In GFX 50S M2 I would expect better ergonomics and better EVF.
The GetDPI Photography Forum
Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
Fuji GFX 100
- Thread starter Christopher
- Start date
k-hawinkler
Well-known member
As an owner of a 50S not very enticing.
More to my liking the GFX 100, while also keeping the 50S. :thumbs:
dougpeterson
Workshop Member
Once a curve, dodge, burn, or highlight/shadow recovery tool is used.
24bit via LUT https://imagescience.com.au/products/monitors/eizo-coloredge-cg319x-4k
[the 10 bit mentioned is the serial digital signal for broadcast BVM monitors]
[the 10 bit mentioned is the serial digital signal for broadcast BVM monitors]
Digital sensors only detect brightness information. The typical Bayer array adds green, blue and red filters in front of each sensor pixel. The other two color channels are interpolated from the colors of the neighbour channels. So, in a certain way, the DR is for example 14 bit brightness, but converted/interpolated to 3x 14 bit color information.
The 24 bit via LUT mentioned here is only 3x 8 bit (isn't it?).
So again: how to really see the DR of a raw image, if not first compressed?
I get your point Doug, if a curve is applied, that theoretically could make a difference. But the compression or cut-off from the raw image to the DR of a monitor or print is so tremendous that I really have doubts that on a on a 8/10 bit DR monitor or a 5 bit DR print the difference between a 14 bit DR sensor and a 16 bit (which actually mostly is only 15 bit real DR) sensor can be seen. Can you show comparison samples?
dougpeterson
Workshop Member
Output dynamic range is almost irrelevant to this conversation. The only question is what range of contrast is in the original scene, and what parts of that scene you wish to contain detail (as opposed to being clipped black or clipped white), and only in situations where you can't control the lighting contrast (e.g. you can't turn off the sun in a landscape or add a fill card like you can in the studio). One can demonstrate this even with 1 bit output (black or white only) provided you can use a bit of dithering. If you record it in the scene, you can represent it in the print. Of course, whether it is aesthetically useful to do so depends on the scene, the artist, and the intention of the work; a drive to always include all tonal range from a scene is a great way to get very low contrast muddy looking prints, or worse, garish HDR-looking images.
This isn't just theory, but can be seen in real-world images, albeit only in problematic scenes (e.g. contrasty light). We have 14 bit vs 16 bit raw file tests we've done with the IQ3 100mp in central park a couple years ago that shows it quite clearly. I'd be stepping on this thread to post them here, but feel free to email me.
JimKasson
Well-known member
This depends entirely on the tone curve. If linear, 5 bits corresponds to a print Dmax of -log10(2^-5) = 1.5 at SNR = 1.
Jim
JimKasson
Well-known member
Agree. In a practical sense, not completely irrelevant, but certainly not very important. That was the basis of my comments about scene-referred vs output-referred files. (That was here, right? [Yes, it was; I see it at the bottom of this page.]).
Also you can't compare the precision of linear files with those that have tone curves applied.*
Jim
*Nerdy clarification: you can do such comparisons if you take into account the first derivative of the tone curve and its effect on precision when mapped back to linear.
Last edited:
dougpeterson
Workshop Member
Jim, you're one of my very favorite posters here. You're always accurate and practically informative without being demeaning, oversimplifying, or pedantic*– a rare ability for someone who knows so much of the technical nitty gritty. If you're ever in NYC I'd be very glad to buy you a beer.
My 2c, as someone who has shot a lot of high quality images on Phase backs. For me 16bit worked at 100asa only, and was just just noticeable in darkest shadow discrimination (around 1-8 of 255). And I mean only noticeable by extreme pixel peeping. Anything without deepest shadow detail, I couldn't see any difference at all, compared to 14bit.
Choosing 16bit brought slower shutter speeds, longer writes (not ready for next capture =missed images) and less storage on cards. I settled happily on 14 bit, especially as 200asa is my default (one extra stop: 1/250 vs 1/500, when handheld) and once you move off base ISO, 16bit was redundant, as DR dropped.
Of course my needs are not yours, so YMMV.
Choosing 16bit brought slower shutter speeds, longer writes (not ready for next capture =missed images) and less storage on cards. I settled happily on 14 bit, especially as 200asa is my default (one extra stop: 1/250 vs 1/500, when handheld) and once you move off base ISO, 16bit was redundant, as DR dropped.
Of course my needs are not yours, so YMMV.
ErikKaffehr
Well-known member
Hi,
That is a good question...
The way I see it we may have a great dynamic range. Than we have the task of converting that range to something we can show on screen or print.
Here is an image I am working on right now:
Dynamic range is great and a piece of magic light. But how can I fit in in the DMax of matte paper?!
Best regards
Erik
It’s my understanding that the 16bit Capture with a P1 back is really advantageous at base ISO. I might have it wrong as very little has been published on it. For sure it slows down the Capture with the ES on both 3100 an IQ4. I am hoping that with the IQ4 the 16 bit Capture is still useable past 50.
Still seems like the thing to use when conditions allow.
Paul C
Still seems like the thing to use when conditions allow.
Paul C
k-hawinkler
Well-known member
When taking an image I naively assume there must be a number of more important considerations than worrying about 14 or 16 bit capture, no? :facesmack:
faberryman
Member
Like a lot of menu choices, choosing 14 or 16 bit is a set and forget operation so you don't have to think about it once you have made your decision.
Erik - that is a lovely lovely image, I think the best you've posted. Congrats!Hi,
That is a good question...
The way I see it we may have a great dynamic range. Than we have the task of converting that range to something we can show on screen or print.
Here is an image I am working on right now:
Dynamic range is great and a piece of magic light. But how can I fit in in the DMax of matte paper?!
Best regards
Erik
k-hawinkler
Well-known member
+1, great light, colors, and composition. :thumbs: