JP7361514B2 - Image sensor and imaging device - Google Patents

Description

The present invention relates to an imaging device and an imaging device.

In recent years, an imaging device called a SPAD (Single Photon Avalanche Diode) has been studied. This is an imaging device that uses the avalanche phenomenon that occurs when an avalanche photodiode (APD) is operated in Geiger mode to measure the number of incident photons itself, that is, treats incident light as a digital value.

When the APD is operated in Geiger mode, for example, when one photon is incident on the APD, an observable level of current is generated due to an avalanche phenomenon. By converting this current into pulses and counting the number of pulses, it is possible to directly measure the number of incident photons, so RTS noise is not generated and S/N is expected to improve. . Such operation is called photon counting.

In photon counting, since incident light is treated as a digital value, the basic condition is that one AD converter is provided in one pixel. This is a type of image sensor called a so-called pixel AD. Since the AD converter only needs to count the number of incident photons, it can be implemented as long as it has a comparator and a counter that convert the current generated by the avalanche phenomenon into pulses.

As an example of a sensing device using photon counting, Patent Document 1 discloses a ranging sensor consisting of a SPAD having multiple pixels.

On the other hand, there is a technique that performs phase-difference focus detection at the same time as imaging by using a configuration in which pixels in an image sensor receive light from different pupil planes of an imaging lens. For example, in Patent Document 2, by dividing a photodiode (hereinafter referred to as "PD") located in one pixel and focused by one microlens, each PD is attached to a different pupil plane of the photographing lens. It is configured to receive light.

As a result, the amount of image shift is detected from the signals of each PD that receives light that has passed through different pupil planes of the photographic lens, that is, phase difference detection is performed, and the amount of focus shift (defocus amount) is correlated from the amount of image shift. The focus state is detected by calculating. It is also possible to generate a photographic signal for viewing by adding signals from each PD that receives light from different pupil planes of the imaging lens.

Japanese Patent Application Publication No. 2014-081253 Japanese Patent Application Publication No. 2001-083407

In order to realize a configuration having a plurality of photoelectric conversion elements in order to obtain phase difference information in the pixel AD, which is typified by photon counting, it is conceivable to provide as many AD converters as there are photoelectric conversion elements. However, since the AD converter requires a large area for just the comparators and counters, providing as many AD converters as the number of photoelectric conversion elements hinders pixel miniaturization. Therefore, in a configuration in which one pixel includes a plurality of photoelectric conversion elements, a configuration that can perform efficient readout even with one AD converter per pixel is required.

The present invention provides a technology that is advantageous in achieving both miniaturization of pixels and improvement in readout speed in an image sensor including a plurality of photoelectric conversion elements in one pixel.

According to one aspect of the present invention, the pixel unit includes a plurality of unit pixels arranged in a matrix, each of the plurality of unit pixels photoelectrically converting light passing through different exit pupils of a photographing lens. a pixel portion having a photoelectric conversion element, a plurality of AD converters provided for each of the plurality of unit pixels, and reading from the plurality of unit pixels using the plurality of AD converters. and a pixel control section that controls the pixel signals of the frame, the pixel control section dividing the frame into a plurality of subframes and controlling the plurality of subframes for each subframe. A readout operation is performed from a unit pixel, and in the readout operation of a plurality of first subframes among the plurality of subframes, the plurality of photoelectric conversion elements obtained in each of the plurality of first subframes are read out from a unit pixel. A first signal based on a signal from one of the photoelectric conversion elements is read out , and in a readout operation of a plurality of second subframes among the plurality of subframes, each of the plurality of second subframes is read out. A second signal based on signals from all of the plurality of photoelectric conversion elements obtained in each of the plurality of photoelectric conversion elements is read out , and in the frame, the plurality of first subframes and the plurality of second subframes are read out. is alternately repeated, and in other frames different from the frame, the number of the plurality of second subframes is greater than the number of the plurality of first subframes. An image sensor characterized by controlling readout is provided.

According to the present invention, it is possible to provide a technique that is advantageous for achieving both miniaturization of pixels and improvement of readout speed in an image sensor including a plurality of photoelectric conversion elements in one pixel.

