Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: Lust4Life on September 02, 2019, 09:39:54 am
-
I've not found any posting on the ACTUAL number of color bits that the IV generates in RAW.
Want not just what Sony says, but what something like RAW Digger states.
Anyone compare this to our world of MF?
Jack
-
Best to simply email Illiah, he'll confirm ASAP, he's that kind of follow. I don't want to post his email publicly so if you don't have it, PM me.
-
Best to simply email Illiah, he'll confirm ASAP, he's that kind of follow. I don't want to post his email publicly so if you don't have it, PM me.
PM sent. I have no idea who Illiah is.
-
i believe he had a little something to do with Raw Digger and related software. You can Search on his name in the LULA forums.
-
I think my question may not be clear.
I am trying to learn about the Sony aR7 IV camera and does IT provide a true 16 bit RAW color file.
-
i believe he had a little something to do with Raw Digger and related software. You can Search on his name in the LULA forums.
He is the author of RD and FastRawViewer.
-
I think my question may not be clear.
I am trying to learn about the Sony aR7 IV camera and does IT provide a true 16 bit RAW color file.
Probably not but Illiah knows.
-
I've not found any posting on the ACTUAL number of color bits that the IV generates in RAW.
Want not just what Sony says, but what something like RAW Digger states.
Anyone compare this to our world of MF?
Jack
You are better off asking questions about a7rIV in the "Mirrorless Cameras" instead of the "Medium Format/..." forum.
Sony's raw format support 12-bit and 14-bit formats. No change to the raw format has been announced.
-
You are better off asking questions about a7rIV in the "Mirrorless Cameras" instead of the "Medium Format/..." forum.
Sony's raw format support 12-bit and 14-bit formats. No change to the raw format has been announced.
Craw is 13 bits near the origin, and less further up the tone curve.
-
I've not found any posting on the ACTUAL number of color bits that the IV generates in RAW.
Want not just what Sony says, but what something like RAW Digger states.
Here's an excerpt from the EXIF of an uncompressed raw shot:
File Type : ARW
File Type Extension : arw
MIME Type : image/x-sony-arw
Exif Byte Order : Little-endian (Intel, II)
Image Description :
Orientation : Horizontal (normal)
Image Width : 9600
Image Height : 6376
Bits Per Sample : 14
Compression : Uncompressed
Photometric Interpretation : Color Filter Array
Samples Per Pixel : 1
Planar Configuration : Chunky
X Resolution : 350
Y Resolution : 350
Resolution Unit : inches
CFA Repeat Pattern Dim : 2 2
CFA Pattern 2 : 0 1 1 2
Sony Raw File Type : Sony Uncompressed 14-bit RAW
Here's a histogram:
-
Hi Jim - that is EXACTLY what I was hoping to see!
Thank you.
Now, to get the same thing on a GFX 100! Bernard - are you listening?? :-)
I don't want to start a "range war" here but my personal experience and test do push me in favor of a 16 bit color file. As I do a lot of "pushing" of the channels when converting to B&W, which is the historical majority of my work until just the last year or two, I find a distinct preference for a 16 bit file.
Best,
Jack
-
I don't want to start a "range war" here but my personal experience and test do push me in favor of a 16 bit color file.
So you'll be hard pressed to find many true 16-bit capture devices nor is it really necessary.... Keep in mind too, when you think you're editing 16-bit in Photoshop, you never are. Actually, it's 15 bits plus one level...done for algorithmic processing reasons. And since there really isn't a real life source of full 16 bit images, that's all the precision Photoshop needs.
-
So you'll be hard pressed to find many true 16-bit capture devices nor is it really necessary.... Keep in mind too, when you think you're editing 16-bit in Photoshop, you never are. Actually, it's 15 bits plus one level...done for algorithmic processing reasons. And since there really isn't a real life source of full 16 bit images, that's all the precision Photoshop needs.
All true, BUT I have played with 14 bit files and used a rather sophisticated Up-Res routine back when I was developing on SGI platform - the 16 bit ramp produced a MUCH smoother gradient in the skys, etc.
Hey, I just have my "dreams".
Have actually thought of developing a little program to offer conversions from 14 to true 16 bits, but doubt there is much market for it.
-
Christ on a crutch, who gives a fiddler's fart?
-
It's comments like that which prove you and I have nothing of value to each other.
-
All true, BUT I have played with 14 bit files and used a rather sophisticated Up-Res routine back when I was developing on SGI platform - the 16 bit ramp produced a MUCH smoother gradient in the skys, etc.
Hi Jack,
That's somewhat to be expected, if the data has very little noise. But there's also a role to be played by the Raw converter. After all, the 14-bit Raw data has yet to be demosaiced, and that can/should be done in 16-bit per Channel, or even in floating-point numbers (reals instead of integers).
Hey, I just have my "dreams".
Have actually thought of developing a little program to offer conversions from 14 to true 16 bits, but doubt there is much market for it.
For a commercial product the market size an issue, but it's also possible to contribute to an initiative like e.g. RawTherapee.
Cheers,
Bart
-
Hi Jack,
That's somewhat to be expected, if the data has very little noise. But there's also a role to be played by the Raw converter. After all, the 14-bit Raw data has yet to be demosaiced, and that can/should be done in 16-bit per Channel, or even in floating-point numbers (reals instead of integers).
For a commercial product the market size an issue, but it's also possible to contribute to an initiative like e.g. RawTherapee.
Cheers,
Bart
Bart, you have several good points there.
I have tried several of the converters but have not found one that just clicked with me.
Given: Luminar is of no interest to me, any more than chewing bubble gum at dinner is.
