, the concatenation operation is to connect the previously recorded performance data up to immediately before the punch-in timing to the performance data written by punch-in, and then connect the previously recorded performance data after the punch-out timing. After that, the operation is to add a new consistent address. After this, proceed to n36.

FIG. 5 shows a pause subroutine. When the pause switch is turned on, the PSE flag is inverted (n43). Inversion result
If PSE is set (n44), a key-off signal that responds to the key-on key is sent (n45) and the process returns. If the PSE is reset as a result of the inversion, the process directly returns. The key-off signal is transmitted so that the key-off event does not occur for the key being sounded during the pause, so that the key-off signal is not kept ringing.

FIG. 6 shows a punch-in / out subroutine. The punch-in / out switch is turned on during the reproduction of a normal tune (in the play mode and reading the singing data).
When the punch-in / out switch is turned on, it is first determined whether or not writing is currently being performed (n46, n57). That is, if PNC ≠ 1 in the record mode, the punch-in / out switch is ignored, since the performance data is being written at present, and the routine returns. Also, PNC =
When it is 1, that is, when this switch is turned on when the punch-in section is set (n46 → n47, n46 → n47 → n52), the punch-in record mode is released. That is, the PNC flag is set to 0 and the LED is turned off (n59). To return.

On the other hand, there is no track with REC (i) = 1 (reading) P
When NC = 0 (n46, n47), the punch-in timing is set, so -1 which means that the punch-in interval has been entered is set in PNC (n48), and the general tempo clock at that time is set in the punch-in timing clock register PICLK. Substitute the value of counter GCLK (n49). Then LED
Blinking (n50), and punch-in timing pointer PMTPI (i) for all tracks is set to the pointer PNT at that time.
Substitute the value of (i) and punch-in event shift register
The number of clock deviations from the last event in TMPI (i) (=
Store GCLK- (EVCLK (i) -PAT (i) (PNT (i))) and return (n51).

On the other hand, there is no track with REC (i) = 1 (reading) P
When NC = -1 (n46, n52), the punch-out timing is set, so 1 which means the completion of punch-in interval setting is set in PNC (n53), and the general tempo clock at that time is set in the punch-out timing clock register POCLK. Substitute the value of counter GCLK (n54). Then LED
Is lit (n55) and the punch-out timing pointer PNTPO (i) for all tracks is set to the pointer PNT at that time.
The value of (i) is substituted, the number of deviation clocks (= EVCLK (i) -GCLK) to the next event is stored in the punch-out event deviation register TMPO (i), and the process returns (n56).

FIG. 7 shows a phrase subroutine. When the phrase switch is turned on, it is determined whether or not it is pressed simultaneously with the function switch (n60). If both are pressed at the same time, it is a phrase point setting operation, so there is an empty PRSB.
Substitute the value of GOLK at that time into UF (n) (n68). After substitution, the contents of PRSBUF (n) are rearranged in ascending order (n69) and the process returns.

If the phrase switch is pressed independently, it means that the phrase mode is set, so it is determined whether the RUN flag is set or reset (n61). If the RUN flag is set, it means that data writing / reading has already been performed, and therefore the phrase mode is not set and the process directly returns. If the RUN flag is reset, the input of the phrase number n from the ten-key pad is accepted at n62, and if it is entered, it is determined whether the key is the 0 key (n63). If 0 is input, the phrase buffer that has been set until then
Since this operation clears all PRSBUF, this process is executed at n70 and the process returns.

When a value other than n = 0 is input, the value of PRSBUF (n) is assigned to GCLK and the phrase start register PRSST (n
64). Set PRSBUF (n + 1) value to phrase end register PRSE
Substitute for ND (n65). Next, set the address of the first duration data of the phrase on each track to PNT (i).
And STPNT (i) (n66). This is the address when the duration data is read from the beginning and exceeds PRSST. After that, both the phrase mode flag PRS and the pause mode flag PSE are set (n67) and the process returns.