FIG. 1 is a diagram showing the configuration of an image sensor in an embodiment. FIG. 3 is a diagram showing the configuration of one pixel of an image sensor in an embodiment. FIG. 2 is a diagram showing the configuration of a pixel and the configuration of a circuit that reads signals from the pixel. FIG. 3 is a diagram schematically showing the operation of an avalanche photodiode. FIG. 3 is a timing chart schematically showing an example of driving an image sensor. FIG. 1 is a diagram showing the configuration of an image sensor in an embodiment. FIG. 3 is a timing chart schematically showing an example of driving an image sensor. FIG. 3 is a timing chart schematically showing an example of driving an image sensor. FIG. 3 is a timing chart schematically showing an example of driving an image sensor. FIG. 1 is a diagram showing the configuration of an imaging device according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Note that the following embodiments do not limit the claimed invention. Although a plurality of features are described in the embodiments, not all of these features are essential to the invention, and the plurality of features may be arbitrarily combined. Furthermore, in the accompanying drawings, the same or similar components are designated by the same reference numerals, and redundant description will be omitted.

<First embodiment>
FIG. 1 is a diagram showing the configuration of an image sensor in an embodiment. The image sensor 100 includes a first semiconductor chip 101 that forms a sensor substrate and a second semiconductor chip 102 that forms a circuit board that are stacked on each other. A pixel array (pixel section) in which a plurality of unit pixels 103 are arranged in a matrix is formed on the first semiconductor chip 101 forming the sensor substrate. The detailed configuration of the unit pixel 103 will be described later. The second semiconductor chip 102 forming a circuit board may include a main control section 104, a pixel control section 105, and a signal processing circuit 106.
The main control unit 104 receives operation instructions, serial communication, etc. from the outside, and supplies timing signals and the like for the pixel control unit 105 and the signal processing circuit 106 to operate. The pixel control unit 105 is electrically connected to each of the plurality of unit pixels 103 of the first semiconductor chip 101 by a bump or the like, and outputs a control signal for driving each of the plurality of unit pixels 103, and also controls the buffer output from the pixel. Receive output.

The pixel control unit 105 may include a plurality of AD converters, one for each of the plurality of unit pixels 103. The pixel control unit 105 controls reading from a plurality of unit pixels 103 using a plurality of AD converters. Specifically, the pixel control unit 105 is provided with a counter for each corresponding unit pixel. The pixel control unit 105 has a circuit that uses this counter to measure the number of pulses output from the buffer output in response to incident photons. The count value measured by the pixel control unit 105 can be outputted to the outside by the signal processing circuit 106.

FIG. 2 is a diagram showing the configuration of one pixel (unit pixel 103) of the image sensor 100. Unit pixel 103 has one microlens 202. Furthermore, the unit pixel 103 includes a plurality of photoelectric conversion elements that photoelectrically convert light passing through different exit pupils of the photographic lens. For example, the unit pixel 103 has two avalanche photodiodes (hereinafter referred to as "APD") that operate in an avalanche state, and are shown as APD 203 and APD 204 in FIG. 2.

In this way, by sharing one microlens 202 with multiple APDs, out of the light from the photographing lens that reaches the image sensor 100, the light passing through different exit pupils is divided and received, and phase difference information is obtained. It is possible to obtain. In addition to the illustrated components, the unit pixel 103 may include a plurality of components described below. Further, although the unit pixel 103 is described here as having two APDs, the number of APDs may be any number as long as it is two or more.

FIG. 3 is a diagram showing the configuration of a pixel and the configuration of a circuit that reads signals from the pixel. The unit pixel 103 in the embodiment has two APDs 203 and 204, as described above. By sharing the microlens 202, the APDs 203 and 204 each receive light that passes through different exit pupils out of the light from the photographing lens.
Quench transistors 301 and 302 operate as quench resistors when turned on, and are driven by driving pulses φQA and φQB, respectively. Switches 303 and 304 are switches for selecting which pixel to output, and are driven by drive pulses φTA and φTB, respectively. AD converter 310 may include a comparator 305 and a counter 306. Comparator 305 may function as a pulse shaping circuit.

APDs 203 and 204, quench transistors 301 and 302, and comparator 305 are electrically connected via switches 303 and 304. A voltage H _VDD is applied to the APDs 203 and 204 via quench transistors 301 and 302, respectively. The voltage H _VDD is set to a voltage at which the APDs 203 and 204 operate in Geiger mode.

The comparator 305 has one of the output voltages V _APD of the APDs 203 and 204 as a first input, and a predetermined comparison voltage V _comp as a second input. By outputting the comparison result V _out as "H" or "L", the comparator 305 functions as a pulse shaping circuit. By counting the output of the comparator 305, the counter 306 performs AD conversion on the signals of the APDs 203 and 204, and records the converted data as digital data.

