JP2016009411A - Recording device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent reduction of recordable areas due to fragmentation and to prevent recording halt.SOLUTION: A recording device includes: a memory for storing input stream data; recording means that records the stream data stored in the memory onto a recording medium; measurement means that measures a degree of fragmentation in each of a plurality of sub-areas of the recording medium; and control means that controls a first mode for selectively writing data onto sub-areas with degrees of fragmentation greater than a predetermined level and a second mode for selectively writing data onto sub-areas with degrees of fragmentation no greater than the predetermined level. The control means switches either to the first mode or the second mode depending on the situation to write initial stream data supplied after start of recording, and uses the second mode to write data after completion of the initial writing until the end of recording.

Description

  The present invention relates to a recording apparatus.

  In recent years, digital cameras and digital video cameras that record moving images and still images as files on a recording medium equipped with a file system such as a memory card have become widespread. Since such a device allows random access to various files, convenience is improved for a recording medium such as a conventional tape camcorder.

  In a NAND flash memory generally used in a memory card or the like, the smallest unit that can be rewritten by a single write request is called a page. In order to overwrite a written page, it is necessary to erase in advance. However, erasing can be performed only in units called blocks larger than the page.

  Therefore, when writing a small size to a memory with a large block size, you want to rewrite the entire block of data, but only the updated part, even though you want to rewrite only some of the pages inside the block. Since the operation of erasing and rewriting the entire block after rewriting is necessary, the speed is reduced as compared with writing of a large size in block units. Such a write operation with internal copy is called moving write. Furthermore, since the block size has increased with the recent increase in capacity of flash memory, the writing speed of small units tends to be relatively lower than the writing of large units.

  Such a decrease in writing speed is a problem in a device such as a video camera that requires continuous recording. In particular, when the recording area allocation status on the file system becomes increasingly fragmented, the writing unit becomes smaller, resulting in a decrease in writing speed. If the average rate of writing to the recording medium is lower than the data bit rate, recording cannot be continued.

  In order to deal with such a problem, in Patent Document 1, a recording area is managed in units of a management area of a predetermined size for allocating a writing area of a memory card, and writing is performed according to a ratio of an unwritten area in the management area. There has been proposed a host device that selectively allocates the management area capable of writing at a desired recording rate by predicting the speed. By performing such control, stable recording can be performed even on a memory card including a region where the file system has been fragmented.

  However, the following problems also exist in devices that perform such control. For example, consider a case in which a writing area is selected in a video camera in accordance with the unused area ratio of a management unit (AU: Allocation Unit) of a predetermined size as in Patent Document 1. FIG. 2A shows the state of the recording area after performing a short recording three times on a certain memory card. “Write 1”, “Write 2”, and “Write 3” in the figure represent the writings generated by three recordings. This is the result of recording a moving image by assigning an AU that has not been fragmented during moving image recording. The shaded area is an area that is not used on the file system, but cannot be used for moving image recording because high-speed writing cannot be performed because a part is on the used AU.

  As described above, generally, when stream data such as a moving image is recorded by the method of Patent Document 1, unfragmented AU is selectively consumed, so that the free space in fragmented AU remaining after recording is a stream. Cannot be used to write data.

  On the other hand, FIG. 2 (b) shows a situation after three short recordings are similarly performed on a memory card having a larger AU size than FIG. It can be seen that if the AU size is large, a hatched portion in the figure, that is, a large area that cannot be used for moving image recording occurs. As described above, particularly in recent memory cards in which the AU size has increased, there is a problem that the recordable capacity is greatly reduced beyond the actual amount of recorded data simply by repeating the recording.

  With respect to this problem, Patent Document 2 proposes a method of suppressing fragmentation by writing the last data at the end of recording into a fragmented area. According to Patent Document 2, stream data is buffered before being written to a memory card, and the data buffered at the end of recording is recorded in an AU where fragmentation has occurred. . As a result, since the AU fragmented at the last writing is used every time recording is performed, it is difficult for fragmentation to proceed, and it is possible to prevent the recordable area from greatly decreasing.