To start reading in the phrase mode, turn on the pause switch to cancel the pause. At this time, when the start / stop switch is turned on, the address set by this phrase subroutine is ignored and the reproduction is started from the beginning of the music.

FIG. 8 shows a fast-forward / rewind subroutine. This operation is performed when there is an on / off event of the fast-forward switch or the rewind switch. Determine which switch was on / off in n70 to n73. If there is an ON event of the fast-forward switch, n70 → n74
Proceed to and determine whether the FR flag is reset.
If the FR flag is set, it means that rewinding has already been performed, so the ON of this fast-forward switch is ignored and the process returns. If the FR flag is reset, the clock control subroutine (see FIG. 7B) is executed (n75).

Here, in the clock control subroutine (B) in the figure, the RUN flag is first judged (n86), and since it is being read when it is set, the clock frequency is set to fast-forward / rewind at an audible speed. Is set to 8 times the normal frequency (n87) and returns. When the RUN flag is reset, the clock frequency is set to the highest speed (n88) because no reading is in progress.

After this clock control subroutine, the FR flag is set, 1 is assigned to the STP register (n76), and the process returns.

N71 when the rewind switch ON event occurs
→ Proceed to n77 to determine whether the FR flag is reset. When the FR flag is set, it means that the fast-forward operation has already been performed, so the on-event of this rewind switch is ignored and a direct return is made. FR
When the flag is reset, the clock control subroutine is executed in n78 and the clock value of the immediately preceding event is set in the event clock register EVCLK of all tracks (i) in order to read the event data in the reverse direction. (N79). That is, EVCLK until then
Since (i) stores the clock timing of the next event, PAT (i) (PNT (i)) is subtracted to set the clock timing of the immediately preceding event. Next, the FR flag is set, -1 is assigned to the STP register (n80), and the process returns.

On the other hand, if an off event of the rewind switch has occurred, the process proceeds from n71 to n81 to determine whether STP is -1. If STP is not -1, this off event is ignored because it is not rewinding at present and the process returns. If the STP register is -1, rewinding is in progress and the operation from n82 onwards is executed to cancel this. To set the event clock to the clock value of the next event (in the forward direction)
The current PAT (i) (PNT (i)) is added to EVCLK (i) (n82), the clock frequency is returned to the normal value, and reading is resumed at the normal tempo (n83). After that, the FR flag is reset, 1 is assigned to STP (n84), and the process returns.

When the fast-forward switch off event occurs, n73 → n85
Go to and determine whether STP is 1. If the STP register is 1, fast-forwarding is in progress, and in order to cancel this, proceed to n83 to execute processing for returning to normal read operation. If the STP register is not 1 (if -1), the rewind operation is in progress, so this off event is ignored and the process returns.

FIG. 9 shows a key input subroutine. This operation is an operation executed when performance data is input from a performance device such as a keyboard via the input / output circuit 8 to correct the performance data. Since performance data is input based on the MIBI format, this data contains track numbers (channel N
o.) etc. are included. When this data is input, first the track number in the data input in n91 is set in the track pointer i. If this track is currently being written (REC = 1 and PNC ≠ 1), the operation of n93 and below is executed. If not writing, return as it is (n9
2). At n93, it is determined whether EVCLK (i) is equal to the current GCLK. If they are equal, at n96 PAT (i) (PNT
Write the key information entered in (i))
GCLK is set to (i) (n96), PNT (i) is incremented (n97), and the process returns. If EVCLK (i) is not equal to the current GCLK, GCLK-EVCLK as duration data between events in PAT (i) (PNT (i)) at n94
(I) is set (n94), the pointer PNT (i) is stepped, and then (n95) to n96.