Next, the specific operation of the pixel will be explained using FIG. 4. FIG. 4 is a diagram schematically showing the operation of the APD. Since the operating principles of APD 203 and APD 204 are the same, only the APD 203 and the circuit that reads out the signals of APD 203 will be explained here. First, the quench transistor 301 is turned on and the switch 303 is turned on to enter a read standby state.

When the APD 203 is on standby, no current flows through the APD 203 and the quench transistor 301, and V _APD indicates H _VDD . That is, when the APD 203 is on standby, a voltage is applied to the APD 203 that can cause an avalanche phenomenon by photon incidence. Note that at this time, the comparator 305 outputs "L".

After that, when photons are incident on the APD 203, an avalanche phenomenon occurs, and current begins to flow through the APD 203 and the quench transistor 301. When current begins to flow through the quench transistor 301, a voltage drop occurs and the voltage at V _APD decreases. When V _APD becomes lower than the comparison voltage V _comp , the output V _out of the comparator 305 becomes "H".

Thereafter, when V _APD drops to a voltage that makes it impossible to generate an avalanche phenomenon, the avalanche phenomenon stops. As a result, V _APD returns to H _VDD , and the output V _out of comparator 305 becomes "L".

Through these operations, incident photons are extracted as pulse signals. When the counter 306 counts the pulse output of the comparator 305, the signal of the APD 203 is AD converted and recorded as digital data.

The APD 204 also operates in the same manner to extract a pulse signal. Furthermore, the comparator 305 may be of a simple inverter type, as long as it can shape V _APD as a pulse.

In the embodiment, one AD converter is provided for each of a plurality of unit pixels. That is, a comparator 305 and a counter 306 are mounted for each pixel to read out the signals of the APDs 203 and 204. Therefore, it is possible to read out signals from all pixels simultaneously. However, since the comparator 305 and the counter 306 are provided for each pixel and not for each of the APDs 203 and 204, the signals of the APDs 203 and 204 cannot be taken out simultaneously as separate signals.

A read operation in an embodiment that addresses this will be described with reference to FIG. FIG. 5 is a timing diagram schematically showing driving of the image sensor in the embodiment. It shows how time passes from left to right in the drawing. In this embodiment, in order to read out pixel signals of one frame, the pixel control unit 105 time-divides one frame into a plurality of subframes and performs a readout operation from a plurality of unit pixels 103 for each subframe. configured. The pixel control unit 105 simultaneously reads out all pixels for each subframe, synthesizes the information of each subframe, and generates one frame of information.

FIG. 5 shows how to read one frame of information, the subframes are divided into a plurality of subframes 1_1 to 1_10 and the read operation is performed ten times. Here, in the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9, only the APD 203 is in the operating state. For this purpose, the quench transistor 301 and the switch 303 are turned on by setting φQA and φTA to high level. On the other hand, by setting φQB and φTB to a low level, the quench transistor 302 and the switch 304 are turned off, and as a result, the counter 306 performs a counting operation corresponding to the number of photons that have entered the APD 203.

On the other hand, in the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10, both APD 203 and APD 204 are in the operating state. Therefore, by setting all of φQA, φTA, φQB, and φTB to a high level, the quench transistors 301 and 302 and the switches 303 and 304 are turned on. At this time, since both APDs 203 and 204 are in the operating state, the counter 306 performs a counting operation on the photon signals incident on both APDs 203 and 204. In this way, one frame can be divided into a plurality of subframes by changing the driving frequencies of φQA, φTA, φQB, and φTB.

Here, all the count values obtained in the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9 are combined. Then, signals of photons incident on the APD 203 (first photoelectric conversion element) for the times of the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9 are obtained. The signal generated in this way is called the "A signal" (first signal). However, within the time of one frame, the period during which the A signal is obtained by the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9 is half of the signal obtained during one frame in this example. Therefore, by applying a double gain to the A signal, a signal close to the signal expected to be obtained in one frame time can be obtained. This is called the "A2 signal."

Similarly, all the count values obtained in the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10 are combined. Then, signals of photons that entered the APD 203 and APD 204, which are all photoelectric conversion elements (second photoelectric conversion elements), are obtained for the time period of the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10. The signal generated in this way is called the "A+B signal" (second signal). However, within the time of one frame, the period during which the A+B signal is obtained by the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10 is half of the signal obtained during one frame in this example. Therefore, by applying a double gain to the A+B signal, a signal close to the signal expected to be obtained in one frame time can be obtained. This is called the "(A+B)2 signal."

