Sensor evolution stuck?

Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
The promise of smartphone-style global shutter sensors is that we'll be able to take and combine ten seamless captures instead of one longer one. Each of the ten will be under-exposed, but the sum can exceed the full-well capacity, thereby exceeding the sensor's native dynamic range.

This requires a "blackout-free" global shutter, which AFAIK only exists for smaller sensors at the moment. But at least it doesn't present a computational challenge, since we're just interested in the straight sum of the images, without any alignment/stacking shenanigans.
 
Does it need to evolve anymore? film was good in 70s and you could have used the same stock in the 1990s.
Agreed. Kodachrome for still photography was introduced in the 1930s, Kodachrome II and X in the 1960s, Kodachrome 25 and 64 in the 1970s and discontinued in the 2000s. That's three generations in around 70 years.

--
Dave, HCL
 
Last edited:
Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
The promise of smartphone-style global shutter sensors is that we'll be able to take and combine ten seamless captures instead of one longer one. Each of the ten will be under-exposed, but the sum can exceed the full-well capacity, thereby exceeding the sensor's native dynamic range.

This requires a "blackout-free" global shutter, which AFAIK only exists for smaller sensors at the moment. But at least it doesn't present a computational challenge, since we're just interested in the straight sum of the images, without any alignment/stacking shenanigans.
Do you have any link to a document or presentation about something like this? I have no idea when that will be feasible, but it would be interesting to know how the implementation would work.
 
Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
The promise of smartphone-style global shutter sensors is that we'll be able to take and combine ten seamless captures instead of one longer one. Each of the ten will be under-exposed, but the sum can exceed the full-well capacity, thereby exceeding the sensor's native dynamic range.

This requires a "blackout-free" global shutter, which AFAIK only exists for smaller sensors at the moment. But at least it doesn't present a computational challenge, since we're just interested in the straight sum of the images, without any alignment/stacking shenanigans.
I guess the original post might have been a bit broad, but I'm mostly talking about camera sensors (as in interchangeable and integrated lens cameras) mostly.
 
Hi all, I am a photographer With more than 30 years of expericene (mostly Nikons, NikkortMat, F90, F100, F3, D700, D750... but also some medium format). I have started to consider upgrading my equipment but when I see some wonderful cameras such as Fuji XT5 OR OM3 and I take a look at these amazing reviews and check the image comparisson pages, I see that the quality is not better than the quality that I get from the >10 years old sensor that I got in the D750....so, are new cameras just focussing on design and ergonomics? Is sensor technology evolving at a lower pace?

Just my thouthgs and my first post on this great site!

Thanks!,
Absolutely. As far as I can tell the IQ of cameras isn't any better now than cameras from 10 years ago. If it is the difference is so small that it takes close examination to see it. What has improved is on sensor PDAF auto focus and tracking which has improved a lot. Sensor read out speed has also advanced quite a bit. So there have been improvements in sensors that have improved the speed and usability of cameras but not the image quality.

--
Tom
 
Last edited:
I think the curve is starting to flatten out in terms of major innovations. Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
Not the way Sony sensors are engineered to achieve global instantaneous readout. The architecture required to store signal prior to readout occupies area that could otherwise be used for light-gathering, hence the reduced dynamic range of the A9III sensor.
Supposedly Apple is developing (or has developed) a sensor that has some 20-stops of DR, but not sure how long it will be until that makes its way (20-stops of DR) into dedicated digital camera like ones that Sony/Canon/Nikon make). Might be another 5 years or so before that is done if it is.
It's tech that's been around for 20+ years. The impact is had in consumer digital imaging systems has been the development of dual-gain sensors...not insignificant but also not exactly game-changing.
We'll see if Apple actually has a practical application in mind.
 
Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
The promise of smartphone-style global shutter sensors is that we'll be able to take and combine ten seamless captures instead of one longer one. Each of the ten will be under-exposed, but the sum can exceed the full-well capacity, thereby exceeding the sensor's native dynamic range.

This requires a "blackout-free" global shutter, which AFAIK only exists for smaller sensors at the moment. But at least it doesn't present a computational challenge, since we're just interested in the straight sum of the images, without any alignment/stacking shenanigans.
I guess the original post might have been a bit broad, but I'm mostly talking about camera sensors (as in interchangeable and integrated lens cameras) mostly.
There's nothing inherently tied to sensor size about this. But since the feature isn't available on the A9, it perhaps can't quite do gapless captures yet at the required high frame rate? Of course it's also debatable how many people really need more dynamic range than what today's large sensors provide.
 
I think the curve is starting to flatten out in terms of major innovations. Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
Not the way Sony sensors are engineered to achieve global instantaneous readout. The architecture required to store signal prior to readout occupies area that could otherwise be used for light-gathering, hence the reduced dynamic range of the A9III sensor.
Supposedly Apple is developing (or has developed) a sensor that has some 20-stops of DR, but not sure how long it will be until that makes its way (20-stops of DR) into dedicated digital camera like ones that Sony/Canon/Nikon make). Might be another 5 years or so before that is done if it is.
It's tech that's been around for 20+ years. The impact is had in consumer digital imaging systems has been the development of dual-gain sensors...not insignificant but also not exactly game-changing.