Lightroom (CS6) is getting a little old for me. (I do not like "subscription software, whether it is Cloud based or not so CS6 is where I get off the Adobe train.)
I have just again updated my C1 software for the GFX as they had a promotion and I had not played with it in a couple of years. But, I don't find that exactly what I want either.
Phocus from Hassie - did use that for many years back when I was shooting the H series, up through the H5D 50.
You have a great point on RAWTherapy - I tried the free trial but lost almost all of the trial time to a health issue that flared up! I think it had some potential but a steep learning curve if one wants to master all it offers. Do you feel that is a valid assessment of it?
Jack
-
Bart, you have several good points there.
I have tried several of the converters but have not found one that just clicked with me.
Given: Luminar is of no interest to me, any more than chewing bubble gum at dinner is.
Lightroom (CS6) is getting a little old for me. (I do not like "subscription software, whether it is Cloud based or not so CS6 is where I get off the Adobe train.)
I have just again updated my C1 software for the GFX as they had a promotion and I had not played with it in a couple of years. But, I don't find that exactly what I want either.
Phocus from Hassie - did use that for many years back when I was shooting the H series, up through the H5D 50.
You have a great point on RAWTherapy - I tried the free trial but lost almost all of the trial time to a health issue that flared up! I think it had some potential but a steep learning curve if one wants to master all it offers. Do you feel that is a valid assessment of it?
IMHO, Capture One and RawTherapee have the best demosaicing algorithms. RawTherapee is an amalgam of interesting techniques, which makes the learning curve quite steep, whereas C1 is very customizable and very powerful under-the-hood.
'Affinity Photo' is interesting as an alternative to Photoshop, but it also has a Raw conversion engine that can demosaic in 32-bit floating point format. It's not quite as good as the C1 or RT demosaicing, but the 32-bit conversions can be helpful with other aspects of the postprocessing.
Cheers,
Bart
-
Bart,
Sounds like we are on the same page about C-1 and RAWTherapee.
I just have not had time this summer to digest more than one interface and set of tools so I choose C-1 with the 1/2 price offer they made for users of the GFX.
Will take another look at RT over the winter if all goes well.
Again, thanks for input here.
Jack
-
Now, to get the same thing on a GFX 100!
Even better, pictures:
https://blog.kasson.com/gfx-100/visual-comparisons-of-fuji-gfx-100-14-and-16-bit-raw-precision/
-
Jim,
Just read the link - decent study BUT I must say that I see a difference in:
1. Ever so slight of an "apparent" sharpness increase with the 16 bit over the 14 bit. Could be what the little bit extra of data provides in each colors ramp.
2. Again, to my eye I see a slight "glow" to the 16 bit file that the 14 does not. This is similar to what I would see on occasion back decades ago when I was working in the development of the Eikonic camera (which Kodak killed in hopes of stalling the digital photography transition.) - we finally attributed it to the ever so slight current changes coming out of the wall.
Jack
-
Jim,
Just read the link - decent study BUT I must say that I see a difference in:
1. Ever so slight of an "apparent" sharpness increase with the 16 bit over the 14 bit. Could be what the little bit extra of data provides in each colors ramp.
2. Again, to my eye I see a slight "glow" to the 16 bit file that the 14 does not. This is similar to what I would see on occasion back decades ago when I was working in the development of the Eikonic camera (which Kodak killed in hopes of stalling the digital photography transition.) - we finally attributed it to the ever so slight current changes coming out of the wall.
Jack
The reason that I post pictures in tests like this is to allow readers to draw their own conclusions.
-
Come on now, do you see what I am describing - very slight but there in my view.
To my eye, like the slight difference in "warmth" between a Sony a7R III and the Hassie 4D WiFi 50MP I had.
-
Come on now, do you see what I am describing - very slight but there in my view.
To my eye, like the slight difference in "warmth" between a Sony a7R III and the Hassie 4D WiFi 50MP I had.
The long pole in the tent is that neither push is usable without extensive post-production because of the PDAF banding, and whatever differences there are due to precision remain unimportant because of that.
Jim
-
The long pole in the tent is that neither push is usable without extensive post-production because of the PDAF banding, and whatever differences there are due to precision remain unimportant because of that.
Jim
Some people might say this is irrelevant because the non-pdaf pixels are what gets measured so the camera meets its specs :)
Others might say that what counts is the practical ability of the camera which appears as you say to be severley constrained by the PDAF banding; interestingly, if the firmware were to fix the banding we would then need some other measuring procedure to detect and quantify the quality of this information synthesis :)
If I were cynical I would say that the messy PDAF pixels are there because someone used existing software to declare the design met its specs.
-
Some people might say this is irrelevant because the non-pdaf pixels are what gets measured so the camera meets its specs :)
Others might say that what counts is the practical ability of the camera which appears as you say to be severley constrained by the PDAF banding; interestingly, if the firmware were to fix the banding we would then need some other measuring procedure to detect and quantify the quality of this information synthesis :)
If I were cynical I would say that the messy PDAF pixels are there because someone used existing software to declare the design met its specs.
I think the banding is there because the firmware is trying to correct for PDAF striping. Look at the a7RIII: striping, but no banding. I think that banding is a self-inflicted injury.
Jim
-
I think the banding is there because the firmware is trying to correct for PDAF striping. Look at the a7RIII: striping, but no banding. I think that banding is a self-inflicted injury.
Not that it would affect the Raw data itself (assuming it's a firmware issue), but if the PDAF 'adjustment' is consistent, the Capture One's LCC correction functionality should allow much cleaner Conversions which can be pushed further without (many of) the visually striking artifacts.
Cheers,
Bart