Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Thursday January 18 2018, @11:21AM   Printer-friendly
from the can-you-picture-that? dept.

A couple years ago, Hasselblad released a 200-megapixel, Multi-Shot version of its H5D medium format camera. Now it's back with a bonkers, 400-megapixel version of the H6D: the H6D-400c.

Hasselblad's Multi-Shot technology is pretty straightforward: it takes four 100-megapixel images, shifting the sensor by one pixel for each capture, and then two more shots that shift the sensor by half a pixel. By combining all six stills, the resulting file is a single 400-megapixel (23200 x 17400 pixel) 16-bit TIFF file that weighs in at 2.4GB. In fact, the images are large enough that the camera needs to be tethered to a computer to capture them.

[...] The camera will go for $47,995 when it launches in March, compared to the H6D-100c's relatively modest $27,000 price tag.

Story at The Verge.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 0) by Anonymous Coward on Thursday January 18 2018, @03:05PM (2 children)

    by Anonymous Coward on Thursday January 18 2018, @03:05PM (#624157)

    They're tethering it by need:

    With an effective resolution of 400MP via 6 shot image capture, or 100MP resolution in either 4 shot Multi-Shot capture or single shot mode, the Multi-Shot capture requires the sensor and its mount to be moved at a high-precision of 1 or ½ a pixel at a time via a piezo unit. To capture Multi-Shot images the camera must be tethered to a PC or MAC.

    You need a piezo stage driver and controller coordinated with the frame capture. I suppose you could build all of that into the camera body, but you'd probably want a hand truck to move it around with you.

  • (Score: 2) by KilroySmith on Thursday January 18 2018, @03:15PM

    by KilroySmith (2113) on Thursday January 18 2018, @03:15PM (#624160)

    They're generating 2.4GB images. To deal with them internally reasonably fast, they'd have to buffer the data to RAM, while in parallel writing it to FLASH. So, they'd need a storage subsystem consisting of a microprocessor, 2.4GB of RAM, and a USB/SD/CF interface. Hate to tell you, but that's a package that fits in an area of less than 1 sq in and a couple of millimeters thick these days. The specs aren't that different from the processing complex in a modern cell phone. There'd be no need for a hand truck.

  • (Score: 4, Interesting) by Kromagv0 on Thursday January 18 2018, @04:23PM

    by Kromagv0 (1825) on Thursday January 18 2018, @04:23PM (#624187) Homepage

    All of that equipment is built into cameras that are already available and even a bunch that don't cost more than my car. While there was the previous 200MP camera from Hasselblad, there are others you can buy now that aren't even a flagship camera. I would suggest looking at the Pentax K-1 (flagship DSLR), K-3II, KP, and K-70(entry level DSLR) as they all do pixel shift. However they don't do it to increase the resolution but instead as a method to eliminate the demosaicing process [wikipedia.org] normally needed when working with a sensor wit a bayer pattern so that each pixel has a real RGB set of readings instead of interpolated. The added color depth as well as higher clarity on edges from the Pentax cameras makes up for the lack of increase in resolution in the images I have seen but if one wanted to they could just do demosaicing on the pixel shifted Pentax image afterwards and get a similar increase in resolution likely with better results as you would likely have access to better algorithms that don't need to meet an embedded system's power and time budget.

    --
    T-Shirts and bumper stickers [zazzle.com] to offend someone