Is the new Sony 44x33 sensor out in the wild?

Started Apr 30, 2018 | Discussions
Erik Kaffehr
Erik Kaffehr Senior Member • Posts: 2,827
Is the new Sony 44x33 sensor out in the wild?
5

Hi,

Some interesting industrial stuff from Phase One: https://industrial.phaseone.com/landing/The_New_iXM_Series.aspx

It seems that Phase One is marketing an industrial camera with the Sony 100 MP back side illuminated 44x33 mm sensor.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

JimKasson
JimKasson Forum Pro • Posts: 25,120
Re: Is the new Sony 44x33 sensor out in the wild?
2

Erik Kaffehr wrote:

Hi,

Some interesting industrial stuff from Phase One: https://industrial.phaseone.com/landing/The_New_iXM_Series.aspx

It seems that Phase One is marketing an industrial camera with the Sony 100 MP back side illuminated 44x33 mm sensor.

You beat me to it, Erik. Sure looks like the IMX461. If it's shipping to P1, I'll bet it's shipping to Hasseblad and Fujifilm, although possibly in limited quantities.

Looking forward to the GFX 100S and X2D announcements.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Erik Kaffehr
OP Erik Kaffehr Senior Member • Posts: 2,827
Re: Is the new Sony 44x33 sensor out in the wild?

JimKasson wrote:

Erik Kaffehr wrote:

Hi,

Some interesting industrial stuff from Phase One: https://industrial.phaseone.com/landing/The_New_iXM_Series.aspx

It seems that Phase One is marketing an industrial camera with the Sony 100 MP back side illuminated 44x33 mm sensor.

You beat me to it, Erik. Sure looks like the IMX461. If it's shipping to P1, I'll bet it's shipping to Hasseblad and Fujifilm, although possibly in limited quantities.

Looking forward to the GFX 100S and X2D announcements.

Jim

Hi Jim,

Phase One seems to release the sensor in an industrial device.

But, it seems that the sensor is available, so I guess that HB and Fuji announcements may be around the corner.

We may see some chicken race... Announcing new tech may harm sales of existing tech, but no vendor wants to make the impression of lagging behind.

I would expect announcements or road maps from major players within a month, or so.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

JimKasson
JimKasson Forum Pro • Posts: 25,120
Re: Is the new Sony 44x33 sensor out in the wild?
2

Erik Kaffehr wrote:

We may see some chicken race... Announcing new tech may harm sales of existing tech,

In the Santa Clara Valley, we used to call the extreme effects of that The Osborne Effect.

https://en.wikipedia.org/wiki/Osborne_effect

Jim

 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Chris Dodkin
MOD Chris Dodkin Forum Pro • Posts: 10,998
Re: Is the new Sony 44x33 sensor out in the wild?

If they are, as anticipated, launching the GFX50R at Photokina, I wouldn't expect them to announce a higher rez model before then, and kill that buzz.

-- hide signature --

Your time is limited, so don't waste it arguing about camera features - go out and capture memories

 Chris Dodkin's gear list:Chris Dodkin's gear list
Fujifilm FinePix X100 Fujifilm X100F Canon EOS-1D Canon EOS-1D Mark II Fujifilm X-Pro1 +41 more
Rick Knepper
Rick Knepper Forum Pro • Posts: 17,126
Re: Is the new Sony 44x33 sensor out in the wild?

Erik Kaffehr wrote:

Hi,

Some interesting industrial stuff from Phase One: https://industrial.phaseone.com/landing/The_New_iXM_Series.aspx

It seems that Phase One is marketing an industrial camera with the Sony 100 MP back side illuminated 44x33 mm sensor.

Best regards

Erik

The iXM 100MP and iXM 50MP metric cameras are available today for advance order – from Phase One Industrial partners worldwide – with delivery in June 2018.

Something delicious in a FujiHassy wrapper may not be far off.

-- hide signature --

Once you've done fifty, anything less is iffy.

