How much bigger is 33x44mm MF than FF?

[DavidMillier]

If you were selling floor space I would bet a room 43ft by 33ft would cost approx 70% more than one 36ft by 24 ft..

You are probably correct.

A 43mm x 33mm sensor likely costs the same as or more than a 36mm x 24mm sensor per unit area.

This isn’t the question Jim has addressed.

We might agree that the benefit of floor space is generally correlated to its area, and that the relationship is linear with respect to area.

The question Jim is addressing is: what is the relationship between the area of a sensor and its benefit?

I believe he has argued persuasively, that the relationship isn’t linear with respect to sensor area.

Jim was responding to me when I said the MF GFX and new Hassy sensor is 70% bigger than FF, which it is. Whether or not it is better is a different argument. I was talking size.

The term ‘size’ is ambiguous.

No it's not. It is 70% bigger. Why does that simple fact seem to offend you?

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

What's calculus? They don't teach it at school here any more, unless you study a specialised exam syllabus. I've always wondered...

 

[JimKasson]

If you were selling floor space I would bet a room 43ft by 33ft would cost approx 70% more than one 36ft by 24 ft..

You are probably correct.

A 43mm x 33mm sensor likely costs the same as or more than a 36mm x 24mm sensor per unit area.

This isn’t the question Jim has addressed.

We might agree that the benefit of floor space is generally correlated to its area, and that the relationship is linear with respect to area.

The question Jim is addressing is: what is the relationship between the area of a sensor and its benefit?

I believe he has argued persuasively, that the relationship isn’t linear with respect to sensor area.

Jim was responding to me when I said the MF GFX and new Hassy sensor is 70% bigger than FF, which it is. Whether or not it is better is a different argument. I was talking size.

The term ‘size’ is ambiguous.

No it's not. It is 70% bigger. Why does that simple fact seem to offend you?

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

What's calculus?

Loading…

en.wikipedia.org

They don't teach it at school here any more, unless you study a specialised exam syllabus. I've always wondered...

Gosh, in the US, they teach it in high school.

 

[Greg7579]

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

Area is not a wrong metric when you are talking about size. There is no argument when you say the GFX sensor is 70% bigger. But there are many other was to express the percentages and other aspects of what you were talking about. No problem.

But the GFX sensor is 70% bigger.

Jim, I forgot all the calculus I ever leaned. You know I went to West Point and got an Engineering degree 45 years ago when West Point was very heavy science and math and there was only one degree - Engineering. (It's not like that now.)

I don't remember any of it. 8 semesters of calculus & Diff EQ, 6 semesters of various levels of physics and 4 semesters EE. Three semesters fluid dynamics and 5 semesters of computer science (mainframe - yuch).

I don't remember any of it. Then ten years later got an Industrial Engineering Masters with absolute tons of prob and stats and queuing theory...

Don't remember any of it.

So yes, I had the education back in the day. All wasted because I spent my career in the US Army infantry and the only time I ever used it was when I was a mortar platoon leader in 1981. We had to try to drop mortar rounds into a very small area far away....

And I remember sniper school where trajectory mattered. Ranger school? No math requirement except to try to figure out how many days more you had to suffer. LOL.

Airborne school? We had 1500 feet before we hit the ground - chute or no chute.

Air Assault School? I hated helicopters and we had to plot navigational vectors before GPS, but I was always so seasick in those things that I was usually throwing up out the door and it would blow back in on my platoon mates. Infantry guys don't mind suffering, but they hate getting puked on.

 

[DavidMillier]

If you were selling floor space I would bet a room 43ft by 33ft would cost approx 70% more than one 36ft by 24 ft..

You are probably correct.

A 43mm x 33mm sensor likely costs the same as or more than a 36mm x 24mm sensor per unit area.

This isn’t the question Jim has addressed.

We might agree that the benefit of floor space is generally correlated to its area, and that the relationship is linear with respect to area.

The question Jim is addressing is: what is the relationship between the area of a sensor and its benefit?

I believe he has argued persuasively, that the relationship isn’t linear with respect to sensor area.

Jim was responding to me when I said the MF GFX and new Hassy sensor is 70% bigger than FF, which it is. Whether or not it is better is a different argument. I was talking size.

The term ‘size’ is ambiguous.

No it's not. It is 70% bigger. Why does that simple fact seem to offend you?

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

What's calculus?

https://en.wikipedia.org/wiki/Calculus

They don't teach it at school here any more, unless you study a specialised exam syllabus. I've always wondered...

Gosh, in the US, they teach it in high school.

I don't know how well you know the UK schools system.

Until recently the end of compulsory secondary schooling was at age 16. That is why the end of school national examinations program happens at that age.

