4 ms·

Physics are a limitation but not a blocker. The only major limitation imho is focal length. There's only so much optical zoom you can fit into the smartphone fo

Physics are a limitation but not a blocker. The only major limitation imho is focal length. There's only so much optical zoom you can fit into the smartphone form factor. AI post processing will probably make digital "zoom" much less of an issue in the near future. I'm an amateur photographer. I do mostly street and architecture with a bit of landscape. Right now my primary cameras are an ASP-C DSLR, a Sony RX100, as well as a selection of 35mm, MF, and instant film cameras. All of these bring different strengths to the table but my DSLR and RX100 stay at home more often than not these days. The in-camera processing just can't compare to smartphones. The fact that it is possible to apply the similar processing to DSLR or full-frame mirrorless images doesn't make those cameras any less expensive, heavy, or bulky. For most people, those are the points that matter. The destination of most images is social media more often than 8x10s, the ever shrinking quality advantages offered by "proper" cameras are outweighed by the convenience of having something more than adequate in their pocket at all times. After all, as the cliched advice goes: "The best camera is the one you have with you.", Or for old film photojournalists: "F8 and be there."

mrandish 4y ago
I agree with everything you've said. For me, there are a few use cases where the capabilities offered by an ILC are required. At the moment my primary "ILC-required" scenario is shooting my teenage daughter's competitive volleyball games. Sports is often a uniquely challenging photography use case and indoor volleyball is known as being especially challenging due to the reduced lighting indoors, unpredictable high-speed action over a large area and the considerable yet constantly changing distance to the camera. I have a top of the line smartphone camera and know how to use it extremely well, yet it's virtually impossible to get consistently good shots in this scenario because it simultaneously presents real-time challenges to magnification, dynamic tracking auto-focus, shutter speed, rapid sequential multi-frame capture and low light resolving. I found it hard even with my mid-range Canon full frame DSLR but I recently upgraded to a Sony A1 and it has made a huge difference (50 megapixel lossless compressed raw at >20 fps with perfect AI eye-tracking auto-focus (through a foreground mesh net!) while at f2.8, 1/500th, ISO 320 and 200mm zoom is remarkable). And the 50mp resolution provides plenty of headroom for me to stay relatively loose yet still crop in for framing in post. When the ball is coming I press the shutter button and hold it down. Three seconds later I have 75+ perfectly focused crop-ready HDR images of the correct subject's face (out of several in frame) and it remains tracked frame-by-frame while the face changes distance to the camera by >12 feet during those three seconds. Smartphone cameras are amazing but certain use cases like this are going to remain beyond them for some time still. Despite claiming "up to 100mp resolution" no smartphone camera I've seen provides images that actually contain enough useful image data to be cropped in while every wildly moving eyelash remains in crystal sharp focus and the shade of my girl's blue eyes is not only visible but correct from >75 feet away. The A1's auto-focused 50 megapixels are qualitatively different than even the best smartphone. The A1's pixels remain correct and unblurred even in low-light, even zoomed, even with a constantly moving, high-speed subject in a complex changing scene. A good smartphone can do any one of those things given ideal conditions. It's the "all-the-hard-things-at-once" part where smartphones still struggle. Gathering this much scene detail this quickly at this accuracy and fidelity may not be physically possible through a hole-punch sized opening into a mechanism which is less than 10mm deep weighing only a few grams and requiring only millivolts of power. The 2 pounds worth of $2,000 ground glass with a photon-gathering opening 100x larger and an imager surface area >25x larger is an enabler at the fundamental physics level. No matter how good the tech gets (and I'm sure it will continue to improve dramatically), there simply may not be enough photons coming in quickly enough through that small of a hole onto that small of an imager area. To be clear, I'm not a photography snob. The minute a pocketable smartphone (at any price) can consistently get the photos I need of my kid's tournament games, I'll sell my bulky, heavy, expensive ILC.

mrandish 4y ago

I agree with everything you've said. For me, there are a few use cases where the capabilities offered by an ILC are required. At the moment my primary "ILC-required" scenario is shooting my teenage daughter's competitive volleyball games. Sports is often a uniquely challenging photography use case and indoor volleyball is known as being especially challenging due to the reduced lighting indoors, unpredictable high-speed action over a large area and the considerable yet constantly changing distance to the camera. I have a top of the line smartphone camera and know how to use it extremely well, yet it's virtually impossible to get consistently good shots in this scenario because it simultaneously presents real-time challenges to magnification, dynamic tracking auto-focus, shutter speed, rapid sequential multi-frame capture and low light resolving. I found it hard even with my mid-range Canon full frame DSLR but I recently upgraded to a Sony A1 and it has made a huge difference (50 megapixel lossless compressed raw at >20 fps with perfect AI eye-tracking auto-focus (through a foreground mesh net!) while at f2.8, 1/500th, ISO 320 and 200mm zoom is remarkable). And the 50mp resolution provides plenty of headroom for me to stay relatively loose yet still crop in for framing in post. When the ball is coming I press the shutter button and hold it down. Three seconds later I have 75+ perfectly focused crop-ready HDR images of the correct subject's face (out of several in frame) and it remains tracked frame-by-frame while the face changes distance to the camera by >12 feet during those three seconds. Smartphone cameras are amazing but certain use cases like this are going to remain beyond them for some time still. Despite claiming "up to 100mp resolution" no smartphone camera I've seen provides images that actually contain enough useful image data to be cropped in while every wildly moving eyelash remains in crystal sharp focus and the shade of my girl's blue eyes is not only visible but correct from >75 feet away. The A1's auto-focused 50 megapixels are qualitatively different than even the best smartphone. The A1's pixels remain correct and unblurred even in low-light, even zoomed, even with a constantly moving, high-speed subject in a complex changing scene. A good smartphone can do any one of those things given ideal conditions. It's the "all-the-hard-things-at-once" part where smartphones still struggle. Gathering this much scene detail this quickly at this accuracy and fidelity may not be physically possible through a hole-punch sized opening into a mechanism which is less than 10mm deep weighing only a few grams and requiring only millivolts of power. The 2 pounds worth of $2,000 ground glass with a photon-gathering opening 100x larger and an imager surface area >25x larger is an enabler at the fundamental physics level. No matter how good the tech gets (and I'm sure it will continue to improve dramatically), there simply may not be enough photons coming in quickly enough through that small of a hole onto that small of an imager area. To be clear, I'm not a photography snob. The minute a pocketable smartphone (at any price) can consistently get the photos I need of my kid's tournament games, I'll sell my bulky, heavy, expensive ILC.