How much bigger is 33x44mm MF than FF?

It is really quite reassuring that the discussion can, at times, become very passionate but without the rudeness and malice which sometimes rears it head in other forums.
 
It is really quite reassuring that the discussion can, at times, become very passionate but without the rudeness and malice which sometimes rears it head in other forums.
Thank you. Some of the rudeness in this thread has been removed by the forum moderators.

Jim, as a moderator.
 
Just saw this. That is fantastic! Congrats. You should be and are very pround....
 
A 33x44 mm sensor is 37.5% larger than a FF sensor by image height. A 33x44 mm sensor is 22% larger than a FF sensor by image width. A 33x44 mm sensor is 28% larger than a FF sensor by image diagonal.
Image diagonal determines lens coverage circle sqrt(36^2+24^2) is roughly 43.27mm whereas sqrt(44^2+33^2) is 55mm. That 28% is the most generous way to quote size difference that is vaguely real.

In truth, 44x33mm is really just a rounded-number version of what would really be a multi-aspect-ratio 135-format (FF) lens sensor. 44mm is just the next integer after 43.27mm, so you can have an aspect ratio as wide as fits in the official FF coverage circle. Similarly, the biggest square format that fits with the FF coverage circle is sqrt(36^2+24^2)/sqrt(2) on a side, or roughly 30.59mm; 33mm is the first integer larger than that which gives a small-integer-ratio aspect. In fact, a 33x33mm square would only need 33*sqrt(2), or about 46.67mm coverage circle diameter -- which the vast majority of FF lenses can support.

In fact, there have been several attempts by sensor makers (e.g., Sony) to market 44x33mm as multi-aspect 135 format, and multi-aspect sensors were fairly successful for the MFT format. However, the (until recently) much higher sensor cost for 44x33mm vs. 36x24mm meant the concept didn't make it into commercial camera products.

An interesting note is that Nikon Z mount has a 55mm throat, so it would be particularly easy for Nikon to upgrade to a 44x33mm sensor and still avoid vignetting even with rear-telecentric lenses. Thus, Z could provide a gentle upgrade path first using FF lenses with the larger sensor as multi-aspect FF and then joining in the claims of MF. It could be a coincidence, but I'm an engineer, and I know engineers usually decide such things for a reason... I even quietly wonder if Nikon has made any of their new Z-mount lenses support 55mm coverage circles.
 
Last edited:
I shoot (mainly) square, and I can confirm that most FF lenses work I've tried work just fine on 33x33mm. Ultra wide lenses seem the most challenged and can nip a mm or so off the extreme corners.
 
A 33x44 mm sensor is 37.5% larger than a FF sensor by image height. A 33x44 mm sensor is 22% larger than a FF sensor by image width. A 33x44 mm sensor is 28% larger than a FF sensor by image diagonal.
Image diagonal determines lens coverage circle sqrt(36^2+24^2) is roughly 43.27mm whereas sqrt(44^2+33^2) is 55mm. That 28% is the most generous way to quote size difference that is vaguely real.

In truth, 44x33mm is really just a rounded-number version of what would really be a multi-aspect-ratio 135-format (FF) lens sensor. 44mm is just the next integer after 43.27mm, so you can have an aspect ratio as wide as fits in the official FF coverage circle. Similarly, the biggest square format that fits with the FF coverage circle is sqrt(36^2+24^2)/sqrt(2) on a side, or roughly 30.59mm; 33mm is the first integer larger than that which gives a small-integer-ratio aspect. In fact, a 33x33mm square would only need 33*sqrt(2), or about 46.67mm coverage circle diameter -- which the vast majority of FF lenses can support.

In fact, there have been several attempts by sensor makers (e.g., Sony) to market 44x33mm as multi-aspect 135 format, and multi-aspect sensors were fairly successful for the MFT format. However, the (until recently) much higher sensor cost for 44x33mm vs. 36x24mm meant the concept didn't make it into commercial camera products.

An interesting note is that Nikon Z mount has a 55mm throat, so it would be particularly easy for Nikon to upgrade to a 44x33mm sensor and still avoid vignetting even with rear-telecentric lenses. Thus, Z could provide a gentle upgrade path first using FF lenses with the larger sensor as multi-aspect FF and then joining in the claims of MF. It could be a coincidence, but I'm an engineer, and I know engineers usually decide such things for a reason... I even quietly wonder if Nikon has made any of their new Z-mount lenses support 55mm coverage circles.
16mm FFD - nice and short for technical camera use.
 
A 33x44 mm sensor is 37.5% larger than a FF sensor by image height. A 33x44 mm sensor is 22% larger than a FF sensor by image width. A 33x44 mm sensor is 28% larger than a FF sensor by image diagonal.
You know Jim, all kidding aside.... I've been doing a lot of fooling around and having some fun on this thread, which has been the subject of endless discussion on countless posts and articles all over the photography community since the dawn of GFX as the FF makers freaked out and started panicking (as my Canon very high executive friend told me 4 years ago)....

Those measurements you repeated above are of course well known and are important. (They have been in my notes for at least 5 years from posts you made back then and also copied from Rico.)

But those are each just little parts of the equation of what really matters, which is area and total size of the sensor. As you and everyone else here knows, the GFX / Hassy MF sensor is 1.7 times bigger and 70% larger that the much smaller FF sensors that Canon, Nikon, Sony and Leica are stuck with. Leica might not be stuck with it, but I will believe it when I see it. 😁

I try not to let that much smaller sensor bother me too much when I shoot my beloved Q3 (which I have yet to do on this road trip but will when I hike tomorrow). It takes really good pictures, even with the much smaller sensor.
 