It used to called O-level exams until 1987, then changed name to GCSE exams. These are single subject qualifications. For example, at 16, I sat O-level examinations in Eng language, English Lit, French, maths, physics, chemistry, British constitution, geography, photography. These are all independent qualifications and you get a grade and a certificate for each one. We have never had a baccalaureate style education system like much of Europe and the US.

At 14, you got to choose which subjects you wanted to study at O-level (English and maths were compulsory, most of the others were optional). This could cause clashes. For example, I never studied O-level history because the options didn't allow me to study geography, british constitution and history all at once, I had to choose 2 from 3.

I know from looking at vintage maths O-level past papers that a basic introduction to integration and differentiation were part of the O-level maths syllabus in the 1960s, but when I took my O-level maths in 1978, it was no longer part of the syllabus and I was never taught it. They moved the calculus topic up to the A-level course studied as tertiary education at age 16-18 by those who decided not to leave school at 16 (about 20% of students). And at 6th form, students typically picked 3 A-Level subjects. I chose English, history and french. Very few people opted for A-level maths back, then, too hard, so few people of my age ever got taught calculus unless they went on to study higher maths courses.

The school leaving age is now 18, so everyone has to go to 6th form and take A-levels (or an equivalent like BTEC diplomas). And because of the increased emphasis on STEM over the last decade, students opting for maths A-level has massively increased. I think it might now be one of the most popular choices, rather than least popular. So, basic intro to calculus used to be taught to 16 years, then got dropped and was studied only by people specialising in maths 16-18 and now it had dramatically made a come back because of the great rise in the popularity of STEM (and that fact that you cannot no longer leave at 16).

 

[Doppler9000]

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

Area is not a wrong metric when you are talking about size. There is no argument when you say the GFX sensor is 70% bigger. But there are many other was to express the percentages and other aspects of what you were talking about. No problem.

But the GFX sensor is 70% bigger.

Size is an ambiguous term. The 33x44mm sensor is 70% bigger by area, 28% by frame diagonal, versus a 24x36mm sensor.

If someone said Person A is twice the size of Person B, what is the answer?

Twice as tall?

Twice as heavy?

Because there are different ways to characterize “size”, Jim is explaining the metric most relevant to image output.

Why does this elude you?

 

[JimKasson]

I don't know how well you know the UK schools system.

Until recently the end of compulsory secondary schooling was at age 16. That is why the end of school national examinations program happens at that age.

It used to called O-level exams until 1987, then changed name to GCSE exams. These are single subject qualifications. For example, at 16, I sat O-level examinations in Eng language, English Lit, French, maths, physics, chemistry, British constitution, geography, photography. These are all independent qualifications and you get a grade and a certificate for each one. We have never had a baccalaureate style education system like much of Europe and the US.

At 14, you got to choose which subjects you wanted to study at O-level (English and maths were compulsory, most of the others were optional). This could cause clashes. For example, I never studied O-level history because the options didn't allow me to study geography, british constitution and history all at once, I had to choose 2 from 3.

I know from looking at vintage maths O-level past papers that a basic introduction to integration and differentiation were part of the O-level maths syllabus in the 1960s, but when I took my O-level maths in 1978, it was no longer part of the syllabus and I was never taught it. They moved the calculus topic up to the A-level course studied as tertiary education at age 16-18 by those who decided not to leave school at 16 (about 20% of students). And at 6th form, students typically picked 3 A-Level subjects. I chose English, history and french. Very few people opted for A-level maths back, then, too hard, so few people of my age ever got taught calculus unless they went on to study higher maths courses.

The school leaving age is now 18, so everyone has to go to 6th form and take A-levels (or an equivalent like BTEC diplomas). And because of the increased emphasis on STEM over the last decade, students opting for maths A-level has massively increased. I think it might now be one of the most popular choices, rather than least popular. So, basic intro to calculus used to be taught to 16 years, then got dropped and was studied only by people specialising in maths 16-18 and now it had dramatically made a come back because of the great rise in the popularity of STEM (and that fact that you cannot no longer leave at 16).

Thank you.

 

[JimKasson]

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

Area is not a wrong metric when you are talking about size. There is no argument when you say the GFX sensor is 70% bigger. But there are many other was to express the percentages and other aspects of what you were talking about. No problem.

But the GFX sensor is 70% bigger.

If you choose area as your metric. If you choose area as your metric, a 100 mm f/2.8 lens is four times the size of a 50mm f/2.8 lens.

 

[Greg7579]

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

Area is not a wrong metric when you are talking about size. There is no argument when you say the GFX sensor is 70% bigger. But there are many other was to express the percentages and other aspects of what you were talking about. No problem.

But the GFX sensor is 70% bigger.

Size is an ambiguous term. The 33x44mm sensor is 70% bigger by area, 28% by frame diagonal, versus a 24x36mm sensor.