JP 2006-285669 A JP 2009-064386 A

  However, the following points are problems when the method of Patent Document 2 is applied to a video camera.

  A video camera is required to have excellent responsiveness in order to reliably record a moment desired by a user as an image. Therefore, even when the next recording start instruction is given before the recording stop instruction is given and the writing of the stream data is finished, it is necessary to execute the recording process without delay. However, when the method of Patent Document 2 is applied, since the next recording is started in a state where the fragmented area is selected and the writing speed is reduced, writing to the memory card is not in time. There is a fear.

  Therefore, there has been a demand for a recording apparatus that can reliably record stream data while suppressing an increase in useless free space due to AU fragmentation, and prevent an extremely low response immediately after the end of recording.

  An object of the present invention is to prevent a recordable area from being reduced due to fragmentation, and to prevent a recording stop due to a decrease in responsiveness or a buffer overflow.

  The configuration of the recording apparatus according to the present invention includes a memory that stores input stream data, a recording unit that records the stream data stored in the memory on a recording medium, and a fragment of each of a plurality of partial areas in the recording medium Measuring means for measuring the degree of fragmentation, a first mode for selecting and writing a partial region with a fragmentation rate larger than a predetermined value, and a second mode for selecting and writing a partial region with a fragmentation rate less than the predetermined value Control means for controlling the mode, and the control means switches between and applies the first and second modes according to the situation for writing the stream data supplied first after the start of recording, The second mode is applied to writing until the end of recording after the first writing is completed.

  According to the present invention, it is possible to prevent a recordable area from being reduced due to fragmentation, and to prevent a recording stop due to a decrease in responsiveness or a buffer overflow.

It is a block diagram of a recording device. It is a figure explaining the usage condition of a recording medium. It is a figure explaining AU. It is a figure explaining speed class control. It is a figure explaining an AU management table. It is a figure explaining a data buffer. It is a figure which shows the condition of a data buffer. It is a flowchart which shows write-in control. It is a flowchart which shows write-in control.

[Example]
Hereinafter, embodiments of the present invention will be described. FIG. 1 shows the configuration of a video camera 100 to which the present embodiment is applied. The video camera 100 can record moving image data shot on the memory card 112.

  The video camera 100 includes an imaging unit 101, a video signal processing unit 102, a frame memory 103, a microphone 104, an audio signal processing unit 105, a PCM buffer 106, a VIDEO codec (CODEC) 107, an AUDIOCODEC 108, a multiplexing (MUX) processing unit 109, and a free space. An area management unit 110, a memory card I / F 111, a memory card 112, an operation unit 113, a display unit 114, a CPU (central processing unit) 115, a file management unit 116, a file system 117, and a data buffer 118 are included.

  In FIG. 1, an imaging unit 101 captures a subject image and converts it into an electrical signal. The video signal processing unit 102 performs A / D conversion and appropriate image processing on the video signal obtained by the imaging unit 101. The frame memory 103 stores video frame data generated by the video signal processing unit during recording. The microphone 104 captures external audio, and the audio signal processing unit 105 performs A / D conversion and appropriate signal processing on the audio signal obtained by the microphone 104. The PCM buffer 106 stores PCM data taken from the microphone 104 and AD-converted during recording.

  A VIDEO codec (CODEC) 107 performs compression encoding on video frame data stored in the frame memory 103 and outputs it as a video stream during recording. The AUDIOCODEC 108 performs compression coding on the PCM data stored in the PCM buffer 106 and outputs an audio stream. A multiplexing (MUX) processing unit 109 multiplexes the video stream output from the VIDEOCODEC 107 and the audio stream output from the AUDIO CODEC 108, and outputs the result as an AV stream.

  The memory card I / F 111 reads / writes data from / into the memory card 112 mounted, and the memory card 112 records moving image data shot by the video camera 100. The operation unit 113 is for receiving an operation from the user and operating the video camera 100. The display unit 114 has a liquid crystal screen (not shown), and displays a video captured by the imaging unit 101 during recording and a reproduced video during playback. It also has a function of displaying a setting menu or the like for the video camera 100 or information such as recording time and remaining battery level.

