Get ready to add a new term to your Android geek lingo, because HTC wants you to forget about megapixels and instead drool over the “ultrapixels” of its new M7 smartphone.
Ok, let’s back up a little. Megapixel literally means “million of pixels”, which is pretty straightforward. So, what’s HTC’s new ultrapixel? Are there more pixels in an ultrapixel than in an old-fashioned megapixel?
The new term actually refers to a type of sensor, which features three layers of pixels, each of a 4.3MP resolution. Using three overlaid sheets of pixels on the sensor will enable the M7 camera to take images that are crisper and more color accurate than the ones taken with conventional sensors. This is because the firmware will combine the information received by the three layers of pixels, to create one single virtual pixel. More information per pixel will also allow the camera to process the image in novel ways, such as focusing on a certain area of the image, after it was taken.
According to unnamed sources quoted by Pocket-lint, the sensor of the HTC M7 camera will somehow resemble the sensor that Nokia fitted on the PureView 808 and the Foveon X3 sensor on some Sigma cameras.
So, why did HTC felt compelled to coin a new term for the M7 camera sensor? Because the actual resolution of the images snapped with the new flagship will supposedly be just 4.3MP. Just like on the Nokia PureView 808, the camera will be able to snap photos at higher resolutions, but only in a special mode.
It’s understandable why HTC would be uneasy about touting a camera resolution of just 4.3MP. According to the report, the ultrapixel sensor would be central to HTC’s marketing for the M7, just like the ImageSense software was a highlight of the marketing campaign for the One X.
A caveat: this is, until further proof, just a rumor. Earlier reports about the M7 all suggested that HTC’s new hero device would come with a conventional 13MP sensor. But one could say that 3 x 4.3MP is approximately 13MP, so there might be some truth in both accounts.
The M7 is due on February 19, so we don’t have to wait too long until the mystery surrounding the ultrapixel is cleared. Stay tuned for mover coverage.
Like this post? Share it!
Oh great, I see the old argument looming again. So in a Bayer sensor, the red, green and blue components are spread spatially in the XY plane of the image, and therefore must be interpolated by a demosaicing algorithm to constitute (R,G,B) triples at a variety of locations totalling the quoted number of megapixels; in a Foveon sensor, you have as few locations and each layer in the Z-axis absorbs some of the incoming photons so you don’t have the same sensitivity in the second and third layers as in the top – e.g. your greens (and therefore luminosity) will be smooth but your reds and blues will be even noisier than with Bayer. And if you want a larger image you’ve lost the resolution before upscaling it.
With Foveon-style layering technology, it’ll be 4.3MP whether they call it that or 13.
Thanks Tim for the technical explanation!
So by having to merge the three layers together, does it mean it won’t be instant capture?
And can I resolution be higher than 4.3? I don’t really care much but just wondering.
All I know is that the Exmor RSV2 sensor in the Oppo Find 5 shoots stunning pictures, both in pixel sharpness and dynamic range, that includes night shots, and that is more than enough for my needs. Samples can be googled. Really impressive.
Yay, 2GB photo’s, sounds great!
It’s not going to be that sharp! :D
Pureview has a 43 MP sensor :)
4.3 MP is *plenty* of pixels. 1600×1200 images, suitable for printing at 5.3″ x 4″. Plenty big for uploading to Facebook or Instagram.
I’d be delighted to have a *better* camera sensor at the cost of reduced megapixel counts.
An explanation of the Foven system: http://www.foveon.com/article.php?a=67