Hasselblad X1D - Actual part number/model of the Sony sensor?

[Lust4Life]

I am doing a little study - have not been able to find the answer myself after many searches.
What are the actual Sony Sensor part numbers/models for:
X1D
Fuji GFX
Sony a7RII

Does anyone know for sure?
I want to review the actual spect sheets on the sensors.

Jack

[landscapephoto]

The X1D sensor is called IMX161. The other sensors are here: http://www.sony-semicon.co.jp/products_en/IS/sensor2/products/index.html

[Lust4Life]

The X1D sensor is called IMX161. The other sensors are here: http://www.sony-semicon.co.jp/products_en/IS/sensor2/products/index.html

Thanks, thats a beginning for what I want to learn BUT I am trying to define without question that the X1D is providing a 16 bit color ramp direct from the array/sensor or is it up-reved in the camera from 14 bit.

Both the Sony and the Fuji are only pulling 14 bits of color from their chip according to their data sheets BUT Hassie is providing 16.

[Paul2660]

They all use the same 50mp sensor P1, IQ350 X1D and Fuji. IMX 161. Over 2 years old now.

There is currently one one physical sensor. The fact the Hasselblad is pulling 16bit info from a 14bit sensor seems to be similar to the claims that the older CCD sensors were 16 bit. Which they epwere not. Hasselblad may be adding something in the processing of the data which is unique to their files only they know.

The only 16 bit MF sensor is the 100mp sensor and P1 does have a unique solution that provides 16 bit data some climate 15 bit.

As for real data on the IMX161 you will not find much real spec sheet data out there on it.

Paul Caldwell

[scyth]

They all use the same

OP mentioned 2 MF cameras (X1D & GFX - for whatever reason omitting Ricoh/Pentax MF dSLR) and one FF camera (A7RII)... there are _NOT_ all the same

[scyth]

Both the Sony and the Fuji are only pulling 14 bits of color from their chip

different chips ... there are no other (manufacturer-wise) implementation of 42mp FF sensor...

[Paul2660]

OP mentioned 2 MF cameras (X1D & GFX - for whatever reason omitting Ricoh/Pentax MF dSLR) and one FF camera (A7RII)... there are _NOT_ all the same

Thanks I overlooked the 645z as it uses the same chip as the others. I assumed he knew the A7RII was 35mm and 14bit 42mp.

Paul Caldwell

[Lust4Life]

Correct, I knew this.

Actually, my experience with Sensors/Arrays goes back to Eikonix.
Worked with a group out of MIT to develop the Eikonix chip decades ago.

What I am seeking is not opinion, but the actual data sheets with FULL specs on the respective chips.
I have written to Sony requesting this but no reply as of this moment.

I have worked with Leaf, Hasselblad and Nikon years back in developing their first digital cameras, so I have learned to
check the actual specifications myself before forming an opinion or making a purchase.

But let's stay on topic of what I initially asked.  Where can I get my mits on the spec sheets?

[Paul2660]

Hi Jack,

That is my point, there is nothing much written that is published on the IMX161. In fact nothing on any of the Sony MF chips as  I looked for this info a while back. Sony's data sheets are very vague on the chip at least what is out there on the web.  Even Sony's future data on the follow chips is only pointing to megapixels and not 16 or 14 bit.

For the 50MP MF chip, the IMX161, its a 14 bit implementation in all other companies that use it, Fuji, Pentax, Phase One.  If Hasselblad has a full 16 bit implementation from this chip (which I totally doubt), then it's being done with filler data after the capture.  It would make no sense to have three other companies bypass the full 16 bit data path (if it was truly there) and just use 14 bit.

The only facts I found were that Sony only has data listed on 1 MF 44 x 33 chip, limited as it is, and that chip is in use by all 4 vendors. 

The only unique features that anyone has developed are the ES and EFC that somehow Fuji was able to develop from the same IMX161, and Fuji uses a thicker coverglass, which Fuji has documented in their literature.