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +9 more
Paul Regular Member • Posts: 293
Re: Is the new Sony 44x33 sensor out in the wild?

these phase bodies are such niche products, i wonder what the sales numbers are compared to fuji and hasselblad....

do we actually know the units sold from fuji and hasselblad? i keep hearing that sales far exceeded expectations for both, but what are the actual numbers?

not sure we will see a gfx100 or x2d before 2019....

the specs of that camera/sensor do look nice though....better DR, 3fps, iso 50 and 4k video 30fps

-- hide signature --

schefz.com

(unknown member) Regular Member • Posts: 165
Re: Is the new Sony 44x33 sensor out in the wild?
1

Paul wrote:

the specs of that camera/sensor do look nice though....better DR, 3fps, iso 50 and 4k video 30fps

The sensor is specced to 6 fps @ 14bits by Sony... but the sensor has 16bit ADCs. We will see, if Fuji brings 16bit RAWs, Hassy already has them.

What I find interesting is that

  • the new sensor actually has 1 dB less DR than the old one at pixel level (will produce about 1 stop more DR when normalized (downscaled))
  • 83dB is below 14 stops, so the 16bits image data will be a bit theoretical, two extra bits will mostly record the noise more precisely
  • ISO50 vs. ISO100 and 83 dB vs. 84 dB seems a bit odd... D810 managed to push about 0.5 stops more DR when it extended the D800's ISO100 to ISO64. Unless the sensor has dual gain, I predict that the high ISO performance looked at 100mpix will be disappointment for some (obviously NR over 100mpix and normalization to 50mpix or less will even out things).
  • Why we do not see 50mpix DR monster made with this same tech (at least for now)? Is it more optimal to push more pixels and take advantage of oversampling instead of improving pixel level DR from current point? Current pixel level DR king is Sony A7iii (AFAIK, haven't gone through every one), which is very close to gap 14bit, but now this MF one has 16bit ADC, so we should be able to take full benefit from over 14 stop pixel level DR (sidenote: A7 series now has to move 16bit ADC if they want to take pixel level DR advantage of any future development after A7iii). Or maybe I have understood something completely wrong.

I personally guestimate that the current 50mpix will be better choice for me, as I do not need 100mpix and these MF tighter aperture lenses will require using higher ISOs in my kinda work so I won't benefit from ISO50. But time will tell how badly I understood these specs :).

Erik Kaffehr
OP Erik Kaffehr Senior Member • Posts: 2,827
Re: Is the new Sony 44x33 sensor out in the wild?

umeet wrote:

Paul wrote:

the specs of that camera/sensor do look nice though....better DR, 3fps, iso 50 and 4k video 30fps

The sensor is specced to 6 fps @ 14bits by Sony... but the sensor has 16bit ADCs. We will see, if Fuji brings 16bit RAWs, Hassy already has them.

What I find interesting is that

  • the new sensor actually has 1 dB less DR than the old one at pixel level (will produce about 1 stop more DR when normalized (downscaled))
  • 83dB is below 14 stops, so the 16bits image data will be a bit theoretical, two extra bits will mostly record the noise more precisely
  • ISO50 vs. ISO100 and 83 dB vs. 84 dB seems a bit odd... D810 managed to push about 0.5 stops more DR when it extended the D800's ISO100 to ISO64. Unless the sensor has dual gain, I predict that the high ISO performance looked at 100mpix will be disappointment for some (obviously NR over 100mpix and normalization to 50mpix or less will even out things).
  • Why we do not see 50mpix DR monster made with this same tech (at least for now)? Is it more optimal to push more pixels and take advantage of oversampling instead of improving pixel level DR from current point? Current pixel level DR king is Sony A7iii (AFAIK, haven't gone through every one), which is very close to gap 14bit, but now this MF one has 16bit ADC, so we should be able to take full benefit from over 14 stop pixel level DR (sidenote: A7 series now has to move 16bit ADC if they want to take pixel level DR advantage of any future development after A7iii). Or maybe I have understood something completely wrong.

I personally guestimate that the current 50mpix will be better choice for me, as I do not need 100mpix and these MF tighter aperture lenses will require using higher ISOs in my kinda work so I won't benefit from ISO50. But time will tell how badly I understood these specs :).

Hi,

Having twice the amount of pixels can improve DR, not at the pixel level but for any given image size.So it may seen that you have a lot of so good pixels, but when viewing the image at a fixed scale the smaller pixels would yield a better image. i would argue that it is mostly the case.