A 33x44 mm sensor is 37.5% larger than a FF sensor by image height. A 33x44 mm sensor is 22% larger than a FF sensor by image width. A 33x44 mm sensor is 28% larger than a FF sensor by image diagonal.
You know Jim, all kidding aside.... I've been doing a lot of fooling around and having some fun on this thread, which has been the subject of endless discussion on countless posts and articles all over the photography community since the dawn of GFX as the FF makers freaked out and started panicking (as my Canon very high executive friend told me 4 years ago)....

Those measurements you repeated above are of course well known and are important. (They have been in my notes for at least 5 years from posts you made back then and also copied from Rico.)

But those are each just little parts of the equation of what really matters, which is area and total size of the sensor. As you and everyone else here knows, the GFX / Hassy MF sensor is 1.7 times bigger and 70% larger that the much smaller FF sensors that Canon, Nikon, Sony and Leica are stuck with.
How do you reconcile this view with your seemingly inconsistent, previously-held position that despite the sensor being less than half the size, images made with Fuji X APS-C cameras were essentially indistinguishable from full frame cameras that used the same sensor tech?
 
Last edited:
A 33x44 mm sensor is 37.5% larger than a FF sensor by image height. A 33x44 mm sensor is 22% larger than a FF sensor by image width. A 33x44 mm sensor is 28% larger than a FF sensor by image diagonal.
You know Jim, all kidding aside.... I've been doing a lot of fooling around and having some fun on this thread, which has been the subject of endless discussion on countless posts and articles all over the photography community since the dawn of GFX as the FF makers freaked out and started panicking (as my Canon very high executive friend told me 4 years ago)....

Those measurements you repeated above are of course well known and are important. (They have been in my notes for at least 5 years from posts you made back then and also copied from Rico.)

But those are each just little parts of the equation of what really matters, which is area and total size of the sensor. As you and everyone else here knows, the GFX / Hassy MF sensor is 1.7 times bigger and 70% larger that the much smaller FF sensors that Canon, Nikon, Sony and Leica are stuck with.
How do you reconcile this view with your seemingly inconsistent, previously-held position that despite the sensor being less than half the size, images made with Fuji X APS-C cameras were essentially indistinguishable from full frame cameras that used the same sensor tech?
I was wrong about that and that was 7 years ago. But remember, back then I was comparing the Fuji APSC files to my Canon 5D III and IV files. I had yet to see GFX! When I did see GFX, I that very week came on here to correct the record. I was stunned 5 years ago when I started shooting GFX.

I love my extensive catalog of Fuji X (APSC) raw files, and they are really nice. But they don't hole up to my 60 MP Q3 files by any stretch.

And yes ... I can see it clearly on my 4K and 6K monitors.

Sensor size matters, which is the only reason we all shoot GFX and new Hassy MF cameras. That is the only reason - much bigger sensor and thus more res which results in much better image fidelity. That's the only reason. And that is a powerful reason not to be taken lightly.
 
But why does sensor size matter?

A bigger sensor collects more light than a smaller sensor and (given equal generation sensors), that means a proportionally increased signal/noise ratio. Which should mean a proportionally reduced noise level. Which should mean a proportionally increased dynamic range.

So, if GFX is 1.7x larger area than FF, that means 2/3 of a stop less noise and 2/3 stop greater dynamic range, all else being equal. So that means that a ISO166 image from a GFX should have about the same image quality as a ISO100 image from a FF.

That's not really a big difference in my book. Is it even visible to most people?

So why do people prefer the GFX? Today, it's my belief that the answer is +40MP extra. But my older GFX has -10MP to my FF. Where does that leave my GFX? Well, there's still that 2/3 of a stop less noise - if that is even visible. And the shrunken pixel aperture which increases acutance and aliasing and adds a little crunch to the image.

This leaves me with a question: is Greg capable of telling the difference between a 60MP FF and a 50MP GFX without any external cues?

One interesting numerical co-incidence is that a square crop from my GFX50s is 38MP. A square crop from my A7Riv is 40MP. A square crop comparison with these would eliminate the aspect ratio difference and almost equalise the pixel count. Which means a comparison would be testing whether FSI 33x33mm with shrunken pixel aperture brings any visible benefit over 24x24mm with BSI tech. It would be an interesting test to do.

--
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:
But why does sensor size matter?

A bigger sensor collects more light than a smaller sensor and (given equal generation sensors), that means a proportionally increased signal/noise ratio. Which should mean a proportionally reduced noise level. Which should mean a proportionally increased dynamic range.

So, if GFX is 1.7x larger area than FF, that means 2/3 of a stop less noise and 2/3 stop greater dynamic range, all else being equal.
Close, but no cigar. If DR is limited by photon noise, it would take a sensor of four times the area to have a stop greater DR.
 
But why does sensor size matter?

A bigger sensor collects more light than a smaller sensor and (given equal generation sensors), that means a proportionally increased signal/noise ratio. Which should mean a proportionally reduced noise level. Which should mean a proportionally increased dynamic range.

So, if GFX is 1.7x larger area than FF, that means 2/3 of a stop less noise and 2/3 stop greater dynamic range, all else being equal.
Close, but no cigar. If DR is limited by photon noise, it would take a sensor of four times the area to have a stop greater DR.
Jim, you are amazing. If cameras were more popular than they are, you could make a fortune as a talking head and expert on the camera network.
 

Keyboard shortcuts

Back
Top