GFX100S 14-bit vs 16-bit

[fcracer]

I was surprised to see the 16-bit option preselected on my 100S and started to research what the real world differences would be of using 14-bit.

As usual, Jim’s blog has the answers: https://blog.kasson.com/gfx-100/visual-comparisons-of-fuji-gfx-100-14-and-16-bit-raw-precision/

It would seem that 14-bit has no real downside and lots of upsides like file size and being able to use more drive modes.

Has anyone had a situation where the 16-bit made a tangible difference? Would love to hear your thoughts.

--
How to rescue an image
Capture One Pro quick workflow
Travel and photography blog: fcracer.com

 

[Erik Kaffehr]

I was surprised to see the 16-bit option preselected on my 100S and started to research what the real world differences would be of using 14-bit.

As usual, Jim’s blog has the answers: https://blog.kasson.com/gfx-100/visual-comparisons-of-fuji-gfx-100-14-and-16-bit-raw-precision/

It would seem that 14-bit has no real downside and lots of upsides like file size and being able to use more drive modes.

Has anyone had a situation where the 16-bit made a tangible difference? Would love to hear your thoughts.

Hi,

In theory, there is no advantage to 16-bit files vs 14-bit as long as engineering DR at the pixel level is below 14.5 EV.

Historically, MFD was claimed to have 16-bits, but that was mostly fake. The 100 MP 54x41 mm sensor did reach 14EV in engineering DR, so 16 bit may have been useful. So Phase One 'developed' a new raw format that was actually 16 bit wide.

So, 16 bits is about old hype. But, modern sensors are on the verge to needing more than 14 bits. But, they are not there.

Here is the simple formula calculating the number of bits needed:

bits = log(FWC/RN)/log(2)

The formula is a bit on the conservative side.

GFX 100 has:

FWC (Full Well Capacity) = 41,534 at 100 ISO

RN (Read Noise) = 3.411 at 100 ISO

Both these numbers are given normally in the units of electron charges. I took my numbers from Bill Claff.

>> log( 41534 / 3.411) /log(2)

ans =

13.5718

Just to say, Bill Claffs data has RN = 2.282 for ISO 50 and that would correspond to 14.2 EV.

Best regards

Erik

--
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

 

[fcracer]

Wow, thanks for the detailed reply Eric! It makes me even more annoyed that Fujifilm has it set to 16-bit by default. Besides marketing, I can see no other reason for them to do that.

 

[bclaff]

...

Has anyone had a situation where the 16-bit made a tangible difference? Would love to hear your thoughts.

I keep forgetting to put a link to this on the main page at PhotonsToPhotos:

Read Noise in DN versus Bit-depth

According to what I have here the Phase One IQ250 and Phase One IQ3 100MP take advantage of 16-bit.

 

[JimKasson]

I was surprised to see the 16-bit option preselected on my 100S and started to research what the real world differences would be of using 14-bit.

As usual, Jim’s blog has the answers: https://blog.kasson.com/gfx-100/visual-comparisons-of-fuji-gfx-100-14-and-16-bit-raw-precision/

It would seem that 14-bit has no real downside and lots of upsides like file size and being able to use more drive modes.

Has anyone had a situation where the 16-bit made a tangible difference? Would love to hear your thoughts.

I don’t have my GFX 100S yet but it looks from early testa that that camera doesn’t have PDAF banding. I will rerun that test was the 100S.

 

[Greg7579]

Jim, I have been shooting lossless 14 bit since day one, but I'm looking on my metadata on Adobe Bridge and it says bit depth 16 on all my files. What does that mean? My raw files are about 100 MB instead of the 200 or 250 if they were uncompressed 16 bit.

 

[JimKasson]

Jim, I have been shooting lossless 14 bit since day one, but I'm looking on my metadata on Adobe Bridge and it says bit depth 16 on all my files. What does that mean? My raw files are about 100 MB instead of the 200 or 250 if they were uncompressed 16 bit.

You sure you're not looking at the precision after demosaicing?

 

[bclaff]

Jim, I have been shooting lossless 14 bit since day one, but I'm looking on my metadata on Adobe Bridge and it says bit depth 16 on all my files. What does that mean? My raw files are about 100 MB instead of the 200 or 250 if they were uncompressed 16 bit.

Remember, lossless is not the same as uncompressed; lossless will be smaller than uncompressed.

 

[TheGrilledCheese]

There is horizontal banding here, more so in the 16-bit file. I thought folks said there was no banding with the GFX 100S. From these images it is clearly there.

 

[JimKasson]

There is horizontal banding here, more so in the 16-bit file. I thought folks said there was no banding with the GFX 100S. From these images it is clearly there.