  A CPU (Central Processing Unit) 115 controls the operation of each unit by software. The file management unit 116 controls file level writing during recording. The file system 117 provides file level access to the memory card 112. The data buffer 118 is a buffer for storing the AV stream generated during recording. By buffering the AV stream with the data buffer 118, the video camera 100 can start the next recording even if all the data has not been written at the end of the recording.

  FIG. 6 shows an image of the data buffer 118. The data buffer 118 operates as a first-in first-out buffer for the AV stream, and transfers the stream data input from the multiplexing processing unit 109 to the memory card I / F 111.

  The SD card defines a speed class so that real-time data can be recorded without failure. The speed class will be described with reference to FIGS. The speed class stipulates that the recording area on the card is divided into units called AUs (Allocation Units) and the degree of fragmentation is measured for each AU as shown in FIG. Specifically, AU is an integer multiple of the cluster size in FAT, and the proportion of used clusters in FAT in AU is the fragmentation rate.

  Depending on the fragmentation rate of the AU, the minimum writing speed to the free cluster in the AU is specified, and by selectively writing to the AU that can write faster than the bit rate of the data to be written Real-time data can be reliably recorded on the card.

  A specific example will be described below with reference to FIG. On some cards, as shown in the figure, if the rate of the free cluster in the AU is 100%, it is 6MB / sec, if it is 75%, 4MB / sec, if it is 50%, 2MB / sec, if it is 25%, it is 1MB / sec. Can be written. Therefore, when recording a stream with a bit rate of 3 MB / sec in real time, it is possible to continue recording without causing a buffer overflow by using an area where the percentage of free clusters is 100% or 75%. is there.

  In the video camera 100, an SD card is adopted as the type of the memory card 112, and information on the speed class is acquired from the card and used for writing control at the time of moving image recording.

(Selection of writing area during recording)
During recording, the free area management unit 110 manages the write area as follows. First, the fragmentation rate is examined for each AU, rearranged in order of the fragmentation rate, and stored as an AU information table (FIG. 5). At this time, each AU is managed by being classified into the following three types according to the fragmentation rate.

Area A: The write speed specified by the real-time writable area speed class is equal to or higher than the stream bit rate r Area B: The write speed specified by the real-time writable area speed class is the stream bit rate r Area C: Area in which no free cluster exists in the AU is an area that cannot be written because it is a used area. The AU information table is updated to the latest state each time data is written to the card during recording. Is done.

  Hereinafter, details of AV stream data writing control to the memory card 112 will be described during steady recording and at the start of recording.

(Write control during regular recording)
FIG. 8 shows the flow of writing processing during steady recording. AV stream data is always written at an integer multiple of the cluster size, and every time the write position reaches the AU boundary (Step 1 Yes in FIG. 8), the free space management unit 110 assigns an AU belonging to region A from the AU management table (FIG. 8). Step2). Subsequently, the AV stream data buffered in the data buffer 118 is written to the memory card (Step 3 in the figure). In this way, recording can be surely continued during steady recording.

(Write control at the start of recording)
When there is unwritten AV stream data from the previous recording in the data buffer 118 at the start of recording, the free space management unit 110 determines that there is no room in the data buffer 118 and switches the free space allocation method. Do.

  For example, when the second recording start operation is performed immediately after the completion of the first recording, the unwritten stream data generated in the first and second recordings is stored in the stream buffer 118 as shown in FIG. Accumulated state. The data buffer 118 includes stream data D1 generated by the first recording and unwritten stream data D2 generated by the second recording. Hereinafter, the size of D1 is S1, and the size of D2 is S2.

  Let Sth be the minimum storage amount necessary to start writing stream data after recording starts. The value of Sth is determined based on the minimum unit for multiplexing stream data. When S1 is not 0 at the time when the stream data D2 by the second recording is accumulated by Sth, the writing start of D2 is delayed until at least the writing of D1 is completed. In this case, a larger amount of data is buffered in the data buffer 118, and the possibility of buffer overflow increases. Therefore, the free space management unit 110 allocates the region A to D2.

