June 20, 2018, 01:20:42 PM

Author Topic: Industry News: Sony Develops a Back-Illuminated CMOS Image Sensor with Pixel-Parallel A/D Converter  (Read 6859 times)

exquisitor

  • EOS Rebel SL2
  • ***
  • Posts: 76
I find this more impressive:
https://photorumors.com/2018/02/17/panasonic-develops-8k-global-shutter-technology-using-organic-photoconductive-film-cmos-image-sensor/

This is obviously not only a 1.5 MP sample, but has at least 32 MP with a series of very interesting features.

canon rumors FORUM


Jim Saunders

  • EOS 5DS R
  • ******
  • Posts: 1061
    • HHA Photo
Would I buy a 10MP full-frame mirrorless with flash sync to 1/1000?  Absolutely.  Anyone else may not, but I know what I want.

Jim
A good image might fit entirely into the histogram, might not.

woodman411

  • EOS Rebel 300D
  • ***
  • Posts: 56
I think this isn't a good idea, for any reason. The mechanical shutter isn't an issue, it works just fine.
This kind of sensor is inefficient in many kind of views.

hmmm, I would like to have 19th century mechanical shutters eliminated from all of my cameras. Pure solid state please.
Why?

1. Zero vibration
2. Zero shutter noise
3. Zero oil/lubricants near sensor
4. Zero particle abrasion near sensor
5. Zero wear and tear ["electronic aging" is a non-issue in cameras - provided hi-quality components are used]

So ... way to go, Sony! Probably only another 20 years until "innovative Canon" introduces it too ...  :P

I would take point 5 off the table - if this was true, laptops and tablets would never fail, but even hi-quality companies/components like Apple, would be impressed if it lasted 5+ years without motherboard/ram/ssd failure. In automobiles, good mechanical components often outlast electronic ones too.

3kramd5

  • Canon EF 300mm f/2.8L IS II
  • ********
  • Posts: 2097
Would I buy a 10MP full-frame mirrorless with flash sync to 1/1000?  Absolutely.  Anyone else may not, but I know what I want.

Jim

How about mirrorless camera with a 50MP sensor about 2/3 larger that will sync at 1/2000?

Don Haines

  • Canon EF 800mm f/5.6L IS
  • ***********
  • Posts: 6852
  • posting cat pictures on the internet since 1986
I think this isn't a good idea, for any reason. The mechanical shutter isn't an issue, it works just fine.
This kind of sensor is inefficient in many kind of views.

hmmm, I would like to have 19th century mechanical shutters eliminated from all of my cameras. Pure solid state please.
Why?

1. Zero vibration
2. Zero shutter noise
3. Zero oil/lubricants near sensor
4. Zero particle abrasion near sensor
5. Zero wear and tear ["electronic aging" is a non-issue in cameras - provided hi-quality components are used]

So ... way to go, Sony! Probably only another 20 years until "innovative Canon" introduces it too ...  :P

I would take point 5 off the table - if this was true, laptops and tablets would never fail, but even hi-quality companies/components like Apple, would be impressed if it lasted 5+ years without motherboard/ram/ssd failure. In automobiles, good mechanical components often outlast electronic ones too.

My TRS-80 pocket computer still works.... and I have a HP-29C calculator at work that is approaching 40 years old.....  and it is rare for a work computer to fail in 5 years. Most of our electronics test equipment is at least 10 years old.... you can most definitely build electronics to last more than 5 years.
The best camera is the one in your hands

traveller

  • EOS 5DS R
  • ******
  • Posts: 843
What do you use a 1.45 MP sensor for?  Machine vision?  I have a old Fujifilm MX700 with a 1.5MP sensor, it works, takes reasonable photos, but 1.5 MP just doesn't cut it.

Lab prototype.....

My first digital camera was 320 X 200 pixels. I told people in my camera club that this was revolutionary and the way of the future.... They laughed at me.... No look where we are......

This is an intermediate step.... The end goal is (probably) counting photons as they hit the photocells.... When you can put over 10 billion transistors on a chunk of silicon, these things start to become more and more likely.
Yes but... this line:

"The inclusion of nearly 1,000 times as many A/D converters compared to the traditional column A/D conversion method*² means an increased demand for current."

Suggests that there may be problems with scaling this up to higher resolutions. I also wonder how large this Sony prototype sensor is? The fact that they don't seem to mention the size in their press release suggests to me that it is small (i.e. not APS-C, FF, or even 1"). I don't know if size scaling would present issues as well...