Image quality is an interesting issue. If the sensor resolution does not match the resolution of the lens we will get a lot of fake detail.

Making the pixels smaller introduces more shot noise for each pixel, but once you show the image on screen, or print, all the pixels will be resampled to the resolution of your screen or printer.

In the end, shot noise will be the same - regardless of the number of pixels. Making the pixels smaller may affect DR, but smaller pixels tend to have less readout noise.

Smaller pixels will probably sharpen better.

My presumption is that you will never gain image quality using smaller pixels, unless:

  • You are shooting video, which pretty much binds you to 2K, 4K and 8K.
  • Shooting extremely low light.

In those cases large pixels may have some benefits.

But, going to small pixels may also have diminishing returns.

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Paul Regular Member • Posts: 293
Re: Is the new Sony 44x33 sensor out in the wild?

i wont comment on the 14 vs 16bit thing....i am fine with what the gfx has now, i dont see any advantage in the X1D files even if they are 16bit, i do see an advantage over the 14 bit sony files, my own opinion.....i dont see this as a reason to buy or not to buy the new version....

i do like having a 50 iso option for outdoor flash.....i cant see the high iso performance being worse then what the GFX has now (with 5? year old tech), i do expect about a stop better there without taking the extra pixels into consideration.....

i dont really need the 100mpix, or the file sizes that come with that.....but it is a huge bonus to have in commercial work, the ability to just crop in on everything is huge....i love being able to simply always shoot wider and cropping in later....

the 100mpix should be better when it comes to moire, which does show up in fabrics quite a bit now....

i just expect everything to be a little quicker, fps, AF (with or without PDAF) with slightly better DR and high iso....and that would make the GFX pretty much a perfect camera for me....

-- hide signature --

schefz.com

Rick Knepper
Rick Knepper Forum Pro • Posts: 17,126
Dumb question.

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

umeet wrote:

Paul wrote:

the specs of that camera/sensor do look nice though....better DR, 3fps, iso 50 and 4k video 30fps

The sensor is specced to 6 fps @ 14bits by Sony... but the sensor has 16bit ADCs. We will see, if Fuji brings 16bit RAWs, Hassy already has them.

What I find interesting is that

  • the new sensor actually has 1 dB less DR than the old one at pixel level (will produce about 1 stop more DR when normalized (downscaled))
  • 83dB is below 14 stops, so the 16bits image data will be a bit theoretical, two extra bits will mostly record the noise more precisely
  • ISO50 vs. ISO100 and 83 dB vs. 84 dB seems a bit odd... D810 managed to push about 0.5 stops more DR when it extended the D800's ISO100 to ISO64. Unless the sensor has dual gain, I predict that the high ISO performance looked at 100mpix will be disappointment for some (obviously NR over 100mpix and normalization to 50mpix or less will even out things).
  • Why we do not see 50mpix DR monster made with this same tech (at least for now)? Is it more optimal to push more pixels and take advantage of oversampling instead of improving pixel level DR from current point? Current pixel level DR king is Sony A7iii (AFAIK, haven't gone through every one), which is very close to gap 14bit, but now this MF one has 16bit ADC, so we should be able to take full benefit from over 14 stop pixel level DR (sidenote: A7 series now has to move 16bit ADC if they want to take pixel level DR advantage of any future development after A7iii). Or maybe I have understood something completely wrong.

I personally guestimate that the current 50mpix will be better choice for me, as I do not need 100mpix and these MF tighter aperture lenses will require using higher ISOs in my kinda work so I won't benefit from ISO50. But time will tell how badly I understood these specs :).

-- hide signature --

Once you've done fifty, anything less is iffy.

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +9 more
Erik Kaffehr
OP Erik Kaffehr Senior Member • Posts: 2,827
Re: Dumb question.
1

Rick Knepper wrote:

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

Hi,

You can get the DR in stops by dividing the dB rating by 6.0, or to be more exact 6.0201. (*)

That gives you engineering DR per pixel. That is the same value DxO-mark uses in "Screen" mode.

DxO-mark's "Print-mode" is normalized to 8 MP.

Sensor makers often use dB, as it is the normal term used in signal processing.

Best regards

Erik

(*) The number is 2 x log(2)

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

JimKasson
JimKasson Forum Pro • Posts: 25,120
Re: Dumb question.

