A better CMOS image sensor for ultra-high definition television

Ultra-high definition television (UHDTV), which promises a clearer and more aesthetic viewing experience than is currently possible, has attracted a lot of attention in recent years. International standards for the technology are being established by the ITU-R (the radiocommunication sector of the International Telecommunication Union)1 and industry standards by SMPTE (Society of Motion Picture and Television Engineers),2 all aimed at making UHDTV a reality. For its part, NHK (Japan Broadcasting Corporation) has researched and developed an “8K Super Hi-Vision” (SHV) system featuring 8K resolution and 22.2 multi-channel sound with the aim of giving the audience the feeling to be “there”. Table 1 shows the video parameter values ​​for SHV. These parameters comply with international standards.

Table 1. SHV settings. H: horizontal. V: Vertical. RGB: Red, Green, Blue.

number of pixels 7680(H)×4320(V)
Frame rate 120Hz, 60Hz
Scanning progressive
Sampling structure 4:4:4, 4:2:2, 4:2:0
Restoration of tones 12 bit, 10 bit
Color rendering Wide Gamut RGB

We are currently developing an image sensor (see Figure 1) for a “complete” SHV camera system. In the full SHV system, the video signal is characterized by 7680 (H)×4320 (V) pixels (about 33Mpixel per frame) with a 4:4:4 (red, green, blue) RGB sampling structure, 120fps ( frames per second) progressive scan, 12-bit tone reproduction and wide color gamut. To achieve a complete VHS, we designed and fabricated CMOS image sensors and prototyped a color camera system.3–5

The image sensor in this system is capable of shooting 33 Mpixel, 120 fps movies with 12-bit ADC (analog-to-digital converter) resolution. The so-called column-parallel ADC architecture (see Figure 2) – a common CMOS image sensor technology – and our newly developed two-stage cyclic ADC enable high-speed operation and power consumption as low as 2 .5 W, simultaneously. In the two-stage cyclic ADC, the first and second cyclic ADCs operate in a pipelined configuration (see Figure 3). However, due to its smaller pixel area and a quarter of the exposure time, our prototype three-chip color camera is not as sensitive as existing HDTV cameras, which typically feature 1920×1080 pixels and 30 frames per second. Therefore, we place a high priority on improving sensitivity.

This requires either increasing the output signal (voltage) of each pixel per incoming light or reducing the noise generated in the sensor (i.e. improving the signal to noise or S/N ratio) . We decided to increase the output signal of the pixels (ie their sensitivity) thanks to nanofabrication. Pixel sensitivity is determined by multiplying “quantum efficiency” by “conversion gain”. Quantum efficiency refers to the rate at which photons incident on a pixel area are converted into electrons. The conversion gain represents the amount of voltage generated by a single electron in the floating diffusion amplifier (FDA) in the pixel. Nanofabrication reduces the capacitance of the ADF, thus increasing the conversion gain, which is inversely proportional to the capacitance. We used a manufacturing process adapted to a 0.11µm CMOS image sensor instead of 0.18µm, as for the previous device.6

The main difference between the two sensors made with the 0.11 and 0.18µm design rules was the pixel structure, in particular the FDA capacity. Figure 1 shows the fabricated image sensor and Table 2 lists its specifications. This sensor exhibited a conversion gain of 110µV/e and a sensitivity of 2.4V/lx·s. The conversion gain matched our design value and the pixel sensitivity was increased by 1.6 times. As a result, the random noise referred to the input, which gives an indication of the sensor’s S/N ratio, has been reduced from the previous sensor’s 3.0e.rms at 2.1erms.

Table 2. Manufactured image sensor specifications. CIS: CMOS image sensor. CIE: International Commission on Illumination.

Process 0.11µm 1P4M IEC
Number of active pixels 7680(H)×4320(V)
Total number of pixels 7836(H)×4372(V)
Pixel spacing 2.8×2.8µm
Frame rate 120Hz (max)
conversion gain 110µV/e
Sensitivity 2.4 V/lx s (CIE A-light, IR cut filter)
Random noise 2.1erms (at 120Hz and gain=15)

These experimental results confirmed the effect of nanofabrication on the conversion gain and therefore on the sensitivity of the pixels. Because further increase in conversion gain could degrade dynamic range, we plan to increase quantum efficiency by optimizing pixel design. Refining the design and installing low profile wiring would increase the aperture ratio and increase sensitivity. We continue to research additional ways to improve the sensitivity of the image sensor for practical use of the full VHS.

This image sensor is developed through a collaboration between NHK and Shizuoka University in Japan.

Toshio Yasue


Tokyo, Japan

Toshio Yasue has done research on CMOS image sensors and camera systems at NHK Science and Technology Research Laboratories. He obtained his ME in Applied Physics from the University of Tokyo in 2008.


1. Parameter values ​​for UHDTV systems for production and international program exchange, Technology. ITU-R Rec. BT2020International Telecommunication Union, Geneva.

2. Ultra-high definition television—image parameter values ​​for program production, Technology. Rep. SMPTE ST 2036-1:2013 Society of Motion Picture and Television Engineers, 2013.

3. T. Watabe, K. Kitamura, T. Sawamoto, T. Kosugi, T. Akahori, T. Lida, K. Isobe, et al., A 33Mpixel 120fps CMOS image sensor using a 12b parallel pipelined cyclic ADC columns, ISSCC search. Technology. Papers, p. 388-389, 2012.

4. K. Kitamura, T. Watabe, T. Sawamoto, T. Kosugi, T. Akahori, T. Lida, K. Isobe, A 33 megapixel 120 frames per second 2.5 watt CMOS image sensor with column-converters parallel two-stage cyclic analog-digital, et al., IEEE Trans. Electron. Devices 59(12), p. 3426-3433, 2012.

5. H. Shimamoto, K. Kitamura, T. Watabe, H. Ohtake, N. Egami, Y. Kusakabe, Y. Nishida, et al., Super High Vision Frame Rate Capture and Display Devices of 120Hz, SMPTEMot. picture J 122(2), p. 55-61, 2013.

6. T. Yasue, T. Hayashida, J. Yonai, K. Kitamura, T. Watabe, H. Ootake, H. Shimamoto, et al., A 33 megapixel 120 fps CMOS image sensor using a CIS process of 0.11 μm, proc. SPIE 9100, 2014. doi:10.1117/12.2049145

Michael C. Garrison