• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Samsung Unveils ISOCELL Image Sensor with Industry's Smallest 0.56μm Pixel

Joined
Nov 11, 2004
Messages
11,871 (1.83/day)
Location
Formosa
System Name Overlord Mk MXVI
Processor AMD Ryzen 7 5800X
Motherboard Gigabyte X570 Aorus Master
Cooling Corsair H115i Pro
Memory 32GB Viper Steel 3600 DDR4 @ 3800MHz 16-19-16-19-36
Video Card(s) Gigabyte RTX 2080 Gaming OC 8G
Storage 1TB WD Black NVMe (2018), 2TB Viper VPN100, 1TB WD Blue 3D NAND
Display(s) Acer XV272K LVbmiipruzx
Case Corsair Carbide 275Q
Audio Device(s) Corsair Virtuoso SE
Power Supply Corsair RM750
Mouse Logitech G502 Lightspeed
Keyboard Wooting Two
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/33u9si
Samsung Electronics Co., Ltd., a world leader in advanced semiconductor technology, today introduced the 200MP ISOCELL HP3, the image sensor with the industry's smallest 0.56-micrometer (μm)-pixels. "Samsung has continuously led the image sensor market trend through its technology leadership in high resolution sensors with the smallest pixels," said JoonSeo Yim, executive vice president of sensor business team at Samsung Electronics. "With our latest and upgraded 0.56μm 200MP ISOCELL HP3, Samsung will push on to deliver epic resolutions beyond professional levels for smartphone camera users."

Since its first 108MP image sensor roll-out in 2019, Samsung has been leading the trend of next-generation, ultra-high-resolution camera development. Through the steady launch of new image sensors and advancements in performance, the company is once again forging ahead with the 0.56μm 200MP ISOCELL HP3. The ISOCELL HP3, with a 12 percent smaller pixel size than the predecessor's 0.64μm, packs 200 million pixels in a 1/1.4" optical format, which is the diameter of the area that is captured through the camera lens. This means that the ISOCELL HP3 can enable an approximately 20 percent reduction in camera module surface area, allowing smartphone manufacturers to keep their premium devices slim.




The ISOCELL HP3 comes with a Super QPD auto-focusing solution, meaning that all of the sensor's pixels are equipped with auto-focusing capabilities. In addition, Super QPD uses a single lens over four-adjacent pixels to detect the phase differences in both horizontal and vertical directions. This paves way for a more accurate and quicker auto focusing for smartphone camera users.

The sensor also allows users to take videos in 8K at 30 frames-per-second (FPS) or 4K at 120fps, with minimal loss in the field of view when taking 8K videos. Combined with the Super QPD solution, users can take movie-like cinematic footage with their mobile devices.

Ultimate Low Light Experience through 'Tetra2pixel'
The ISOCELL HP3 also provides an ultimate low-light experience, with the Tetra2pixel technology that combines four pixels into one to transform the 0.56μm 200MP sensor into a 1.12μm 50MP sensor, or a 12.5MP sensor with 2.24μm-pixels by combining 16 pixels into one. The technology enables the sensor to simulate a large-sized pixel sensor to take brighter and more vibrant shots even in dimmed environments, like in-doors or during nighttime.

To maximize the dynamic range of the mobile image sensor, the ISOCELL HP3 adopts an improved Smart-ISO Pro feature. The technology merges image information made from the two conversion gains of Low and High ISO mode to create HDR images. The upgraded version of the technology comes with a triple ISO mode - Low, Mid, and High - that further widens the sensor's dynamic range. In addition, the improved Smart-ISO Pro enables the sensor to express images in over 4-trillion colors (14-bit color depth), 64 times more colors than the predecessor's 68 billion (12-bit). Furthermore, by supporting staggered HDR along with Smart-ISO Pro, the ISOCELL HP3 can switch between the two solutions depending on the filming environment to produce high-quality HDR images.

Samples of the Samsung ISOCELL HP3 are currently available, and mass production is set to begin this year.

View at TechPowerUp Main Site | Source
 
Joined
Sep 15, 2011
Messages
5,850 (1.47/day)
Processor Intel Core i7 3770k @ 4.3GHz
Motherboard Asus P8Z77-V LK
Memory 16GB(2x8) DDR3@2133MHz 1.5v Patriot
Video Card(s) ZOTAC GAMING GeForce RTX 3080 Trinity
Storage 59.63GB Samsung SSD 830 + 465.76 GB Samsung SSD 840 EVO + 2TB Hitachi + 300GB Velociraptor HDD
Display(s) Acer Predator X34 3440x1440@100Hz G-Sync
Case NZXT PHANTOM410-BK
Audio Device(s) Creative X-Fi Titanium PCIe
Power Supply Corsair 850W
Mouse Anker
Software Win 10 Pro - 64bit
Benchmark Scores 30FPS in NFS:Rivals
This is interesting. How does this compare to simmilar big sensors found in professional DLSR cameras?
 
Joined
Aug 13, 2010
Messages
4,977 (1.14/day)
This is interesting. How does this compare to simmilar big sensors found in professional DLSR cameras?
The marketing shot comes down to optics. Samsung was cheeky putting a giant lens on their sensor to produce the marketing behind their new sensor. They also probably collected the rawest output they could out of the sensor, and has equipped it with clean circuitry.
In reality, when dealing with such tight spaces as phones, your optics and circuitry are going to be far inferior in their ability.