Erik Kaffehr wrote:

Sensor makers often use dB, as it is the normal term used in signal processing.

And we pay homage to this man every time we use it:

https://en.wikipedia.org/wiki/Alexander_Graham_Bell

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Rick Knepper
Rick Knepper Forum Pro • Posts: 17,126
Re: Dumb question.

Erik Kaffehr wrote:

Rick Knepper wrote:

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

Hi,

You can get the DR in stops by dividing the dB rating by 6.0, or to be more exact 6.0201. (*)

That gives you engineering DR per pixel. That is the same value DxO-mark uses in "Screen" mode.

On the iXM 100MP web page, it claims 83 dB. You didn't use the term "stop" above but I infer the following equation and result: 83/6.0201 = 13.78714639291760 stops? If this is correct, what do you conjecture the PDR might be?

DxO-mark's "Print-mode" is normalized to 8 MP.

Sensor makers often use dB, as it is the normal term used in signal processing.

Best regards

Erik

(*) The number is 2 x log(2)

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

-- hide signature --

Once you've done fifty, anything less is iffy.

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +9 more
JimKasson
JimKasson Forum Pro • Posts: 25,120
Re: Dumb question.

Rick Knepper wrote:

Erik Kaffehr wrote:

Rick Knepper wrote:

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

Hi,

You can get the DR in stops by dividing the dB rating by 6.0, or to be more exact 6.0201. (*)

That gives you engineering DR per pixel. That is the same value DxO-mark uses in "Screen" mode.

On the iXM 100MP web page, it claims 83 dB. You didn't use the term "stop" above

See the word that I bolded above.

but I infer the following equation and result: 83/6.0201 = 13.78714639291760 stops?

Wow! Look at all those (significant?) figures. I think we probably need only one after the decimal point.

If this is correct, what do you conjecture the PDR might be?

You can't compute PDR from EDR alone. We need to know the FWC, and AFAIK, we don't know that yet.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Erik Kaffehr
OP Erik Kaffehr Senior Member • Posts: 2,827
Re: Dumb question.

JimKasson wrote:

Rick Knepper wrote:

Erik Kaffehr wrote:

Rick Knepper wrote:

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

Hi,

You can get the DR in stops by dividing the dB rating by 6.0, or to be more exact 6.0201. (*)

That gives you engineering DR per pixel. That is the same value DxO-mark uses in "Screen" mode.

On the iXM 100MP web page, it claims 83 dB. You didn't use the term "stop" above

See the word that I bolded above.

but I infer the following equation and result: 83/6.0201 = 13.78714639291760 stops?

Wow! Look at all those (significant?) figures. I think we probably need only one after the decimal point.

If this is correct, what do you conjecture the PDR might be?

You can't compute PDR from EDR alone. We need to know the FWC, and AFAIK, we don't know that yet.

Jim

Hi,

DxO-mark gives an EDR of 13.49 EV for the X1D (their precision...) and the new sensor has 13.7 - 13.9 EV according to Sony. But the pixel are is half the size. I guess that there is some improvement in full well capacity per area and that some of the DR is coming lower readout noise.

What we essentially know is that the new 100 MP sensor is probably a tiny bit better in DR at the pixel level and that should give a significant improvement with the number of pixels taken into account.

My guess is that we may se an 0.7 EV improvement at the sensor level, compared to the 50 MP sensor.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Rick Knepper
Rick Knepper Forum Pro • Posts: 17,126
Re: Dumb question.

Erik Kaffehr wrote:

JimKasson wrote:

Rick Knepper wrote:

Erik Kaffehr wrote:

Rick Knepper wrote:

This is the first time I've seen dBs used in quoting sensor DR. How many stops does this represent?

Hi,

You can get the DR in stops by dividing the dB rating by 6.0, or to be more exact 6.0201. (*)

That gives you engineering DR per pixel. That is the same value DxO-mark uses in "Screen" mode.

On the iXM 100MP web page, it claims 83 dB. You didn't use the term "stop" above

See the word that I bolded above.

but I infer the following equation and result: 83/6.0201 = 13.78714639291760 stops?

Wow! Look at all those (significant?) figures. I think we probably need only one after the decimal point.

