Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Fuji Forum | Join Upload & Sell

1
       2       3       4       5       end
  

How big do you need to print to see the advantage of a 100 MP 44 X 33 se...

  
 
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #1 · p.1 #1 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Almost everyone acknowledges that the bigger 44 X 33 mm 100 MP sensor in the GFX cameras produces better results than even the best of FF 35mm cameras. An important question, however, is whether those better results really matter and can actually be seen in the photos people shoot. One question related to that issue is how large do you need to print to generalize see the advantage of the larger sensor? One member here G. Dan Mitchell has repeatedly claimed that you would need to print at least a 30 X 40 print to see the difference.

Dan Wells at Luminous Landscapes has extensively compared prints and even shown them to non-photographers and he comes to a very different conclusion. He writes:

"The central question of this part is whether that difference is visible in prints, and, if so, how big you have to print in order to see it.

More details further on in the article, but the simple answer is that, to my eyes and those of several non-photographer observers, the difference is visible even in 8x10” prints compared to the best of APS-C files (Fujifilm’s own X-T4 – also a Highly Recommended camera, although for a different job), and by 12x18” prints even against the best of full-frame digital, the Sony A7r IV."

The whole article can be found here, but much of it is behind a paywall:

https://luminous-landscape.com/the-proof-is-in-the-printing-part-iii-of-a-slew-of-similar-sony-sensors/

To me this is an important question. If Dan Mitchell is correct and differences can only be seen with prints as large or larger than 30 X 40, then the advantage of the larger format sensor is quite limited, IMO. In contrast if Dan Wells is right and differences can be seen with 12 X 18 prints, which would include almost all my prints, then the advantage is quite substantial, IMO.

What is your own experience if you have shot both a modern FF 35mm camera and the Fuji GFX 100 or 100s? Personally, I only have the 50s and I find the advantage primarily to be that is shoots in my preferred aspect ratio or close for little cropping, but I suspect that Dan Wells might be more accurate in his description of the differences and the advantage of the 100MP sensor. I am currently deciding whether to upgrade to the GFX 50s II or the 100s, so this is a really relevant question for me.



Jun 13, 2022 at 09:42 AM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #2 · p.1 #2 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


The pixel aperture of the GFX 100x and GFX 50x is the same when measured in micrometers. The sensor's effect on sharpening is the same on a per-um basis for both cameras. The sensors are the same physical size, so the sensor's effect on sharpening is the same on a per-picture-height basis for both cameras. The difference is that the GFX 100x image shows less aliasing. Since aliased frequencies go all the way to dc, there is no practical lower limit of print size below which the GFX 100x and GFX 50x images will look the same.

So the answer to your question comes down to the choice of subject matter, and photographic technique.

Jim

Edited on Jun 13, 2022 at 06:46 PM · View previous versions



Jun 13, 2022 at 10:21 AM
molson
Offline
• • • • • •
Upload & Sell: On
p.1 #3 · p.1 #3 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
If Dan Mitchell is correct and differences can only be seen with prints as large or larger than 30 X 40, then the advantage of the larger format sensor is quite limited, IMO.


Are you referring to the infamous blurry bush snapshots, taken with a borrowed camera, solely for the purpose of confirming a personal bias against a camera somehow deemed not worthy of the term "medium format"?



Jun 13, 2022 at 10:37 AM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #4 · p.1 #4 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


molson wrote:
Are you referring to the infamous blurry bush snapshots, taken with a borrowed camera, solely for the purpose of confirming a personal bias against a camera somehow deemed not worthy of the term "medium format"?


Well the claims were made long before Dan Mitchell did that "test." I may be the only person who actually printed the images from the "test" and I was able to tell the difference between the cameras pretty clearly, but it was an awful "test," IMO. Very flat light, very limited subject that one would never shoot, and was not really informative at all, IMO. Still the claims by Dan Mitchel have been made repeatedly both before and after his "test," and I wonder what other people's experience has been.



Jun 13, 2022 at 11:20 AM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #5 · p.1 #5 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
Still the claims by Dan Mitchel have been made repeatedly both before and after his "test," and I wonder what other people's experience has been.


I can often see differences in C-sized prints. But it's highly subject-related. For example, if you shoot fabrics, you're going to see a lot of difference on some images. With natrual subjects, sometimes the aliasing is harder to see, but the false color is a giveaway.




Jun 13, 2022 at 11:34 AM
Makten
Offline
• • • • •
Upload & Sell: On
p.1 #6 · p.1 #6 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Not gonna dive into the resolution debate since I don't print, but there are some other things worth considering...

