We back to the same old question. Some say no diff between GFX and FF yet choose to shoot with GFX. Others say there is a diff and they can see if even at sizes posted here. Personally for me I liked what GFX50s could do, even when comparing with A7rIV. I had both for 2 yrs, shot side by side enough. Sold the Sony when I upgraded to GFX100s as having 2 systems is hard and expensive.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
How much bigger is 33x44mm MF than FF?
- Thread starter JimKasson
- Start date
SrMi
Veteran Member
Larger sensors require less magnification. Apart from technical aspects (e.g., dual conversion gain), magnification difference is the main reason for the increased DR of larger sensors.
yogi4fitness
Veteran Member
Those are marketing guys at Hassy and Fuji and you know there are 4 degrees of lies in this world (in increasing order of magnitude)...
1) lies
2) damn lies
3) statistics
4) marketing
Which is why you must always trust the engineer.
I rest my case.
DavidMillier
Veteran Member
Did a quick test comparing A7Riv with Zeiss Loxia 50mm f/2 @ f/5.6, f/8, f/11 vs GFX50s with 35-70mm @ f/8, f/11, f/16.
As I said, testing is hard to get right. Looking at the images, I think I overexposed the Sony and had to do some highlight recovery to put colour back in the sky in the top left. This put some more light in the shadows and probably cancelled out the GFX noise advantage.
And the depth of field seems different (which it shouldn't be) which makes me think the Fuji didn't focus on the same spot. I manually focused the Loxia (no AF!) and autofocused the Fuji with a single point positioned on the same spot, but maybe I should have switched to manual focus as well to be sure.
Anyway, because of these little errors, I'm not going to post the results until I've re-shot, otherwise I'd just get criticised (rightly) for methodology. But I am willing to share some preliminary impressions...
I shot with a square crop, so that means it's 38MP (Fuji) vs 40MP (Sony). I imported raw files into darktable and applied minimal edits: Crop, lens correction profile, Filmic RGB (to pull back the Sony highlights). I felt I had to lift the shadows on the Fuji to make up for the extra exposure the Sony got, else the Fuji shots looked murky. Looking at the results at 1:1, I see no significant difference. Even adding some sharpening didn't distinguish the files.
As I said, testing is hard to get right. Looking at the images, I think I overexposed the Sony and had to do some highlight recovery to put colour back in the sky in the top left. This put some more light in the shadows and probably cancelled out the GFX noise advantage.
And the depth of field seems different (which it shouldn't be) which makes me think the Fuji didn't focus on the same spot. I manually focused the Loxia (no AF!) and autofocused the Fuji with a single point positioned on the same spot, but maybe I should have switched to manual focus as well to be sure.
Anyway, because of these little errors, I'm not going to post the results until I've re-shot, otherwise I'd just get criticised (rightly) for methodology. But I am willing to share some preliminary impressions...
I shot with a square crop, so that means it's 38MP (Fuji) vs 40MP (Sony). I imported raw files into darktable and applied minimal edits: Crop, lens correction profile, Filmic RGB (to pull back the Sony highlights). I felt I had to lift the shadows on the Fuji to make up for the extra exposure the Sony got, else the Fuji shots looked murky. Looking at the results at 1:1, I see no significant difference. Even adding some sharpening didn't distinguish the files.
Greg7579
Veteran Member
The Fuji Hassy MF sensor is 70 percent bigger than FF sensors. That is the main reason we buy those cameras other than the glass in Fuji's case.
There are people and many companies that don't like that fact and would prefer it not to be harped on.
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
If you get a test result that shows a significant difference, I would suspect a problem with the test.
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
70% bigger by area, 28% bigger by diagonal.
You finally figured it out - this day had to come, I suppose.
Jim is on the payroll of the FullFrame Agency.
His mission - to somewhat understate the advantages of 33x44mm cameras.
His cover, decades of experience, including a focus on color management, color processing for digital photography and gamut mapping, and seven patents related to image processing including “Computationally efficient low-artifact system for spatially filtering digital color images,” and “Method and Apparatus for Tone Correction of a Digital Color Image with Preservation of the Chromaticity of the Image,”.
Greg7579
Veteran Member
An amazing but not surprising statement on our Medium Format Board.
David, if your test shows that FF = MF image fidelity, that will make a lot of people, camera companies and marketing entities very happy if they have any confidence in your test.
Of course, any direct comparison between cameras with different sensor sizes and using different lenses is difficult with all of the equivalency, scene, setting and focusing requirements.
If you shoot a chart or flat target, people will say they never shoot charts or flat targets.
If you don't use a test chart, others will say you should have.
It's the old DPR challenge, no matter what Forum you are on, and no matter what you find, the "test" can be shot full of holes.
I've attempted it many times. Good luck.
BY the way, I'm looking at a Q3 60MP shot right now on my 4K 15-inch laptop screen and the 100II shots of the same scene and fairly close to the same FOV. Not completely equivalent by any means. But still....
Is anyone interested in what I see and think?
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
Not surprising because you agree, or because you have figured out the agenda?
If it’s the latter, perhaps you can explain your thesis as to the source of a significant difference?
Yes indeed. Canon is incapable of conducting such a test, and would be delighted.
Darn the scientific method and the pesky “disprove-ability” thing.
Last edited:
Greg7579
Veteran Member
True fa
This is a good lesson. Expressing the description as a diagonal vs an area is what they do with TVs and computer screens. People can relate to it.
I went from a 70-inch to an 80-inch TV recently. TVs are marketed and identified with those diagonal measurement. That doesn't sound like much diagonally, but it is a big difference in area.