Like 3kramd5 states, developing a prototype is one thing, but getting it to market is quite another. Where are the Canon cameras equipped with that global shutter announced 18 months ago?

It's about 1.5 megapixels....

Erm, what? Have they changed the definition of a pixel? Last time I looked it was a dimensionless unit...

Tremotino

  • PowerShot G7 X Mark II
  • **
  • Posts: 7
What do you use a 1.45 MP sensor for?  Machine vision?  I have a old Fujifilm MX700 with a 1.5MP sensor, it works, takes reasonable photos, but 1.5 MP just doesn't cut it.

Lab prototype.....

My first digital camera was 320 X 200 pixels. I told people in my camera club that this was revolutionary and the way of the future.... They laughed at me.... No look where we are......

This is an intermediate step.... The end goal is (probably) counting photons as they hit the photocells.... When you can put over 10 billion transistors on a chunk of silicon, these things start to become more and more likely.
Yes but... this line:

"The inclusion of nearly 1,000 times as many A/D converters compared to the traditional column A/D conversion method*² means an increased demand for current."

Suggests that there may be problems with scaling this up to higher resolutions. I also wonder how large this Sony prototype sensor is? The fact that they don't seem to mention the size in their press release suggests to me that it is small (i.e. not APS-C, FF, or even 1"). I don't know if size scaling would present issues as well...

Like 3kramd5 states, developing a prototype is one thing, but getting it to market is quite another. Where are the Canon cameras equipped with that global shutter announced 18 months ago?

I don't what I should do with a rolling shutter wikia page, but he will know..  :o
I only attend one university lesson of industrial image processing in Europe, maybe in Japan the don't cook with water, how knows?! ::)
The draw back of this technology is at the moment way too huge :/
For 2 clear reasons:
1° a 14-bit adc with low now noise of the size of a fotodiode haha.. :'D I know what I'm talking about, that's simply crazy  ;D the image quality can't  be better with smaller adc
2° when this technology will be mature, we will have 100MP sensor cameras.. something like 12000x8000 diodes. In the prototype  version it might by 1000 times more current. In the actual sensor about 12000 times more current that's inefficient and nonsense.

Thank you, sxquisitor, for you post, the Panasonic concept is simple and astonishing! That's what we will see in the future.
The serial readout of the sensor isn't an issue, the simultaneous exposure of the sensore indeed is, the readout of the sensor hasn't to occur at the same time.
« Last Edit: February 18, 2018, 06:41:20 PM by Tremotino »

canon rumors FORUM


AvTvM

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3165
My TRS-80 pocket computer still works.... and I have a HP-29C calculator at work that is approaching 40 years old.....  and it is rare for a work computer to fail in 5 years. Most of our electronics test equipment is at least 10 years old.... you can most definitely build electronics to last more than 5 years.

exactly. If electronic components with some defect/weakness are rigidly weeded out during QC testing at factory immediately after production (including burn-ins) ... and are operated within specs [eg temperature, humidity, voltage, etc.] they will probably work 100+ years ... definitely long after being "technically obsoleted" by newer tech.

bwud

  • EOS 80D
  • ****
  • Posts: 284
My TRS-80 pocket computer still works.... and I have a HP-29C calculator at work that is approaching 40 years old.....  and it is rare for a work computer to fail in 5 years. Most of our electronics test equipment is at least 10 years old.... you can most definitely build electronics to last more than 5 years.

exactly. If electronic components with some defect/weakness are rigidly weeded out during QC testing at factory immediately after production (including burn-ins) ... and are operated within specs [eg temperature, humidity, voltage, etc.] they will probably work 100+ years ... definitely long after being "technically obsoleted" by newer tech.

Typically QC doesn't find the types of infant mortality defects you describe. Testing like ESS does. Beyond that, the most common failure modes not related to abuse, if you consider only the types of components found in cameras, are likely fatigued solder joints from temperature cycling.

Talys

  • EOS-1D X Mark II
  • *******
  • Posts: 1646
  • Canon 6DII
Would I buy a 10MP full-frame mirrorless with flash sync to 1/1000?  Absolutely.  Anyone else may not, but I know what I want.

Jim

Can't you live with HSS (or equivalent-functioning technology)?

Mt Spokane Photography

  • CR GEEK
  • ***************
  • Posts: 14229
What do you use a 1.45 MP sensor for?  Machine vision?  I have a old Fujifilm MX700 with a 1.5MP sensor, it works, takes reasonable photos, but 1.5 MP just doesn't cut it.

