“But there are — to be honest, the project has been a little bit delayed… because every time we run the wafer and get the prototype sensor, we found some technical issues because we have never designed a sensor by ourselves."
Foveon while a good idea, is extremely complicated to get right from what I understand.
I imagine that different silicon fabs or even different runs of the same fab have slightly different absorption on each focal plan in the sensor and calibrating that at high quality is tricky? Can you even do BSI? I have no idea
New Foveon is a long ways away still, they basically scraped what they had -
https://petapixel.com/2025/10/27/sigma-says-full-frame-foveo...
“But there are — to be honest, the project has been a little bit delayed… because every time we run the wafer and get the prototype sensor, we found some technical issues because we have never designed a sensor by ourselves."
They bought this technology in 2008. What have they been doing?!
Foveon while a good idea, is extremely complicated to get right from what I understand.
I imagine that different silicon fabs or even different runs of the same fab have slightly different absorption on each focal plan in the sensor and calibrating that at high quality is tricky? Can you even do BSI? I have no idea
This uses a partially-stacked bayer sensor, same as that in the Nikon Z6III and Panasonic S1 II (most likely, or very close to it).