The viewing experience difference between the two is really amazing. The 80 inch seems huge compared to the 70 inch when watching any program or sport. Teresa was furious with me because the TV looked so much bigger. That is a lot of extra area which has a big impact on the everything about the viewing experience.
Same with 27 and 32 inch monitors.
Screens aren't sensors, but the analogy is good when talking about area vs diagonal measurement of a rectangle when describing the difference in size.
This is not about Jim. I'm a big fan of Jim. (But that is funny - good job.) FF Agency! I like it.
What we need is a MF agency. LOL.
No, that is not Jim's mission. However, that effort does appear to be in the interest of someone I know in the camera equipment social media arena and who is probably a pretty good guy. I would love to meet him.
Greg7579
Veteran Member
Good points Doppler and I really like David, so I wish him well on his test. I doubt if he will claim any statistical significance or that it was scientific, but it is good for us all to shoot a lot of shots and observe our raw files on a good monitor so we can make our own purchasing decisions.
One would be a fool to buy expensive MF equipment if they think FF image fidelity is equal or even very close to MF.
I have said this 200 times in the past 5 years.
DavidMillier
Veteran Member
You are correct, it is very difficult to do testing perfectly, especially if you are not practiced at it and don't have a well polished method. That's me all right. Which I happily accept means I'm not in a position to make a definitive statement that other people can trust and rely on.
I do think, however, that having said all the above, if you set about repeating tests as best you are able, you are eventually going to average out the worst of the problems and end up with a reasonable idea of the typical differences, at least as far as your own workflow is concerned. Maybe not good enough to publish as a formal review, but good enough to inform your own opinion.
Now, I am only really interested in big differences. The kind of differences that would have any viewer say "Wow, that one is so much better". I don't really care at all about the sort of minute differences that means you have to squint at the pictures for half an hour, scratching your head and umming and aaaing until you eventually, doubtfully, say "I think that one might be a tiny bit better, but I'm not really sure". What I'm looking for is differences that would routinely show up in normal fieldwork, not just under careful test conditions.
In that spirit, I can overlook small methodological problems that don't effectively change the result for my own purposes but would give critics cause to complain. But in the interest of fair play, I'm happy to keep repeating the test in an attempt to remove these caveats.
I'd be very surprised if tightening the test parameters made me change my conclusions substantially, the images are just too close to call.
--
2024: Awarded Royal Photographic Society LRPS Distinction
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)
I do think, however, that having said all the above, if you set about repeating tests as best you are able, you are eventually going to average out the worst of the problems and end up with a reasonable idea of the typical differences, at least as far as your own workflow is concerned. Maybe not good enough to publish as a formal review, but good enough to inform your own opinion.
Now, I am only really interested in big differences. The kind of differences that would have any viewer say "Wow, that one is so much better". I don't really care at all about the sort of minute differences that means you have to squint at the pictures for half an hour, scratching your head and umming and aaaing until you eventually, doubtfully, say "I think that one might be a tiny bit better, but I'm not really sure". What I'm looking for is differences that would routinely show up in normal fieldwork, not just under careful test conditions.
In that spirit, I can overlook small methodological problems that don't effectively change the result for my own purposes but would give critics cause to complain. But in the interest of fair play, I'm happy to keep repeating the test in an attempt to remove these caveats.
I'd be very surprised if tightening the test parameters made me change my conclusions substantially, the images are just too close to call.
--
2024: Awarded Royal Photographic Society LRPS Distinction
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)
Last edited:
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
You’re being far too modest!
DavidMillier
Veteran Member
Perhaps I'm on the way to proving myself a fool, because what I'm seeing so far is that the image quality is essentially the same, at least as far as detail and sharpness are concerned.
Noise levels are still up in the air because I think I gave the Sony about 1 stop more exposure which would negate any medium format advantage.
Let's not jump the gun, I'll repeat the test until I feel I've eradicated mistakes then publish the pics here.
--
2024: Awarded Royal Photographic Society LRPS Distinction
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)
Last edited:
Greg7579
Veteran Member
Go ahead and give it a go David. I won't say anything no matter what the result and I won't give you any flak.
This is more for you than for us anyway. It will be fun, and you will learn a lot playing around with the equivalencies. Pick a good deep scene and just experiment. DOF and OOF pixels (wrong term I know) on the MF vs FF shots are a bit of a challenge, but you'll get a good impression if you get it somewhat close on the equivalencies.
If you want to get more technical, use a slant target and/or star chart and test like Jim does with lenses and some other specific cases. But that can quickly become more about the lens differences between the two cameras and equivalency problems.
But you should see something that effects your opinion of FF vs MF image fidelity.
I do it a lot and I already know what I think.
Greg7579
Veteran Member
I do sometimes think I am a fool for doing what I do, but it is fun for me, and I enjoy shooting landscapes in the mountains on a road trip, even though it is in the mountains where I always drop lenses out of car doors and hatches.
Teresa tells all my friends I am an idiot for buying all of this camera gear. People don't really relate to cameras anymore in the general population.
But she is saying that to guys who drive porches to the yacht club or golf course, and what I spend on camera gear is less than what their wives spend on paintings or jewelry.
They say, Teresa, let Greg have his fun. He won't live forever.
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
A question of logic:
Jim started a thread proposing a metric that he believes more accurately describes the advantages of going to a larger sensor.
I agree that the metric, a linear rather than an area measure, better describes deltas.
You disagree with me.
How can it be that you agree with Jim, if Jim and I agree?
- Messages
- 4,188
- Solutions
- 1
- Reaction score
- 4,056
Life is short, and we should do what we enjoy if we can.
Similar threads
Keyboard shortcuts
- f
- Forum