We'll see if Apple actually has a practical application in mind.
The Apple patent I mentioned in my earlier post is dated July 2, 2025.
 
Last edited:
I think the curve is starting to flatten out in terms of major innovations. Of course Sony with the global shutter is the next logical step, although this doesn't necessarily push the IQ boundary forward necessarily.
Not the way Sony sensors are engineered to achieve global instantaneous readout. The architecture required to store signal prior to readout occupies area that could otherwise be used for light-gathering, hence the reduced dynamic range of the A9III sensor.
Supposedly Apple is developing (or has developed) a sensor that has some 20-stops of DR, but not sure how long it will be until that makes its way (20-stops of DR) into dedicated digital camera like ones that Sony/Canon/Nikon make). Might be another 5 years or so before that is done if it is.
It's tech that's been around for 20+ years. The impact is had in consumer digital imaging systems has been the development of dual-gain sensors...not insignificant but also not exactly game-changing.

We'll see if Apple actually has a practical application in mind.
The Apple patent I mentioned in my earlier post is dated July 2, 2025.
This will explain things: https://www.dpreview.com/forums/post/68332406
 
There's nothing inherently tied to sensor size about this. But since the feature isn't available on the A9, it perhaps can't quite do gapless captures yet at the required high frame rate? Of course it's also debatable how many people really need more dynamic range than what today's large sensors provide.
I agree but I think it's human nature to want something more or better even if we don't need it.
 
Hi all, I am a photographer With more than 30 years of expericene (mostly Nikons, NikkortMat, F90, F100, F3, D700, D750... but also some medium format). I have started to consider upgrading my equipment but when I see some wonderful cameras such as Fuji XT5 OR OM3 and I take a look at these amazing reviews and check the image comparisson pages, I see that the quality is not better than the quality that I get from the >10 years old sensor that I got in the D750....so, are new cameras just focussing on design and ergonomics? Is sensor technology evolving at a lower pace?

Just my thouthgs and my first post on this great site!

Thanks!,
Absolutely. As far as I can tell the IQ of cameras isn't any better now than cameras from 10 years ago. If it is the difference is so small that it takes close examination to see it. What has improved is on sensor PDAF auto focus and tracking which has improved a lot. Sensor read out speed has also advanced quite a bit. So there have been improvements in sensors that have improved the speed and usability of cameras but not the image quality.
 
For just outright picture quality, and modern software like de noise etc. I have pictures taken with a Panasonic G1 , that compare well with pictures taken with my modern cameras. It's other areas where the development is taking place. Area's i don't want or need. Some people want and need it.
 
I think what I'm seeing is that the most obvious measure of sensors, megapixels, has plateaued. That leaves other areas - dynamic range, readout speed, noise reduction, - as areas of improvement.
 
A factor influencing our perception of whether or not sensors continue to evolve and improve is the recent lengthening of the generational cycle. At the height of the DSLR era, the typical generational cycle for flagship cameras was 4 years with mid-cycle refreshes. We saw this in the Nikon ecosystem with the D3, D4, and D5. There were four years - almost to the day - between the release of each. Mid-cycle refreshes included the D3S, D3X, and D4S.

However, by the end of the 2010s, the interchangeable lens camera (ILC) market had seen seven consecutive years of declining sales and manufacturers adjusted, accordingly. Nikon did not refresh the D5. Their next flagship body was also their last DSLR flagship, the D6 which was released 4 years after the D5.

The mirrorless era of digital ILCs is still young. Yes, we've had mirrorless ILCs since the Panasonic G1 in 2008, but Canon and Nikon only joined the party in earnest less than ten years ago. Based on the sensor design and specs in the Sony AII and A9III, Canon R5 and R5II, R3 and R1, and rumors about the upcoming Nikon Z9II, manufacturers appear to be moving toward a 6- to 8-year generational cycle between wholly new sensors and processors in flagship bodies.

Some may misconstrue this as evidence that sensor evolution has stopped. However, as evidenced by resolution, data processing, autofocus, burst rate, buffer, and video performance gains, evolution continues. But the pace of evolution has slowed. We're not seeing completely new flagship products every 4 years. The market can't sustain the financial investment required to maintain that pace.

It'll be every 6 years or possibly 8 years between successive generations of flagship bodies...but the evolution of sensor technology continues.
 
I think what I'm seeing is that the most obvious measure of sensors, megapixels, has plateaued. That leaves other areas - dynamic range, readout speed, noise reduction, - as areas of improvement.
Megapixels are like horsepower in cars. After a certain point they are of use to very few.
 