Here, by subtracting the A2 signal from the (A+B)2 signal, a signal as a phase signal is obtained. This signal is called the "B2 signal" (third signal). The (A+B)2 signal corresponds to the signal of the light that has passed through the exit pupil of the photographic lens and has been photoelectrically converted by the APDs 203 and 204 obtained during one frame, so it is used as an imaging signal for display, recording, etc. be done. On the other hand, the A2 signal and the B2 signal are signals of light that has passed through different exit pupils of the photographing lens, photoelectrically converted by the APD 203 and APD 204, so the A2 signal and the B2 signal are phase signals having phase difference information.

By performing a known correlation calculation from the A2 signal and the B2 signal, it is possible to extract phase difference information, and it is possible to calculate the defocus amount of the photographing lens. The A2 signal and the B2 signal are information necessary to obtain phase difference information, and are not used for display or the like. Therefore, as long as the signals necessary for the calculation are obtained, it is not necessarily necessary to combine the signals of each subframe, and the correlation calculation can be performed using information from a single subframe. You don't have to put it on.

Similarly, for the second and subsequent frames, one frame is divided, read out, and combined to obtain an imaging signal and a phase signal.

In this embodiment, since a SPAD type image sensor capable of photon counting is used, even if signals of a plurality of subframes are combined to form one frame, there is almost no noise deterioration due to the combination. Note that the above-mentioned synthesis operation, difference operation, correlation calculation, etc. may be performed by the signal processing circuit 106, or may be performed by a signal processing unit outside the image sensor, which will be described later.

In FIG. 5, one frame is divided into 10 subframes, but the present invention is not limited to this. One frame can be divided and read out, and the imaging signal and the phase signal can be divided and read out. It is characterized by Further, the (A+B)2 signal is used as pseudo information in one frame by doubling half of the information in one frame. Therefore, if the number of divisions of one frame is small, when photographing a subject moving at high speed, there will be a large deviation from the image signal that has all the information in one frame. Therefore, one frame may be divided into, for example, 30 or more subframes.

In addition, in the above example, the gain multiplied by the A signal and the A+B signal is set to 2 as an example, but this means that the proportion occupied by the first subframes and the second subframes among the subframes is 2. That's because it's half and half. More generally, the A signal is multiplied by a gain corresponding to the proportion of the first subframes among the subframes, and the A+B signal is multiplied by the proportion of the second subframes among the subframes. It is preferable to multiply by a gain according to the proportion occupied by the frame.

As mentioned above, even with a configuration in which one AD converter is provided for one pixel, that is, even if the number of AD converters is not the same as the number of photoelectric conversion elements, high-speed readout operations can be achieved. can.

<Second embodiment>
FIG. 6 is a diagram showing the configuration of an image sensor according to the second embodiment. In this embodiment, a general photodiode (hereinafter referred to as "PD") that does not perform an avalanche operation is used. Like the APDs 203 and 204, the PDs 601 and 602 are photoelectric conversion elements that photoelectrically convert light that has passed through different exit pupils of the photographic lens. Transfer switches 603 and 604 transfer charges generated in PDs 601 and 602, respectively, to floating diffusion sections (hereinafter referred to as "FD"), which will be described later. The FD 605 has the role of a memory that temporarily stores charges. A reset switch 606 resets the FD 605 using VDD. AD converter 210 may include a comparator 205 and a counter 206.

Even in the configuration shown in FIG. 6, if the FD 605 is sufficiently small, it is possible to read signals from incident photons. Therefore, if driven in the same manner as shown in Fig. 5, even with one AD converter per pixel, it is possible to realize a readout operation that separately acquires the image signal and the phase signal as separate signals. It becomes possible. These are the same even if the photodiode is replaced with a photoelectric conversion element such as a photoelectric conversion film, so the present invention does not limit the type of photoelectric conversion element.

Furthermore, even if the photodiode signal is a weak voltage that cannot be separated by the comparator 205, by changing V _REF over time, the comparator 205 operates as a slope type AD converter, and the counter 206 converts the signal. It is possible to count. However, in this case, read noise occurs, so unlike the photon counting method in which photons are counted, noise increases when subframes are combined. However, even with one AD converter for one pixel, it is possible to implement a readout operation in which the imaging signal and the phase signal are individually acquired as separated signals.

