Samsung began working on 200MP sensors in 2021. Development and design of the HP2 started a long time ago but was unfortunately not used for the S22 Ultra.
After using the same 1/1.33″ 108MP sensor on their flagship phones, Samsung is finally moving to a slightly larger, higher-resolution sensor in 2023.
Update February 1: Samsung Galaxy S23 Ultra is now official. The handset features a 200MP ISOCELL HP2 sensor and comes with advanced pixel-binning technology, Tetra pixel. The new tech produce details and sharp images in different lighting conditions.
You can watch Danny Winget’s “Samsung Galaxy S23 Ultra 200MP Camera Tour” video on youtube.
They announced the Samsung ISOCELL HP2 sensor on 17th January 2023. The company made vast claims of significant improvements to dynamic range and enhanced colour reproduction.
ISOCELL HP2 200MP Sensor: How does it stack up?
The new HP2 sensor is often confused with the HPX and HP3 sensors. Samsung’s sensor naming scheme is incredibly confusing. The HP3 and HPX are 200MP sensors, and neither is as good as the newly announced HP2 sensor.
The HPX and HP3 are 200 MP 1/1.4″ sensors, and there’s not much that differentiates the two. However, the HP2 has some significant benefits over those two other sensors. The HPX was used on a $400 midrange phone, so it’s evident that the flagship HP2 is better.
The HP2 sensor is 1/1.3″ in size, bigger than the 1/1.33″ 108MP sensor on the S22, S21, and S20 Ultras. The HP2 is still slightly smaller than the 48MP 1/1.28″ Sony IMX 803 sensor on the iPhone 14 Pro/ Max.
The new sensor can do four-in-one Pixel Binning for a 50MP (1.2 μm Pixels) output or sixteen-to-1 Pixel Binning for a 12.5MP (2.4 μm Pixels) output.
The sensor is capable of 8K videos with four-in-one binning at 30FPS. It’s still capable of 4K videos at 60FPS with OIS+EIS.
The sensor can also do 15FPS bursts at 200MP, which is better than the HPX and HP3. Also, individual pixels now gather more light before closing up. To achieve this, Samsung uses Dual Vertical Transfer Gate (DVTG) technology.
This results in better low-light photos. It also massively improves highlight control and reduces the chances of blown-out lights or bright skies. Bright areas of your photos will not overexpose.
Autofocus and shutter speeds were significant complaints with the 108MP sensor on the S22 Ultra. To improve autofocus speeds, Samsung uses QPD Technology to recognize horizontal and vertical patterns in photos.
This results in faster, snappier, and more accurate autofocus across several lighting conditions. In 50MP mode, Samsung is using Dual Slope Gain Tech for capturing multiple exposures at once, improving HDR performance; so shadow areas won’t be too dark.
It applies two different conversion values to analogue signals. And that results in multiple exposures blended correctly.
To simplify, it’s a more advanced signal transfer method that avoids overexposure and underexposure. “Smart ISO Pro” merges multiple ISO readouts from a single exposure value.
Galaxy S23 Ultra Camera Sample Surfaces
The Samsung ISOCELL HP2 has entered mass production and will debut with the S23 Ultra. Leakster Ice Universe has also shared camera samples of Galaxy S23 Ultra. The camera sample shows significant improvement over S22 Ultra in the camera department.
To conclude, we can expect faster autofocus speeds, better shutter speeds, more reliable colour reproduction, and much better dynamic range performance.
Thanks to the slightly larger sensor, we will also see a better natural field depth, much sharper details, and more details in low-light shooting conditions. We’ll also see slightly more details thanks to the higher 200MP resolution.
Samsung is competing with new technologies for better HDR, but the sensor size is still smaller than most competitors. The iPhone’s sensor is still slightly larger, but it does not matter since the dynamic range performance is abysmal.
Xiaomi’s 12S Ultra and Vivo’s 1″ sensors are far larger than the 1/1.3″ size. The S23 Ultra might fall behind in low-light performance. However, Samsung can still beat the competition with computational processing.