Samsung Debuts ISOCELL Sensor Tech, Promises up to 30% More Dynamic Range


Although the pixel war probably isn’t ending anytime soon, a new sensor technology from Samsung shows how yet another company is focusing on improving the tech instead of stacking the spec sheet.

We’ve seen amazing low-light sensors and dual-pixel AF tech from Canon, organic sensors with insane dynamic range from Fuji and Panasonic, and now new ISOCELL technology from Samsung, which promises substantial increases in color and light sensitivity.

The technology was announced yesterday through Samsung Tomorrow, and will supposedly appear in “premium mobile devices” before long. According to the announcement, the advancement is based around some fancy pixel technology that will increase light sensitivity and produce better colors even in low light.

Here’s a comparison between a photo taken with a typical BSI sensor and one taken with an ISOCELL sensor:


The technical details behind the improvement have to do with physical barriers between each pixel. As pixel size has continued to shrink so that companies can pack ever-more megapixels into ever-smaller devices, a side-effect has been an increase in “crosstalk” between pixels (among other problems).

ISOCELL technology addresses this crosstalk by creating a barrier between pixels, thus allowing more light to be focused on the correct pixel. Crosstalk between pixels is reduced by 30%, which translates into higher color fidelity and an increase in dynamic range of up to 30% as well.

These sensors are also thinner than their non-ISOCELL counterparts, which Samsung explains will make them perfect for the thinner and sleeker smartphones of the future.

You can find out more about the technology on the official announcement page. The first iteration ISOCELL sensor features 8 megapixels of resolution, and is set to go into mass production in Q4 of this year.

(via Engadget)

  • tomdavidsonjr

    I wonder how much of this “30%” boost in dynamic range is actually mathematical manipulation under ideal circumstances. Marketing people have this sleazy tendency to play fast and loose with numbers that engineers give them (I have seen it in my own line of work, first hand). I am guessing that there is probably a numerical difference in the crosstalk between pixels, but that it does not ACTUALLY produce a 30% increase in dynamic range except perhaps under laboratory conditions. That would be (if I am doing the math correctly) like jumping from 6 stops to 9 stops. It is highly unlikely that they can achieve the kind of dynamic range increase that one gets moving (for example) from an APS-C sensor to a full frame sensor. Larger pixels gather more light, with less noise – and adding 30% more dynamic range to an entire sensor simply by shielding each pixel from crosstalk seems like a bit of a stretch. I don’t trust the numbers here, I’m sorry.

  • Mike

    Well, if the crosstalk they eliminated was taking away a potential ~30% making the highlights clip too soon, then now we have these same 30% back.

  • Mike

    Also- does that mean we can ETTR better now? :D

  • 42323423432

    i dont see more details.. just blown yellow insead of blown white.
    im with tomdavidsonjr here…. too much PR talk in the sensor biz.
    in the end it´s just a marginal increase…blown out of proportion to sell a few more samsung cameras to the clueless.

  • log vs linear scale

    Jumping from 6 stops to 9 corresponds to 8-fold increase in DR, 30% DR improvement corresponds to about 0.4 stops (linear DR improvement converted to log scale).

  • Mangap

    I want to see good low light photos. now many new sensor targeting on better low light photos