– Both the 50S II and 100S are smaller than the 50S (edit: if you take the "rear hump" into account).
– Both the 50S II and 100S have IBIS (huge advantage IMO).
– The 100S gives lower noise and better DR.
– The 100S has probably (hard to quantify) much better AF.
– The 100S has a little bit "different" colors (not necessarily better, depending on taste).
– The 100S gives roughly twice the file size (50 is bad enough, IMO!).

Personally I never even considered the 100S when upgrading from 50R. I only wanted IBIS, and absolutely not huge files.
Now I might have changed my mind solely because of the almost useless AF of the 50S II. I'm sure it's a software related issue, because 9 out of 10 times it focuses on the farthest object within the focusing area. Not the closest. This was not the case with the 50R. Flat or large objects, fine. But try focusing on a branch or something similar. The camera will focus on the background, every single time.



Jun 13, 2022 at 11:38 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
Almost everyone acknowledges that the bigger 44 X 33 mm 100 MP sensor in the GFX cameras produces better results than even the best of FF 35mm cameras. An important question, however, is whether those better results really matter and can actually be seen in the photos people shoot. One question related to that issue is how large do you need to print to generalize see the advantage of the larger sensor? One member here G. Dan Mitchell has repeatedly claimed that you would need to print at least a 30 X 40 print to see the difference.

Dan Wells
...Show more

Back when I was considering the GFX 50S, I did a series of tests along with a friend who owned the camera, shooting boring subjects under relatively controlled conditions side-by-side with a Canon 5DsR.

Afterwards I produced print-ready versions of test image pairs from both cameras at sizes up to 60" x 80". I created sets of three images in which sample A came from one of the cameras, sample B came from the other, and sample X came from either of the two, but which of them was not known to the test participants. (In other words if A was the Fujifilm and B was the Canon, X could have been either of those.) Each set randomized which camera was A, B, and X.

In the first test I handed out sets of three prints to a group of fellow photographers. (The group included one who won a series of awards for his book of photography documenting changes in a particular Asian culture, and another member who was a protege of Ansel Adams and who has been represented by the Ansel Adams gallery for decades, and who has also written for Luminous Landscape.) The three-print sets I handed them were letter sized crops of the full size images. I asked them to respond to several prompts: What differences did they observe between A and B? Did they prefer A or B? Did X come from the same source as A or B?

The purpose of the A/B/X test is to determine whether expressed preferences represent a verifiable ability to discriminate between A and B or are random.

Basically, the test design allows us ask three questions: Do participants detect differences between samples? Do their observations of differences reflect a better/worse perception or a not-the-same perception? Are their perceptions of differences consistent and persistent?

I later conducted the same test with two other groups using an online format. Here participants were able to download the print-ready cropped files from the various full print sizes and print them on their own printers and compare in the same manner as above. Separately, those without printers were able to participate solely on the basis of looking at screen images. (I consider that latter to be inferiors and far less meaningful.)

In all cases up to. a 30" x 40" print size, there was no consistent preference between the 5DsR-sources prints and the 50S-sourced prints. Many stated that there was no apparent better/worse difference among the prints, and among those who expressed a preference, it was split between the two. The "X" test also failed to support the idea that those with a preference were consistently observing a real difference.

At the 30" x 40" size we began to see a very small difference in preference that was potentially outside the range of expected random variation. So it is possible but not proven that a difference that is visible to a small number of very careful observers doing side-by-side comparisons might start to appear at about that size. This led me to hold that "at print sizes of 30" x 40" and above there many be some detectible differences." (I'd add that these differences are not "night and day," but subtle differences between two very good examples.)

If I had access to a 100MP miniMF system and newer Fujifilm lenses than we had at the time, it would be fascinating to do the test again and pit the 100s against the 5DsR and the high MP systems from the other FF manufacturers.

Knowing what I know about comparisons between other systems with different MP resolutions, I suspect that there would be some difference above that 30" x 40" size, but how large would it be, when would it be visible, and to whom?

In order of the test to be meaningful, it should really be constructed similarly to what I did. I did know which cameras had been used for each image, but I took myself out of the process of evaluation the images in order to avoid letting my own biases/preferences/motivated reasoning affect the evaluations. All of the evaluations were done by others who did not even know what the test was about (in my initial print-based test), much less what cameras had been used for the samples.