<Third embodiment>
FIG. 7 is a timing diagram schematically showing driving of the image sensor according to the third embodiment. In the example of FIG. 7, only the APD 203 is in the operating state in the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9. For this purpose, the quench transistor 301 and the switch 303 are turned on by setting φQA and φTA to high level. On the other hand, by setting φQB and φTB to a low level, the quench transistor 302 and the switch 304 are turned off, and as a result, the counter 306 performs a counting operation corresponding to the number of photons that have entered the APD 203.

On the other hand, in the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10, only the APD 204 is in the operating state. For this purpose, quench transistor 301 and switch 303 are turned off by setting φQA and φTA to low level. On the other hand, by setting φQB and φTB to high level, the quench transistor 302 and the switch 304 are turned on, and as a result, the counter 306 performs a counting operation corresponding to the number of photons that have entered the APD 204.

Here, all the count values obtained in the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9 are combined. Then, signals of photons incident on the APD 203 (first photoelectric conversion element) for the times of the plurality of first subframes 1_1, 1_3, 1_5, 1_7, and 1_9 are obtained. This generates the A signal (first signal).

Similarly, all the count values obtained in the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10 are combined. Then, signals of photons incident on the APD 204, which is one photoelectric conversion element (second photoelectric conversion element) different from the APD 203, are obtained for the time period of the plurality of second subframes 1_2, 1_4, 1_6, 1_8, and 1_10. As a result, a B signal (second signal) is generated.

By such driving, the A signal and the B signal can be directly obtained. The pixel control unit 105 can generate an imaging signal (fourth signal) by adding the A signal and the B signal. In such driving, even in a configuration in which one AD converter is provided for one pixel, the imaging signal and the phase difference detection signal are individually acquired as separate signals, as in the first embodiment. It is possible to implement a read operation.

<Fourth embodiment>
FIG. 8 is a timing chart schematically showing another example of driving the image sensor. In FIG. 8, in the first frame, as in FIG. 5, subframes belonging to a plurality of first subframes and subframes belonging to a plurality of second subframes are set alternately. On the other hand, in the second frame (another frame), the number of second subframes (subframes indicated by A+B) is greater than the number of first subframes (subframes indicated by A).

By driving in this manner, only the necessary amount of phase difference information is extracted and the remaining subframes are used for reading the A+B signal, thereby making it possible to increase the signal ratio of the A+B signal in one frame. Naturally, the gain when used for the imaging signal is not doubled, but it is sufficient to apply the gain by the amount of the missing subframe.

FIG. 9 is a timing diagram schematically showing still another example of driving the image sensor. In FIG. 9, in the first frame, as in FIG. 8, subframes belonging to a plurality of first subframes and subframes belonging to a plurality of second subframes are set alternately. On the other hand, the second frame (other frame) is composed of only a plurality of second subframes. That is, in the second frame, only the A+B signal is read out.

Note that this is an example of changing the readout between the first frame and the second frame, but it is not limited to this, and there are frames in which the A signal and A+B signal are acquired independently and frames in which only the A+B signal is read out. Good to have. By driving in this way, it is possible to extract phase difference information only from frames that require it. On the other hand, when phase difference information is not required, the entire signal is extracted instead of pseudo one frame information. It becomes possible to obtain.

The ratio of frames for reading out the A signal described in FIGS. 8 and 9 may be made variable depending on the brightness of the subject and the amount of defocus of the photographing lens. Even in such driving, even in a configuration in which one AD converter is provided for one pixel, as in the first embodiment, a readout operation is realized in which the imaging signal and the phase signal are acquired individually as separate signals. It is possible to do so.

Next, based on FIG. 10, an embodiment in which the above-described image sensor is applied to a digital camera, which is an imaging device, will be described. In FIG. 10, an image sensor unit 1005 is a unit for capturing a subject imaged by the photographing lens 1001 as an image signal, and the image sensor shown in FIG. 1 described above is included in this image sensor unit 1005. The photographic lens 1001 forms an optical image of a subject on the image sensor 100 in the image sensor unit 1005.

A lens driving device 1002 performs zoom control, focus control, aperture control, etc. on the photographic lens 1001. Mechanical shutter 1003 is controlled by shutter drive device 1004. The image sensor unit 1005 includes the image sensor 100 as described above, and is also equipped with an optical low-pass filter and the like. The signal processing unit 1006 performs various corrections, data compression, etc. on the image signal output from the image sensor unit 1005. The signal processing unit 1006 can also perform synthesis, subtraction, correlation calculation, etc. of the above-mentioned A signal and A+B signal.

