Curved sensor ready for mass production

[Akira]

https://www.dpreview.com/news/7542036825/french-startup-is-preparing-its-curved-cmos-image-sensor-for-mass-production

I'm really interested but suspect that most of the existing lenses, old or current, would be unusable even if they can be easily adapted.  I also suspect that the focal plane shutter would be unusable.

[Airy]

Same here. That could give rise to some interesting new generation cameras, with relatively low mass and high quality, à la Leica Q (and hopefully cheaper). As you said, incompatibility with existing lenses might make this a niche product. "No aspherical elements" is not such an advantage since moulded aspherical lenses have become common (these were introduced in the eighties by Canon, among others).

[Akira]

Indeed.  The curved sensor may not necessarily be ideal for the interchangeable lens camera systems.  It might help smartphone cameras or action cameras to made thinner so that they can be flush with the smartphone bodies.

[Bill De Jager]

The concept is technically sound, but as manufacturing improves the curve radius may change to become more optically advantageous.  Think of the curve of the human retina as being one possible endpoint.

The problem here is that any change in the curve would require a new set of lenses.  That would make our current F- vs. Z-mount lens update situation look like a minor issue by comparison. So I see this type of sensor mainly being used in future fixed-lens cameras where the lens is matched to the sensor and optimized for that sensor in all respects.

[Hugh_3170]

An optically active adapter might be created in the future to map the focus plane of the legacy lenses onto the sensor. 

That said, such an adapter might also be able to map our current lenses onto current sensors.

(A bit of blue sky thinking from me I will confess.)

[Akira]

The concept is technically sound, but as manufacturing improves the curve radius may change to become more optically advantageous.  Think of the curve of the human retina as being one possible endpoint.

The problem here is that any change in the curve would require a new set of lenses.  That would make our current F- vs. Z-mount lens update situation look like a minor issue by comparison. So I see this type of sensor mainly being used in future fixed-lens cameras where the lens is matched to the sensor and optimized for that sensor in all respects.

In addition to the lens design, I suspect the control of both the in-body and the optical image stabilizations could be more tricky.  The more advanced digital image stabilization might solve the problem?

An optically active adapter might be created in the future to map the focus plane of the legacy lenses onto the sensor. 

That said, such an adapter might also be able to map our current lenses onto current sensors.

(A bit of blue sky thinking from me I will confess.)

That can be an idea.  But it also indicates that the curved sensor may open another can of worms for the interchangeable lens cameras...

[Nasos Kosmas]

One thing for sure lens will be less complex and that’s good