Perhaps someone is up for such a test? Hint: it is not an easy process, you'll need to go to some lengths to remove the potential for preexisting biases to affect the results, and you'll need to be prepared for attacks by folks who don't like the outcomes! ;-)

Dan

(I have to say that I'm extremely skeptical of claims that observers would be able to consistently select as the best sample an image from a 100s in a comparison of 8" x 10" prints from a range of camera formats including MFT, APS-C, FF, miniMF, MF, and LF in a blind or double-blind test.)



Jun 13, 2022 at 12:26 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #8 · p.1 #8 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


JimKasson wrote:
The pixel aperture of the GFX 100x and GFX 50x is the same when measured in micrometers. The sensor's effect on sharpening is the same on a per-um basis for both cameras. The sensors are the same physical size, so the sensor's effect on sharpening is the same on a per-picture-height basis for both cameras. The difference is that the GFX 100x image shows more aliasing. Since aliased frequencies go all the way to dc, there is no practical lower limit of print size below which the GFX 100x and GFX 50x images will look the same.

So the answer
...Show more

Hi Jim,

Thanks for responding. If I understand you rightly (which I may not be), then you are saying there are no differences in sharpness between the 50x and 100x sensor, just the 100x files have more aliasing, even if you downsize the files to the same size. I don't really care about sharpness that much, what I see as the potential advantages of the 100x files is the higher DR especially at higher ISO. I can tolerate aliasing (I have a Leica M10, so that should tell you something about my toleration for aliasing), but I would prefer not to have it. I generally see more DR as a good thing, however. Plus, I would prefer the better AF of the 100s.

All that said, I plan to use the GFX mostly for portraits and with manual focus much of the time. I plan to use the Irix 45 f/1.4, Fuji GF 80 f/1.7, and Zeiss 135 f/2 as my primary lenses. I will shoot some macro and close focus as well with the Contax 645 120 f/4 APO which is manual focus as well. So my only AF lens with likely be the 80 f/1.7, but I do want focus with that to be fairly accurate if not fast. Less than ten percent of my shooting with it will be landscapes. I think for this subject matter and lenses I might prefer the GFX 50s II, but I am still considering both cameras.



Jun 13, 2022 at 12:29 PM
molson
Offline
• • • • • •
Upload & Sell: On
p.1 #9 · p.1 #9 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
Well the claims were made long before Dan Mitchell did that "test." I may be the only person who actually printed the images from the "test" and I was able to tell the difference between the cameras pretty clearly, but it was an awful "test," IMO. Very flat light, very limited subject that one would never shoot, and was not really informative at all, IMO. Still the claims by Dan Mitchel have been made repeatedly both before and after his "test," and I wonder what other people's experience has been.


I was being sarcastic, Steve...



Jun 13, 2022 at 12:35 PM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #10 · p.1 #10 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
Thanks for responding. If I understand you rightly (which I may not be), then you are saying there are no differences in sharpness between the 50x and 100x sensor, just the 100x files have more aliasing, even if you downsize the files to the same size.


No, I['m saying that the 100S files will have less aliasing.




Jun 13, 2022 at 12:35 PM
 


Search in Used Dept. 

rdeloe
Offline
• • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


A few thoughts on this...

(1) Double-blind or bust. And make sure to control for printing-related confounding variables, e.g.,

- Subject and scene, including lighting
- Printer, ink, paper
- Post processing, including output sharpening
- Print viewing conditions

(2) In an 8x10 print, an X-T2 will have 500ppi on the 8" edge, a GFX 50S will have 774 ppi, and a GFX 100S 1,092 ppi. I'm skeptical that differences that can be seen reliably and consistently would be due to number of pixels sent to the printer.

(3) I don't need more pixels to make better prints. I need better ideas. Your results may vary.



Jun 13, 2022 at 12:41 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #12 · p.1 #12 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


gdanmitchell wrote:
Back when I was considering the GFX 50S, I did a series of tests along with a friend who owned the camera, shooting boring subjects under relatively controlled conditions side-by-side with a Canon 5DsR.

Afterwards I produced print-ready versions of test image pairs from both cameras at sizes up to 60" x 80". I created sets of three images in which sample A came from one of the cameras, sample B came from the other, and sample X came from either of the two, but which of them was not known to the test participants. (In other words if A
...Show more

The thing is Dan, I participated in that test and I did see consistent differences even at much smaller sizes than you claim and as I have said previously the test was very insensitive to matters that are important to me. The light was very flat, so the scene had very low DR. One of the huge differences between the Canon 5DsR and the Fuji GF 50s is the base ISO DR. Photons to Photos measures the differences in DR at base ISO as just over 2 stops (i.e., the Fuji can handle over 4 times as bright of light). That is of course not a small difference, Yet your test was insensitive to this potential difference because the light was so flat. Let's say I am highly skeptical that with a high DR scene that the Canon 5DsR can keep up with the Fuji 50s, but your test didn't have a chance of capturing this potential difference.