I was trying to be "exact", See what I boldedabove.

If this is correct, what do you conjecture the PDR might be?

You can't compute PDR from EDR alone. We need to know the FWC, and AFAIK, we don't know that yet.

OK

Jim

-- hide signature --

Hi,

DxO-mark gives an EDR of 13.49 EV for the X1D (their precision...) and the new sensor has 13.7 - 13.9 EV according to Sony. But the pixel are is half the size. I guess that there is some improvement in full well capacity per area and that some of the DR is coming lower readout noise.

What we essentially know is that the new 100 MP sensor is probably a tiny bit better in DR at the pixel level and that should give a significant improvement with the number of pixels taken into account.

My guess is that we may se an 0.7 EV improvement at the sensor level, compared to the 50 MP sensor.

Best regards

Erik

So, nothing to get excited about for DR junkies then.

-- hide signature --

Once you've done fifty, anything less is iffy.

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +9 more
KWNJr Contributing Member • Posts: 924
Re: Dumb question.
1
  • Rick Knepper wrote:

Erik Kaffehr wrote:

What we essentially know is that the new 100 MP sensor is probably a tiny bit better in DR at the pixel level and that should give a significant improvement with the number of pixels taken into account.

My guess is that we may se an 0.7 EV improvement at the sensor level, compared to the 50 MP sensor.

Best regards

Erik

So, nothing to get excited about for DR junkies then.

Don't DR junkies get excited about anything >.01 increase?😀

-- hide signature --

Once you've done fifty, anything less is iffy.

(unknown member) Regular Member • Posts: 165
Re: Dumb question.
1

Erik Kaffehr wrote:

Hi,

DxO-mark gives an EDR of 13.49 EV for the X1D (their precision...) and the new sensor has 13.7 - 13.9 EV according to Sony. But the pixel are is half the size. I guess that there is some improvement in full well capacity per area and that some of the DR is coming lower readout noise.

What we essentially know is that the new 100 MP sensor is probably a tiny bit better in DR at the pixel level and that should give a significant improvement with the number of pixels taken into account.

My guess is that we may se an 0.7 EV improvement at the sensor level, compared to the 50 MP sensor.

Best regards

Erik

PhaseOne says the old 50mpix sensor has 84dB DR and this new 100 mpix has 83dB DR.

So this new one likely will not give better scores than the X1D has at "screen" or pixel level DR at DxOmark.

This same deal has been with FF sensors since D810. The pixel level max DR has stagnated to around 13.5 EVs, with A7iii being only notable exception with its almost 14 EVs (obv. because it has bigger pixels). One reason probably has been 14bit ADCs.

But most certainly the normalized "print" measurement of X1D will be beaten easily by this 100mpix.

JimKasson
JimKasson Forum Pro • Posts: 25,120
Re: Dumb question.

umeet wrote:

Erik Kaffehr wrote:

Hi,

DxO-mark gives an EDR of 13.49 EV for the X1D (their precision...) and the new sensor has 13.7 - 13.9 EV according to Sony. But the pixel are is half the size. I guess that there is some improvement in full well capacity per area and that some of the DR is coming lower readout noise.

What we essentially know is that the new 100 MP sensor is probably a tiny bit better in DR at the pixel level and that should give a significant improvement with the number of pixels taken into account.

My guess is that we may se an 0.7 EV improvement at the sensor level, compared to the 50 MP sensor.

Best regards

Erik

PhaseOne says the old 50mpix sensor has 84dB DR and this new 100 mpix has 83dB DR.

So this new one likely will not give better scores than the X1D has at "screen" or pixel level DR at DxOmark.

This same deal has been with FF sensors since D810. The pixel level max DR has stagnated to around 13.5 EVs, with A7iii being only notable exception with its almost 14 EVs (obv. because it has bigger pixels). One reason probably has been 14bit ADCs.

But most certainly the normalized "print" measurement of X1D will be beaten easily by this 100mpix.

I think it needs emphasizing that you're talking about engineering dynamic range (EDR) here. EDR is an easy to measure, but not very useful, indicator of dynamic range in actual photographic use.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Fujifilm GFX 50S Sony a9 Sony a7R III Sony a7 III +5 more
Keyboard shortcuts:
FForum MMy threads