FIG. 10 is a tempo setting subroutine. In this operation
The n100 controls the tempo clock oscillator based on the value of the TMP register.
Tempo down key 22 or tempo switch 19 → Rewrite TMP by entering tempo value from numeric keypad.
Track (10) is currently writing (REC (10) = 1 and P
If NC ≠ 1), then the operation from n102 onward is executed, and if writing is not in progress, the process returns as it is (n101). n1
In 02, EVCLK (10) = GCLK is determined, and if they match, the increase / decrease rate of the current tempo with respect to the initial tempo value ITMP of this song data is calculated and stored in the DTMP register (n10
Five). The rate of change is Calculated by If the value of DTMP is greater than 0, it means that the tempo has increased, so PAT (10) (PNT (10)) contains the tempo increase event data DO _H , DTMP (see Fig. 14 (E)).
After writing (n107), the pointer PNT (10) is incremented (n109) and the process returns. Further, after written in DTMP is the case was negative a tempo down data for a tempo down C0 _H, the -DTMP PAT (10) (PNT ( 10)), is incremented the PNT (10) (N109) to return.

When EVCLK (10) and GCLK do not match in n102, GCLK-EVCLK (10) as duration data
Is set (n103), the pointer PNT (10) is stepped, and the process proceeds to (n104) and n105.

FIG. 11 shows a clock interrupt subroutine. This operation is executed at every predetermined timing by the interruption of the tempo clock oscillator 6. In this operation, first, it is determined that reading / writing or the like based on the clock signal is performed, that is, RUN = 1 and PSE = 0 or FR = 1. (N121). If this condition is not satisfied, the processing operation by the clock interrupt is not necessary and the process returns as it is. If the condition is satisfied, the operation from n122 onward is executed.

For n122 or lower, first set the general clock register GCLK.
The clock is updated by adding the STP value (n122). Since STP is +1 at the time of normal read / write and fast-forward, GCLK is incremented by 1, and at rewind, it is -1, and GCLK is decremented by 1. After that, in n123, n126, and n131, it is determined whether or not it is the repeating point in the phrase mode, whether or not it is the punch-in timing, and whether or not it is the punch-out timing. In other cases, the normal event clock step of n136 or less,
Proceed to the event data processing operation.

If it is determined in n123 that the phrase is repeated, the process proceeds to n124 and thereafter. The phrase repeat point is the phrase mode flag PRS = 1 and GCLK.
= It can be judged by PRSEND (phrase end register). The n124 sends a key-off signal, which responds to the key being turned on, to the tone generator in order to mute the tone being sounded, and repeats from the beginning of the phrase GCLK ← PR
Set SST and set all PNT (i) ← STPNT (i), EVCLK
(I) ← Set STCLK (i) and proceed to (n125) n150.

If n126 determines that it is the punch-in timing,
Executes the operation of n127 and below. The punch-in timing is REC (i) = 1, PNC (i) = for any track.
It can be determined by 1 and GCLK = PICLK. At n127, the PNC is set to -1 indicating that punch-in recording is in progress, and the LED is made to blink (n128). After that, set PICLK-TMPI (i) to EVCLK (i) in track i with REC (i) = 1, and set PN
After setting the start address of the punch-in interval to T (i) (n129), the PAT pointed to by this PNT (i)
(I) After writing the key-off information corresponding to the currently key-on key to (PNT (i)) (n130), the process proceeds to n150.

On the other hand, if it is determined in n131 that it is the punch-out timing, the operations in and after n132 are executed. PNC for n132
Set 1 to. Punch-out timing means that REC = 1, PNC = -1 and GCLK on any track.
= It can be judged by the fact that it is POCLK. Next, in n133, the LED is turned on, and PAT (i) (PNT (PNT) (PNT (i)
The data of GCLK-EVCLK (i) is stored in (i)) (n13
4), write GCLK + TMPO (i) to EVCLK (i) PNT
After writing PNTPO (i) in (i), proceed to (n135) n150.

The operation of n136 and below is a normal event clock stepping and event data processing operation for each track. First, at n136, 1 is set to the track pointer i, and at n137 and n138, performance data is being read (PLY (i) =
1) or punch-in record (REC (i) = 1 and PNC = -1) and EVCLK (i) = GCLK are determined. If this condition is met
Executes the operations below n139. If this condition is not satisfied, it is not necessary to perform the following operation, so jump to 145 and add 1 to i (n145), and n137 until i exceeds 10.
The following operations are repeated (n146).