  On the other hand, when S1 is 0, there is room in the data buffer 118, so the free space management unit 110 assigns the region B to D2.

  When area B is allocated, the greater the usage rate of the allocated area, the slower the writing speed and the higher the possibility of buffer overflow. In consideration of this, control is performed so that the writing start timing is advanced as the usage rate increases. That is, the writing is started when S2 reaches the free area size in the AU of the selected area B. On the other hand, when the area A is allocated, the writing is started when the writing of D1 is completed and S2 reaches the free area size in the AU of the selected area A.

  The above flow is shown in the flowchart of FIG. That is, after the start of the second recording, it waits for S2 to reach Sth (Step 1). If S1 is 0 (Step 2 Yes) at this time, the free space management unit 110 sets the region B for D2. Assign (Step 3). The number Nfree of free clusters in the AU of the allocated area B is obtained (step 4), and the D2 accumulation amount S2th at the start of writing is obtained from this value (step 8). S2th is obtained by S2th = Nfee × Scl, where the cluster size is Scl. After that, it waits for D2 to reach S2th (Step 9), and starts writing stream data (Step 10).

  On the other hand, when S1 is not 0 when S2 reaches S2th (Step 2 No), the free space management unit 110 assigns region A to D2 (Step 5). Then, it waits for the completion of the writing of D1 (Step 6), and obtains the number of free clusters Nfree in the AU of the allocated area A (Step 7). The control is the same as the case where the area B is allocated until the subsequent writing of D2 is started.

  As described above, the stream data can be reliably recorded while suppressing the increase of useless free space due to AU fragmentation, and the decrease in responsiveness is minimized even when the recording is repeated continuously. It is possible to provide a video camera with a limited amount.

100 video camera, 112 memory card, 101 imaging unit,
102 video signal processing unit, 103 frame memory, 104 microphone,
105 audio signal processing unit, 106 PCM buffer,
107 VIDEO CODEC (CODEC), 108 AUDIO CODEC,
109 multiplexing (MUX) processing unit, 110 free space management unit,
111 memory card I / F, 112 memory card, 113 operation unit,
114 display unit, 115 CPU (central processing unit),
116 file management unit, 117 file system, 118 data buffer

Claims (4)

A memory for storing the input stream data;
Recording means for recording the stream data stored in the memory on a recording medium;
Measuring means for measuring the degree of fragmentation of each of the plurality of partial areas in the recording medium;
Control means for controlling a first mode for selecting and writing a partial area having a fragmentation rate larger than a predetermined value and a second mode for selecting and writing a partial area having a fragmentation rate equal to or less than the predetermined value; Prepared,
The control means applies the first and second modes by switching according to the situation for writing the stream data supplied first after the start of recording, and after completing the initial writing, until the end of recording. A recording apparatus, wherein the second mode is applied for writing. After the recording end instruction, it is possible to start the supply of stream data in response to the next recording start instruction at a time before writing all of the stream data stored in the memory to the recording medium,
Of the stream data generated by the i-th recording when the stream data generated by the (i + 1) -th recording is accumulated in the memory by a predetermined amount after the start of the (i + 1) -th recording, Depending on the amount of data stored in the memory and before being written to the recording medium, the first mode and the second mode for the first stream data by the (i + 1) th recording, The recording apparatus according to claim 1, wherein the recording apparatus is applied by switching.   When the stream data is written to the recording medium at the start of recording in the first mode, the timing for starting the writing is changed according to the fragmentation rate of the partial area to be written. Item 2. The recording device according to Item 1.   When writing the stream data to the recording medium at the start of recording in the first mode, the time when the amount of stream data stored in the memory exceeds a predetermined threshold is set as the write start timing, The recording apparatus according to claim 3, wherein the write start timing is changed by setting the threshold according to a fragmentation rate of the partial area.