If you are able to find anything more detailed from Sony, please share it here as I would like to read up more on this chip.


Paul Caldwell

[Lust4Life]

Hi Paul,

Thank you for sharing your thoughts.  Yes, that is what I have been finding.

Thus, the H5D 50c that I had produced dramatically better color ramps than the Sony Rii that I am now using.
There is no contest between the two.  I find it really shows up when I convert from color to B&W, which is my normal color space, and they go in and play with the color channels to emphasize the clouds, sky, etc.  Frankly, the ramp from the Sony demonstrates a "banding" look, not smooth "creamy" steps of grays.

It has to be that Hassie has incorporated on their processor the ability to grab the color values from the sensor and up rev them to probably a linear 16 bit data set. 

Thus, it leads me to question, in the absence of what I suspect is Hassies up raving the color space, does the Fuji image produce as good of a base of color for my conversions to B&W?

I need to get my mitts on a RAW set of Fuji landscape shots that have a lot of Dmax.  Can take them and twist and pull like I enjoy is LR6 to see if they hold up like the Hassie does.

Jack

[scyth]

Thus, the H5D 50c that I had produced dramatically better color ramps than the Sony Rii that I am now using.

how did you test exactly to separate a specific raw converter and "camera profile" from what firmware writes in the file (compressed raw/uncompressed raw) ? one 'd assume you used a tool like RawDigger to check per channel raw data before any demosaicking/color transform - no ? just curiousssss ... there is no dispute that larger sensor can collect more light (so you can get better S/N) and that CFA are different and H might have them "better" discriminating

[Lust4Life]

Yes, I use RAW Digger.

But my opinion is taken from actually pushing images with LR6, etc. to get the conversion quality to B&W that I demand for my work.

As said, I push the color image from the Sony and harshness sets in as hyper contrast that can't be extracted without destroying the balance of the file.  Also banding will appear when you zoom in 100%.

Make sense?

[scyth]

Also banding will appear when you zoom in 100%.

banding upon conversion to B/W in LR, right ?

[Lust4Life]

Yes, Lightroom/LR6.8
I've used other methods of converting to B&W but LR is my current favorite.

[Doug Peterson]

I am trying to define without question that the X1D is providing a 16 bit color ramp direct from the array/sensor or is it up-reved in the camera from 14 bit.

The X1D sensor is 14 bit. Without question.

As others point out it is the same Sony 50mp sensor as the others.

[Lust4Life]

"I want to review the actual spect sheets on the sensors."

As stated in my opening.

Jack

[eronald]

They all use the same 50mp sensor P1, IQ350 X1D and Fuji. IMX 161. Over 2 years old now.

There is currently one one physical sensor. The fact the Hasselblad is pulling 16bit info from a 14bit sensor seems to be similar to the claims that the older CCD sensors were 16 bit. Which they epwere not. Hasselblad may be adding something in the processing of the data which is unique to their files only they know.

The only 16 bit MF sensor is the 100mp sensor and P1 does have a unique solution that provides 16 bit data some climate 15 bit.

As for real data on the IMX161 you will not find much real spec sheet data out there on it.

Paul Caldwell

The P1 50 might have a different CFA

[Paul2660]

fuji claims also to have a different CFA, but I don't believe that has anything to do with 14 or 16 bit output does it?

Paul Caldwell

[HBIEVP]

Jack,

The X1D-50c sensor provides 14 bit data but the color space is defined in 16 bit. This doesn’t add any information but ensures that we don’t lose anything in all transformations/calculations performed on the sensor data. The X1D-50c data sheet only describe the Color Definition as 16 bit depth.

Eric P.

[scyth]

The X1D-50c sensor provides 14 bit data but the color space is defined in 16 bit.

such input device (camera or scanner) does not have any "color space"

This doesn’t add any information but ensures that we don’t lose anything in all transformations/calculations performed on the sensor data.

you can have 8 bit raw data recorded in raw file and then 32+ bit floating point processing of it in software afterwards...