So, that leaves me wondering whether I should care about the much smaller DR difference between the 100x and 50s sensors.



Jun 13, 2022 at 12:42 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #13 · p.1 #13 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


molson wrote:
I was being sarcastic, Steve...


I know I think we share similar views of the value of Dan Mitchell's "test."



Jun 13, 2022 at 12:43 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #14 · p.1 #14 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


JimKasson wrote:
No, I['m saying that the 100S files will have less aliasing.



Thanks, Jim. That makes sense. So, you see the 100x as having a clear advantage in less aliasing. Do you see the expected difference in DR as well?



Jun 13, 2022 at 12:45 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #15 · p.1 #15 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


rdeloe wrote:
A few thoughts on this...

(1) Double-blind or bust. And make sure to control for printing-related confounding variables, e.g.,

- Subject and scene, including lighting
- Printer, ink, paper
- Post processing, including output sharpening
- Print viewing conditions

(2) In an 8x10 print, an X-T2 will have 500ppi on the 8" edge, a GFX 50S will have 774 ppi, and a GFX 100S 1,092 ppi. I'm skeptical that differences that can be seen reliably and consistently would be due to number of pixels sent to the printer.

(3) I don't need more pixels to make better prints. I need better ideas. Your results may vary.
...Show more


You are of course right that the X-T2 will have lots of pixels for an 8 X 10. I don't think it is the number of pixels that will make a difference, but the GFX 50s and 100s will have over 2 stops more DR. I don't know for sure if that would show up in an 8 X 10, but given the right (or wrong depending on your perspective) light I would not be surprised if that showed up even in fairly small prints. I am not sure it would, but I would not be surprised if it did.

I agree that better ideas are the key to better prints, but I am not sure that some ideas (especially with high DR light or perhaps very low light) might not be better able to be implement with a camera with a larger sensor with more DR and better high ISO capability.



Jun 13, 2022 at 12:51 PM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #16 · p.1 #16 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


rdeloe wrote:
(2) In an 8x10 print, an X-T2 will have 500ppi on the 8" edge, a GFX 50S will have 774 ppi, and a GFX 100S 1,092 ppi. I'm skeptical that differences that can be seen reliably and consistently would be due to number of pixels sent to the printer.


I think whether you'll see differences depends on the subject of the test image and the level of craft used in the capture. Sometimes false color persists through a lot of downsampling. BTW, downsampling in and of itself can introduce aliasing. The highroad is to do the error diffusion sampling at the printer marking engine resolution, and not downsample before halftoning at all. But most printer drivers don't work that way.




Jun 13, 2022 at 12:53 PM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #17 · p.1 #17 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Steve Spencer wrote:
Thanks, Jim. That makes sense. So, you see the 100x as having a clear advantage in less aliasing. Do you see the expected difference in DR as well?


At ISOs high enough for the conversion gain to switch to high.




Jun 13, 2022 at 12:55 PM
molson
Offline
• • • • • •
Upload & Sell: On
p.1 #18 · p.1 #18 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


I would also imagine you need a printer with high enough resolution, and a printing medium that will allow faithful reproduction of the fine details as well. Inkjet prints at low dpi on watercolour paper will probably make it hard to tell how much detail you started out with, no matter what camera you used.


Jun 13, 2022 at 01:05 PM
Steve Spencer
Online
• • • • • •
Upload & Sell: On
p.1 #19 · p.1 #19 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


JimKasson wrote:
At ISOs high enough for the conversion gain to switch to high.



Yes, that is where the differences would be expected to emerge, IMO. The differences before the conversion gain are fairly small and I would not expect them to be visible.



Jun 13, 2022 at 01:06 PM
JimKasson
Offline
• • •
Upload & Sell: On
p.1 #20 · p.1 #20 · How big do you need to print to see the advantage of a 100 MP 44 X 33 sensor vs. FF 35mm?


Here's a test I did a while back:

https://blog.kasson.com/gfx-50s/visibility-of-aliasing-gfx-50r-100-aliasing-in-prints/

There are test subjects that are more demanding that the Siemens star; those have concentrations of energy just above the Nyquist frequency.



Jun 13, 2022 at 01:21 PM
1
       2       3       4       5       end






FM Forums | Fuji Forum | Join Upload & Sell

1
       2       3       4       5       end
    
 

You are not logged in. Login or Register

Username      Reset password