If optics are completely out of the picture, which they shouldn't be, this sensor shouldn't be massively different than any other flagship phone sensor you will find today. The title boasts smallest pixel size for a phone sensor, which is sort of like boasting lower clocks for a proessor. Very backwards.

If your aim is well produced and post processed photos, typically the 1" type sensor category of cameras with proper sized lenses are still ahead of what sensor phones could do on phones with heavy in-device processing. Your dynamic range and optical characteristics will still look more natural, and are still much better resolved as raw data. We aren't going further in sensor size to Micro Four Thirds, APS-C and fullframe type since the argument still resolves itself at the 1" type sensor compact cameras of today.
 
Last edited:
Joined
Nov 11, 2004
Messages
11,871 (1.83/day)
Location
Formosa
System Name Overlord Mk MXVI
Processor AMD Ryzen 7 5800X
Motherboard Gigabyte X570 Aorus Master
Cooling Corsair H115i Pro
Memory 32GB Viper Steel 3600 DDR4 @ 3800MHz 16-19-16-19-36
Video Card(s) Gigabyte RTX 2080 Gaming OC 8G
Storage 1TB WD Black NVMe (2018), 2TB Viper VPN100, 1TB WD Blue 3D NAND
Display(s) Acer XV272K LVbmiipruzx
Case Corsair Carbide 275Q
Audio Device(s) Corsair Virtuoso SE
Power Supply Corsair RM750
Mouse Logitech G502 Lightspeed
Keyboard Wooting Two
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/33u9si
This is interesting. How does this compare to simmilar big sensors found in professional DLSR cameras?
Not at all? Pixel binning to 50 Megapixels should be decent though.
 
Joined
Mar 20, 2019
Messages
505 (0.41/day)
System Name Hermes
Processor 9600k
Motherboard MSI Z390I Gaming EDGE AC
Cooling Scythe Mugen 5
Memory 32GB of G.Skill Ripjaws V 3600MHz CL16
Video Card(s) MSI 3080 Ventus OC
Storage 2x Intel 660p 1TB
Display(s) Acer CG437KP
Case Streacom BC1 mini
Audio Device(s) Topping MX3
Power Supply Corsair RM750
Mouse R.A.T. DWS
Keyboard HAVIT KB487L / AKKO 3098 / Logitech G19
VR HMD HTC Vive
Benchmark Scores Shameful
This is interesting. How does this compare to simmilar big sensors found in professional DLSR cameras?
This is just marketing aimed at people who stopped at the "higher number is obviously better" phase. Those high density tiny sensors are just a noisy mess which needs a lot of processing just to be barely acceptable, not to mention binning to get anything at all from them. Also, 1/1.4" is not big. It's barely fifth of APS-C which is the small DSLR format and roughly 1/12 of full frame.
 
Joined
Jan 2, 2018
Messages
201 (0.12/day)
So i don't get it, why do we need these super small pixels?

The general rule of thumb, if you want better quality, is to search for phones with as big image sensor as possible and as big pixel size as possible.
What they do is exactly opposite.
Also why bother grouping pixels together if you can get best results when producing them bigger in the first place.
The grouping also involves a lot more processing, thus needing more powerful image processors.
 
Joined
Nov 11, 2004
Messages
11,871 (1.83/day)
Location
Formosa
System Name Overlord Mk MXVI
Processor AMD Ryzen 7 5800X
Motherboard Gigabyte X570 Aorus Master
Cooling Corsair H115i Pro
Memory 32GB Viper Steel 3600 DDR4 @ 3800MHz 16-19-16-19-36
Video Card(s) Gigabyte RTX 2080 Gaming OC 8G
Storage 1TB WD Black NVMe (2018), 2TB Viper VPN100, 1TB WD Blue 3D NAND
Display(s) Acer XV272K LVbmiipruzx
Case Corsair Carbide 275Q
Audio Device(s) Corsair Virtuoso SE
Power Supply Corsair RM750
Mouse Logitech G502 Lightspeed
Keyboard Wooting Two
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/33u9si
So i don't get it, why do we need these super small pixels?
Because most moder phones uses four pixels to create one pixel, as it says in the press release. It seems to work quite well and I'm on my second phone that does this now, although this time around, I don't have a choice to use all the pixels.
The general rule of thumb, if you want better quality, is to search for phones with as big image sensor as possible and as big pixel size as possible.
What they do is exactly opposite.
Also why bother grouping pixels together if you can get best results when producing them bigger in the first place.
The grouping also involves a lot more processing, thus needing more powerful image processors.
It's not the exact opposite though, at least not the way it's being used by most phone makers.
It's a different means to reach a similar result.
Because there are some advantages of pixel binning vs. having bigger pixels.
One of the main reasons for pixel binning is said to be lower image noise in low light situations.
Most phone SoCs handles that natively though, as they have dedicated co-processors for all of that stuff.
That said, some of the PRC based phone makers have developed their own chips for handling all of the image processing.
 
Top