2 ms·

This is perfect for Apple since good image sensor design requires a small highly-skilled team. It's very different from, say, cellular modems.

by wmf 3y ago

sudosysgen 3y ago
Is that really true? Isn't Sony I&SS a pretty huge division, with over 10'000 employees? From my understanding, sensor development is highly tied into process development, unless you do fairly minor implementation requests to Sony. But so far, neither Canon nor Nikon have been especially successful speccing out significantly better sensors from Sony than are available to anyone else (and Canon's in-house sensors are pretty much uniformly inferior). But who knows, maybe Apple can pull off something neither Nikon, Canon nor Samsung have been able to.
- wmf 3y ago
  Historically, small fabless companies like FillFactory and Red have been able to do pretty well. Apple's goal may be reducing Sony's margin rather than trying to beat their quality.
  - sudosysgen 3y ago
    Source on RED making their own sensors? Word on the street is that they essentially use minor tweaks of Canon sensors nowadays (the Komodo sensor is suspiciously similar to an APS-C crop of the R5, and of course the highly proprietary RF mount is a strong hint to close ties). FillFactory was successful, but that was 20 years ago. Sony is really dominant now in a way it never was before.
SavageBeast 3y ago
Someone on the Qualcomm modem team should comment here - No clue about making image sensors but having spent some significant time with just such an engineer, the modem team people face some fairly absurd challenges. Their world seemed to be a vast collection of edge/corner/special cases to me. But they were all real world cases that were liable to happen all the time. When is a good 3g signal superior to a bad 4g signal? On and on and on. Making an image sensor at least naively seems like a smaller problem surface.

This is perfect for Apple since good image sensor design requires a small highly-skilled team. It's very different from, say, cellular modems.

by wmf 3y ago

This is perfect for Apple since good image sensor design requires a small highly-skilled team. It's very different from, say, cellular modems.

sudosysgen 3y ago

Is that really true? Isn't Sony I&SS a pretty huge division, with over 10'000 employees? From my understanding, sensor development is highly tied into process development, unless you do fairly minor implementation requests to Sony. But so far, neither Canon nor Nikon have been especially successful speccing out significantly better sensors from Sony than are available to anyone else (and Canon's in-house sensors are pretty much uniformly inferior). But who knows, maybe Apple can pull off something neither Nikon, Canon nor Samsung have been able to.

wmf 3y ago
Historically, small fabless companies like FillFactory and Red have been able to do pretty well. Apple's goal may be reducing Sony's margin rather than trying to beat their quality.
- sudosysgen 3y ago
  Source on RED making their own sensors? Word on the street is that they essentially use minor tweaks of Canon sensors nowadays (the Komodo sensor is suspiciously similar to an APS-C crop of the R5, and of course the highly proprietary RF mount is a strong hint to close ties). FillFactory was successful, but that was 20 years ago. Sony is really dominant now in a way it never was before.

SavageBeast 3y ago

Someone on the Qualcomm modem team should comment here - No clue about making image sensors but having spent some significant time with just such an engineer, the modem team people face some fairly absurd challenges. Their world seemed to be a vast collection of edge/corner/special cases to me. But they were all real world cases that were liable to happen all the time. When is a good 3g signal superior to a bad 4g signal? On and on and on. Making an image sensor at least naively seems like a smaller problem surface.