Where is here? Do you have a link?

 

[Greg7579]

Jim, I have been shooting lossless 14 bit since day one, but I'm looking on my metadata on Adobe Bridge and it says bit depth 16 on all my files. What does that mean? My raw files are about 100 MB instead of the 200 or 250 if they were uncompressed 16 bit.

Remember, lossless is not the same as uncompressed; lossless will be smaller than uncompressed.

Bill, I'm an idiot and I haven't thought about this in a while. I set my GFX 100 on lossless 14 bit the day I got it and haven't looked back. That seemed to be the consensus. It is still set there.

I'm not sure I understand the difference between the three menu choices:

Uncompressed (That is obviously the full-blown raw file and would be big.)

Lossless (Which I thought was really "Lossless Compressed" and defacto choice.)

and Compressed (Which I thought was the smallest but had some loss).

I will check the manual and research it a bit, so sorry to ask.

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

 

[Greg7579]

There is horizontal banding here, more so in the 16-bit file. I thought folks said there was no banding with the GFX 100S. From these images it is clearly there.

Where is here? Do you have a link?

Here is down the rabbit hole and straight into the dark reaches of Hell.

 

[bclaff]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

 

[SrMi]

Jim, I have been shooting lossless 14 bit since day one, but I'm looking on my metadata on Adobe Bridge and it says bit depth 16 on all my files. What does that mean? My raw files are about 100 MB instead of the 200 or 250 if they were uncompressed 16 bit.

Remember, lossless is not the same as uncompressed; lossless will be smaller than uncompressed.

Bill, I'm an idiot and I haven't thought about this in a while. I set my GFX 100 on lossless 14 bit the day I got it and haven't looked back. That seemed to be the consensus. It is still set there.

I'm not sure I understand the difference between the three menu choices:

Uncompressed (That is obviously the full-blown raw file and would be big.)

Lossless (Which I thought was really "Lossless Compressed" and defacto choice.)

and Compressed (Which I thought was the smallest but had some loss).

I will check the manual and research it a bit, so sorry to ask.

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

[Deleted, wrong comment]

 

[Greg7579]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

Bill, when I look at the metadata on Adobe Bridge 2021 (the latest version) the Metadata on all my GFX 100 files that I have shot says 16 bit on every one of them.

When I check my recently shot Leica Q2 files, they all say 14 bit.

 

[bclaff]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

Bill, when I look at the metadata on Adobe Bridge 2021 (the latest version) the Metadata on all my GFX 100 files that I have shot says 16 bit on every one of them.

When I check my recently shot Leica Q2 files, they all say 14 bit.

I don't trust that; let's cross-check with something like Exiftool.

 

[Greg7579]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

Bill, when I look at the metadata on Adobe Bridge 2021 (the latest version) the Metadata on all my GFX 100 files that I have shot says 16 bit on every one of them.

When I check my recently shot Leica Q2 files, they all say 14 bit.

I don't trust that; let's cross-check with something like Exiftool.

OK, I'll download it tomorrow. But I just dug in the data a loittle deeper and clicked on raw data and it had reams of data.... In one place it said bits per sample - 16....

 

[JimKasson]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

Bill, when I look at the metadata on Adobe Bridge 2021 (the latest version) the Metadata on all my GFX 100 files that I have shot says 16 bit on every one of them.

When I check my recently shot Leica Q2 files, they all say 14 bit.

I don't trust that; let's cross-check with something like Exiftool.

OK, I'll download it tomorrow. But I just dug in the data a loittle deeper and clicked on raw data and it had reams of data.... In one place it said bits per sample - 16....

You can use RawDigger:







--

the last word the last word - Photography meets digital computer technology. Photography wins -- most of the time.

Photography meets digital computer technology. Photography wins -- most of the time.

blog.kasson.com

 

[bclaff]

...

But yes, my metadata says 16 bit and I have my camera set on Lossless - 14 bit.

Who's reporting this metadata?

For example, what does Exiftool say?

Bill, when I look at the metadata on Adobe Bridge 2021 (the latest version) the Metadata on all my GFX 100 files that I have shot says 16 bit on every one of them.

When I check my recently shot Leica Q2 files, they all say 14 bit.

I don't trust that; let's cross-check with something like Exiftool.

OK, I'll download it tomorrow. But I just dug in the data a loittle deeper and clicked on raw data and it had reams of data.... In one place it said bits per sample - 16....

You can use RawDigger:

Sure. RawDigger simply uses Exiftool in a separate process

--
Bill ( Your trusted source for independent sensor data at PhotonsToPhotos )

 

[Greg7579]

Thanks Jim. Just checked raw digger and it is 14. So why would Adobe say it is 16? Very odd.....