The timing generator 1007 outputs various timing signals to the image sensor 100 and the signal processor 1006. A control unit 1009 controls the entire imaging apparatus. The control unit 1009 can also perform various other calculations. Memory 1008 temporarily stores image data. The recording medium I/F 1010 is an interface for recording on or reading from the recording medium 1011. The recording medium 1011 is a removable recording medium such as a semiconductor memory that records image data. The display unit 1012 displays various information and photographed images.

Next, the operation of the digital camera in the above-described configuration during photographing will be explained. When the main power is turned on, the power of the control system is turned on, and further the power of the imaging system circuits such as the signal processing section 1006 is turned on. Thereafter, when a release button (not shown) is pressed, the signal processing unit 1006 performs a correlation calculation based on data from the image sensor.

The control unit 1009 drives the photographic lens 1001 by the lens driving device 1002 based on the result of the correlation calculation to determine whether or not the photographic lens 1001 is in focus. to perform correlation calculations. Note that the correlation calculation may be performed not only by the signal processing unit 1006 using data from the image sensor but also by a dedicated calculation device (not shown).

After the focus is confirmed, the photographing operation is started. When the photographing operation is completed, the signal processing unit 1006 performs image processing on the image signal output from the image sensor 100 to generate a photographed image. The data of the photographed image is written into the memory 1008 by the control unit 1009. The signal processing unit 1006 also performs rearrangement processing, addition processing, or selection processing thereof. The data stored in the memory 1008 is recorded on the recording medium 1011 via the recording medium I/F 1010 under the control of the control unit 1009. Note that at this time, the data may be transferred to an external device (such as a computer) via an external I/F unit (not shown), and the image may be processed by the external device.

(Other embodiments)
The present invention provides a system or device with a program that implements one or more functions of the embodiments described above via a network or a storage medium, and one or more processors in a computer of the system or device reads and executes the program. This can also be achieved by processing. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

The invention is not limited to the embodiments described above, and various changes and modifications can be made without departing from the spirit and scope of the invention. Therefore, the following claims are hereby appended to disclose the scope of the invention.

100: Image sensor, 104: Main control unit, 105: Pixel control unit, 106: Signal processing circuit

Claims (9)

A pixel unit in which a plurality of unit pixels are arranged in a matrix, each of the plurality of unit pixels having a plurality of photoelectric conversion elements that photoelectrically convert light passing through different exit pupils of a photographing lens. and,
a plurality of AD converters, one provided for each of the plurality of unit pixels;
a pixel control unit that controls readout from the plurality of unit pixels using the plurality of AD converters;
An image sensor having
The pixel control section includes:
In order to read out pixel signals of a frame, the frame is divided into a plurality of subframes and a readout operation is performed from the plurality of unit pixels for each subframe,
In the readout operation of a plurality of first subframes among the plurality of subframes, a signal from one photoelectric conversion element among the plurality of photoelectric conversion elements obtained in each of the plurality of first subframes; read out a first signal based on
In the readout operation of the plurality of second subframes among the plurality of subframes, the readout operation from all of the plurality of photoelectric conversion elements obtained in each of the plurality of second subframes is performed. reading a second signal based on the signal;
In the frame, readout is controlled so that the plurality of first subframes and the plurality of second subframes are alternately repeated, and in another frame different from the frame, the plurality of first subframes are An image sensor characterized in that readout is controlled so that the number of the plurality of second subframes is greater than the number of frames . The image sensor according to claim 1, wherein each of the plurality of photoelectric conversion elements is an avalanche photodiode that operates in an avalanche state. The image sensor according to claim 1, wherein each of the plurality of photoelectric conversion elements is a photodiode. The imaging according to any one of claims 1 to 3, wherein the pixel control unit generates a third signal as a phase signal by subtracting the first signal from the second signal. element. The image sensor according to any one of claims 1 to 4, wherein the other frame is composed of only the plurality of second subframes. The pixel control unit multiplies the first signal by a gain corresponding to a ratio occupied by the plurality of first subframes among the plurality of subframes, and 6. The image sensor according to claim 1, wherein the second signal is multiplied by a gain corresponding to a proportion of the second signal. The image sensor according to any one of claims 1 to 6 , wherein the plurality of AD converters are included in the pixel control section. comprising a first semiconductor chip and a second semiconductor chip stacked on each other,
8. The image sensor according to claim 1, wherein the pixel section is formed on the first semiconductor chip, and the pixel control section is formed on the second semiconductor chip. An image sensor according to any one of claims 1 to 8 ,
a signal processing unit that generates a captured image based on the signal output from the image sensor;
An imaging device comprising:

Images