JimH123 wrote:
There is an update yesterday on Global Shutter progress at Sony. No word on when this will show up in APS-C and FF cameras, but my guess is that they are going for it.
https://www.sonyalpharumors.com/press-text-sony-develops-a-stacked-cmos-global-shutter-image-sensor/I am also including a cross section of the design comparing conventional vs BSI. Notice the much wider incident light angle. One restraint that new lenses have is that the deeper wells used in conventional designs limit the angle of usable incident light. And notice also the much wider incident light angle in the BSI design.
Why is this important?
In reading David Busch's book on the Sony A7iii, he explains how digital lenses for conventional sensors are constrained to project the light in a straight path onto the sensor whereas the BSI design is perfectly content with light coming at lower angles which is how the older film cameras were designed. As an example of that, I have been playing with my Sony A7iii (which has a BSI sensor) and older Asahai Pentax Takumar lenses, and I am really liking the results more so than in using my older A99 which has now been sold.
Additionally, the article goes on to say how this sensor has really high precision and high processing speeds. Placing the circuitry and wiring onto the other side allows room for the pixels to use the sizes they need to use for a given design. Sounds like earlier tradeoffs for global shutters is no longer a concern.
The way Sony is working on global shutter, it certainly sounds like it is coming. And just guessing, I suspect that this is why the A7000 and A7siii are taking longer to be released. Also makes me wonder if Global Shutter is perfected, will they completely remove the mechanical shutter? What sense does having a mechanical shutter have if the Global Shutter can do everything the Mechanical Shutter can do?
A Global Shutter opens other possibilities too. Since the entire array can be accessed at once, perhaps they can come up with new ways to combine images and to reduce noise where a series of shots can be processed real time and to then produce a RAW file from the processed results. Sony today can do multi-frame noise reduction, and it works really well. But it requires shooting in JPEG and requires all the overhead time needed to shoot the multiple shots before it generates the final result. With Global Shutter, I expect that this process will be faster, perhaps much faster and increase the types of shots it can be used for.
I'm sure that there are many other things that can be thought of too.
There is an update yesterday on Global Shutter pro... (
show quote)
Thanks, Jim, for posting something really interesting and on-topic.
That link on Pregius S is exciting. Unfortunately, it doesn't talk about cost
or pricing.
"Back-illuminated pixel function" makes design and fabrication somewhat easier--
but doesn't necessarily reduce cost. Sony does seem to have improved resolution--
though it still can't match the best rolling shutter sensor (or straight sensor with
focal plane mechanical shutter --- still the benchmark of IQ).
As I'm sure you know, the angle thing comes down to geometry: the node-to-sensor length
and the wide dimension of the format. If a camera designer wants a short FFD and doesn't
want a tiny sensor, then the light must strike the corners of the sensor at a high angle. So
this advance by Sony is significant.
On the other hand, there is nothing forcing a short FFD other than consumer preferences
for dinky cameras. If consumers were to prefer miniature bicycles or minature pianos,
so what? It wouldn't make me want to ride one or play one.
I have to give Sony credit for tackling the problem. For many makers of mirrorless
cameras, the "solution" is a tiny sensor--and big advertising.
After hearing about global shutters for more than a decade....I'm jaded. Llike fusion power,
a perfected global shutter sensor is always "just around the corner".
It''s going to require 3D IC, which is at the bleeding edge of what is possible
to fabricate. It can be done....for industrial and military applicaitons. Sony already
has soe great industiral sensors. But very expensive.
Will prices come down? Well, high-end processor prices haven't. Multi-core Xeons
are mighty expensive. And the iTanium is cancelled.
The problem, I think , is that a "global shutter" isn't a shutter. It's a shutterless design.
Shutters interrupt a light beam, a "global shutter" sensor tries to move massive
amounts of data in a very short time. That has never been easy.
Solid sttate LCD shutters are off-the-shelf items. They just aren't fast enough. That
seems to me a much more tractable problem than what to do with heat when you've
got no air flow.
That link also doesn't talk about heat. The heat problem at high frame rates is not going
to go away. If you want to refrigerate the sensor, then it's manageable. But consumer cameras
have no air flow, let alone room (or batttery capacity) for refrigeration.
How long would an Intel Xeon processor last with no heat sink and no air flow? Probably
about 1 minute. Then it would be hot enough to fry an egg.
If you are willing to give up speed, then you can get the power consumption way down.
Bu tnobody wants a camera with a maximum shutter speed of 100. And these days, a
lot of confusmers are demanding high frame rates for video.
Besides, if Sony ever gets it working with APS-C, "full frame" is larger still. And medium f
format....
When I was a kid, the fastest airplane in the world was the Lockheed SR-71. I built the Revel
scale model--made from the Lockheed blueprints--when the entire design was still classified.
It may have been secret, but I had an accurate scale model of the airframe hanging in my bedroom.
Today the SR-71 is still the fastest airplane in the world. Could Lockheed Martin build a faster one?
Sure. Only, nobody is willing to pay for it. There is no mission that is worth the enormous cost.
The SR-17 is retired, and the U2 is still flying. How ironic!
I once watched two U-2s take off from Beal AFB in CA.: a black one belonging to the USAF and
a white one belonging to NASA. First one, then the other climbed through our altitude, very slowly,
then turned the nose up and climbed out of sight--as if it was on an escalator.. Fantastic.
Probably landed somewhere in Europe.
The U-2 turned out to be more cost-effective and less "bleeding edge" than the SR-71. I mean,
triethylborane ignition on contact with air--not something Cessna is likely to adopt. The SR-71
is like a race car. I understand there is only one two-seater SR-71B trainer still in existence
(the other one crashed).
There's technology, and then there's consumer technology. Global shutters are the former.
Great for industrial and military applications, but as unlikely a consumer product as a
ramjet engine.
But if you can find 500,000 consumers each willing to pay $10,000 for a perfected global shutter
digital camera, I believe Sony could deliver it. Otherwise it will remain an industrial and
military product--a lucratative market in which Sony is doing quite well.