A factor influencing our perception of whether or not sensors continue to evolve and improve is the recent lengthening of the generational cycle. At the height of the DSLR era, the typical generational cycle for flagship cameras was 4 years with mid-cycle refreshes. We saw this in the Nikon ecosystem with the D3, D4, and D5. There were four years - almost to the day - between the release of each. Mid-cycle refreshes included the D3S, D3X, and D4S.

However, by the end of the 2010s, the interchangeable lens camera (ILC) market had seen seven consecutive years of declining sales and manufacturers adjusted, accordingly. Nikon did not refresh the D5. Their next flagship body was also their last DSLR flagship, the D6 which was released 4 years after the D5.

The mirrorless era of digital ILCs is still young. Yes, we've had mirrorless ILCs since the Panasonic G1 in 2008, but Canon and Nikon only joined the party in earnest less than ten years ago. Based on the sensor design and specs in the Sony AII and A9III, Canon R5 and R5II, R3 and R1, and rumors about the upcoming Nikon Z9II, manufacturers appear to be moving toward a 6- to 8-year generational cycle between wholly new sensors and processors in flagship bodies.

Some may misconstrue this as evidence that sensor evolution has stopped. However, as evidenced by resolution, data processing, autofocus, burst rate, buffer, and video performance gains, evolution continues. But the pace of evolution has slowed. We're not seeing completely new flagship products every 4 years. The market can't sustain the financial investment required to maintain that pace.

It'll be every 6 years or possibly 8 years between successive generations of flagship bodies...but the evolution of sensor technology continues.
My estimation is that Japan is in a state comparable to Europe and the USA in the late 70s, when the far better Japanese products wiped the European and US Hi-Fi audio products away within a few years. Anyone remembering Saba, Telefunken, Grundig, Revox, Uher etc.?

Just look at Sony TV's, Panasonic stepped out of the game etc.

A lot of the brilliant Japanese engineers are retired, they have a tough time finding younger engineers that can compete with the older ones due to what is called “the lost generation” in Japan. And they never had honoured good software developers.

It's just that China refuses to step in and disrupt the market of the Japanese dinosaurs with their mediocre and overpriced products, often reassembled from decade old products. The new 5.000,- Sony is a cautionary example.
 
Last edited:
It's just that China refuses to step in and disrupt the market of the Japanese dinosaurs with their mediocre and overpriced products, often reassembled from decade old products. The new 5.000,- Sony is a cautionary example.
They would need to come up with some sort of technological advance like autofocus or mirrorless, which were the last things that really shook up the camera market.
 
It's just that China refuses to step in and disrupt the market of the Japanese dinosaurs with their mediocre and overpriced products, often reassembled from decade old products. The new 5.000,- Sony is a cautionary example.
China is capable. DJI has shown that with their drones. They pretty much beat anything available in the U.S. market. I think if DJI got into the ILC market, they would be successful. But there is this fear that their products are used for gathering data and sending back to the motherland for spying purposes. This is what the U.S. Government is all up in arms about with DJI and a potential ban looming.
 
Last edited:
I think what I'm seeing is that the most obvious measure of sensors, megapixels, has plateaued. That leaves other areas - dynamic range, readout speed, noise reduction, - as areas of improvement.
Megapixels are like horsepower in cars. After a certain point they are of use to very few.
It maybe too early to tell (how great the advancement is) from the recent news releases from Sony [ https://www.dpreview.com/forums/post/68370978 ] and patient application news from Apple, but , as to the OP question as to 'Sensor evolution stuck(?) ' -We may see a real and quite usable improvement for segments of digital camera users that use moble phones as cameras. Specifically, Moble phones will see improvements in dynamic range and functional editing with larger MP available. Apple is all about moble devices so that will be where to look first.

Further real world testing needs to be seen though. Sony's new sensor looks good in example but it needs to be compared to current sensors examples.

Interchangeable lenses will always be where the market is for non-phone cameras.

 
Last edited:
I think what I'm seeing is that the most obvious measure of sensors, megapixels, has plateaued. That leaves other areas - dynamic range, readout speed, noise reduction, - as areas of improvement.
Megapixels are like horsepower in cars. After a certain point they are of use to very few.
It maybe too early to tell (how great the advancement is) from the recent news releases from Sony [ https://www.dpreview.com/forums/post/68370978 ] and patient application news from Apple, but , as to the OP question as to 'Sensor evolution stuck(?) ' -We may see a real and quite usable improvement for segments of digital camera users that use moble phones as cameras. Specifically, Moble phones will see improvements in dynamic range and functional editing with larger MP available. Apple is all about moble devices so that will be where to look first.

Further real world testing needs to be seen though. Sony's new sensor looks good in example but it needs to be compared to current sensors examples.

Interchangeable lenses will always be where the market is for non-phone cameras.

https://www.dpreview.com/forums/post/68368306
Phone sensors have not advanced much. Almost all the advances in phone cameras have come from the in phone processing.
 

Keyboard shortcuts

Back
Top