Lab prototype.....

My first digital camera was 320 X 200 pixels. I told people in my camera club that this was revolutionary and the way of the future.... They laughed at me.... No look where we are......

This is an intermediate step.... The end goal is (probably) counting photons as they hit the photocells.... When you can put over 10 billion transistors on a chunk of silicon, these things start to become more and more likely.

I don't think its entirely a prototype limitation that can be practically upscaled using current technology, its not transistors we are talking about, its A/D convertors.  They do not even mention upscaling to a photographic high MP sensor in the press release.  I think its currently limited by power requirements, so in that respect, its a lab prototype, but not something that we will see in production of 40 MP sensors soon.

The strange thing about the press release is that they make it sound like its a product you will soon buy.

Don Haines

  • Canon EF 800mm f/5.6L IS
  • ***********
  • Posts: 6852
  • posting cat pictures on the internet since 1986
What do you use a 1.45 MP sensor for?  Machine vision?  I have a old Fujifilm MX700 with a 1.5MP sensor, it works, takes reasonable photos, but 1.5 MP just doesn't cut it.

Lab prototype.....

My first digital camera was 320 X 200 pixels. I told people in my camera club that this was revolutionary and the way of the future.... They laughed at me.... No look where we are......

This is an intermediate step.... The end goal is (probably) counting photons as they hit the photocells.... When you can put over 10 billion transistors on a chunk of silicon, these things start to become more and more likely.

I don't think its entirely a prototype limitation that can be practically upscaled using current technology, its not transistors we are talking about, its A/D convertors.  They do not even mention upscaling to a photographic high MP sensor in the press release.  I think its currently limited by power requirements, so in that respect, its a lab prototype, but not something that we will see in production of 40 MP sensors soon.

The strange thing about the press release is that they make it sound like its a product you will soon buy.

Yes, but logic gates, memory, and a/d converters are made out of transistors...
The best camera is the one in your hands

Normalnorm

  • EOS 6D Mark II
  • *****
  • Posts: 366
Would I buy a 10MP full-frame mirrorless with flash sync to 1/1000?  Absolutely.  Anyone else may not, but I know what I want.

Jim

Can't you live with HSS (or equivalent-functioning technology)?

HSS is a kludge that provides greatly reduced power in order to sync at higher speeds.
A leaf shutter allows the full power of a flash at all speeds with fall-off only at the highest speeds where  along duration flash acts more like ambient light.

While the promise of a global shutter seems nice, in practice, flash exposure  may be uneven at high speeds.
Uneven shutter performance at low speeds is invisible in video but high shutter speeds in still imaging will be magnified.
We will just have to wait and see what will be delivered before we can make any proclamations about what we think will happen.

canon rumors FORUM


epsiloneri

  • EOS 6D Mark II
  • *****
  • Posts: 416
1. Zero vibration
2. Zero shutter noise
3. Zero oil/lubricants near sensor
4. Zero particle abrasion near sensor
5. Zero wear and tear ["electronic aging" is a non-issue in cameras - provided hi-quality components are used]
You forgot

6. Higher fps than possible with mechanical shutter.

Current global shutter sCMOS technology from e.g. Andor provides up to 50 fps with 4 Mpix (monochrome); see their Zyla 5.5 model.

BobG

  • PowerShot G1 X II
  • ***
  • Posts: 22
What do you use a 1.45 MP sensor for?  Machine vision?  I have a old Fujifilm MX700 with a 1.5MP sensor, it works, takes reasonable photos, but 1.5 MP just doesn't cut it.

Lab prototype.....

My first digital camera was 320 X 200 pixels. I told people in my camera club that this was revolutionary and the way of the future.... They laughed at me.... No look where we are......

This is an intermediate step.... The end goal is (probably) counting photons as they hit the photocells.... When you can put over 10 billion transistors on a chunk of silicon, these things start to become more and more likely.
Yes but... this line:

"The inclusion of nearly 1,000 times as many A/D converters compared to the traditional column A/D conversion method*² means an increased demand for current."

Suggests that there may be problems with scaling this up to higher resolutions. I also wonder how large this Sony prototype sensor is? The fact that they don't seem to mention the size in their press release suggests to me that it is small (i.e. not APS-C, FF, or even 1"). I don't know if size scaling would present issues as well...

Like 3kramd5 states, developing a prototype is one thing, but getting it to market is quite another. Where are the Canon cameras equipped with that global shutter announced 18 months ago?

size is 16.08 by 12.73 mm

canon rumors FORUM