In n139, the pointer PNT is referred to the next clock information (duration data) in the clock progress direction by referring to STP.
Jump (i) (n139). If there is no next clock information in this jump operation, it is considered that the music is finished and the process proceeds from n140 to n141. When the next clock information exists, the event information (key-on / key-off, etc.) at this time is processed (n143), except in the case of fast forward / rewind (n142). Forward playback (STP =
In the case of 1), the processing according to the content is executed, and in the case of reverse feed reproduction (STP = -1), the processing having the opposite content is executed. The process having the opposite content is, for example, a key-off process in the case of a key-on event and a tempo-down process in the case of a tempo up event. Then n144
In EVCLK (i), PAT (i) which is clock information
Multiply (PNT (i)) by STP and add. Therefore, in the case of rewinding, PAT (i) (PNT (i)) is subtracted. After this, proceed to n145.

When there is no next clock information and the operation proceeds to n141, in the case of the phrase mode, it proceeds to n141 → n147, sets GCLK ← PRSST to repeat from the beginning of the phrase, and sets all PNT (i) ← STPNT (i ), EVCLK (i) ← STCL
Set K (i) and proceed to n150. In other cases, the RUN flag is reset (n148) to end the performance, and the process returns.

In n150, the display shows the number of measures and the number of clocks being written / read at that time. The number of displayed bars is calculated by GCCK / 32, and the number of clocks is calculated by GCLK.MOD.32. After this display, return.

(G) Effects of the Invention As described above, according to the performance data processing device of the present invention,
Since one or more phrase timings can be input in real time during playback of performance data, the user can input phrase timings at the same timing as the performance, and the phrase timings can be easily and accurately set. It becomes possible to enter. Since the input timing is managed by the number of clocks, it is possible to manage the phrase timing at the user's desired timing even if the melody or rhythm is syncopated or in the middle of a detailed passage, and the correct section playback Will be possible. Further, since the start timing and end timing of a section can be arbitrarily specified from three or more phrase timings, it is possible to input many phrase timings in advance and then reproduce various sections. This can contribute to the improvement of the operability of the performance data processing device.

[Brief description of drawings]

1 to 11 are flowcharts showing the operation of the sequencer according to the embodiment of the present invention. FIG. 1 shows the main processing operation, and FIGS. 2 to 10 show the on / off operation of each key switch. FIG. 11 shows a clock interrupt operation. FIG. 12 is a block diagram of a control unit of the sequencer and a diagram showing an operation panel, FIG.
The figure shows the structure of the automatic performance memory of the sequencer.
14 (A) to (G) show the format of various data stored in the automatic performance memory, FIG. 15 shows the configuration of the register group of the sequencer, and FIGS. 16 to 18 show various modes. It is a figure for demonstrating the content of the data set to a pointer and various clock registers at time.

Claims (3)

[Claims] 1. A performance data processing apparatus having a reproducing means for sequentially reading and reproducing prerecorded performance data in accordance with a clock generated at a predetermined interval, wherein at least three or more performance data are reproduced while the reproducing means reproduces the performance data. Timing input means for receiving an input of the phrase timing, phrase timing management means for managing each input phrase timing by using a clock number which is the count number of the clock, and The designating means for arbitrarily designating the start timing and the end timing of the section and the number of clocks for the start timing and the end timing of the section designated by the designating means are calculated by the timing management means, and the performance data in the meantime is read out. Control playback means And a reproduction control means for controlling the performance data processing device. 2. A means for designating and inputting one phrase timing among the at least three or more phrase timings as a section start timing,
2. The performance data processing apparatus according to claim 1, further comprising means for determining a phrase timing immediately after the designated and input phrase timing as an end timing of the section. 3. The reproducing means is capable of reproducing performance data a plurality of times and additionally inputting phrase timing, and the phrase timing management means is means for managing phrase timing in ascending order of timing, and
2. The performance data processing apparatus according to claim 1, further comprising means for inserting, when a new phrase timing is input, the phrase timing into an appropriate order of the phrase timings already managed.