If someone said Person A is twice the size of Person B, what is the answer?

Twice as tall?

Twice as heavy?

Because there are different ways to characterize “size”, Jim is explaining the metric most relevant to image output.

Why does this elude you?

Let me make this simple for you Doppler. Nothing you say eludes me. I understand completely what you are alluding to.

 

[Doppler9000]

If you choose the wrong metric, it's 70% bigger. I can probably come up with a metric that makes the percentage difference much greater. In fact, if you give me a target difference -- call it delta -- I can probably come up with a metric -- call it epsilon -- that makes the difference delta.

You may recognize the formulation from calculus class, or real variable analysis, if you got that far.

Area is not a wrong metric when you are talking about size. There is no argument when you say the GFX sensor is 70% bigger. But there are many other was to express the percentages and other aspects of what you were talking about. No problem.

But the GFX sensor is 70% bigger.

Size is an ambiguous term. The 33x44mm sensor is 70% bigger by area, 28% by frame diagonal, versus a 24x36mm sensor.

If someone said Person A is twice the size of Person B, what is the answer?

Twice as tall?

Twice as heavy?

Because there are different ways to characterize “size”, Jim is explaining the metric most relevant to image output.

Why does this elude you?

Let me make this simple for you Doppler. Nothing you say eludes me. I understand completely what you are alluding to.

I’m not ‘alluding’ to anything. Just making the plain statement of fact. Not clear why some get emotional about relevant size metrics.

Now that Jim has explained the most relevant metric for sensor size, hopefully we can put the question behind us and stop rearguing the point on every thread.

 

[DavidMillier]

As we have diverted into this little by water into eduction, I'll take the liberty of expressing some fatherly pride. This year's A-level examination results day was on Thursday. My daughter sat A-levels in Eng lit, British and American history, classical civilisations and drama & theatre studies and her grades were AAAB, plus an A* in her EPQ research project and she has been accepted at Worcester College, Oxford University to read for a BA(Hons) in Classics. She's crazy about ancient greek and roman culture and judging by the conversations I had with the professors, so are they. We are very chuffed and she is over the moon. I don't personally know anyone who got into Oxford, so this is a big thing for us.

 

[Doppler9000]

As we have diverted into this little by water into eduction, I'll take the liberty of expressing some fatherly pride. This year's A-level examination results day was on Thursday. My daughter sat A-levels in Eng lit, British and American history, classical civilisations and drama & theatre studies and her grades were AAAB, plus an A* in her EPQ research project and she has been accepted at Worcester College, Oxford University to read for a BA(Hons) in Classics. She's crazy about ancient greek and roman culture and judging by the conversations I had with the professors, so are they. We are very chuffed and she is over the moon. I don't personally know anyone who got into Oxford, so this is a big thing for us.

Congratulations - that is fantastic!

 

[Stan Disbrow]

Hi,

And then that all depended on the film used. I was using Panchromatic ISO 400 and processing it myself. And I could see a difference at 8x10.

When I shot slower color film and sent it off for processing and printing, those 8x10s were stunning. A word I can't say as I could apply to the same film in 135.

These days, with digital, I am printing my own 16x20s. It's amazing how much easier processing is. But I can't exactly compare reasonably equivalent sensors the way I did film back when. Not much reasonably equivalent between a 16 MP CMOS and a 40 MP CCD....

But that's where I am these days. But maybe not for long. At least on the 40 MP MF side.

Stan

 

[DavidMillier]

As we have diverted into this little by water into eduction, I'll take the liberty of expressing some fatherly pride. This year's A-level examination results day was on Thursday. My daughter sat A-levels in Eng lit, British and American history, classical civilisations and drama & theatre studies and her grades were AAAB, plus an A* in her EPQ research project and she has been accepted at Worcester College, Oxford University to read for a BA(Hons) in Classics. She's crazy about ancient greek and roman culture and judging by the conversations I had with the professors, so are they. We are very chuffed and she is over the moon. I don't personally know anyone who got into Oxford, so this is a big thing for us.

Congratulations - that is fantastic!

Thank you

 

[JimKasson]

As we have diverted into this little by water into eduction, I'll take the liberty of expressing some fatherly pride. This year's A-level examination results day was on Thursday. My daughter sat A-levels in Eng lit, British and American history, classical civilisations and drama & theatre studies and her grades were AAAB, plus an A* in her EPQ research project and she has been accepted at Worcester College, Oxford University to read for a BA(Hons) in Classics. She's crazy about ancient greek and roman culture and judging by the conversations I had with the professors, so are they. We are very chuffed and she is over the moon. I don't personally know anyone who got into Oxford, so this is a big thing for us.

Congratulations to you both!

 

[Stan Disbrow]

Hi,

The diagonal measurement is the only sensible one to use here. Remember, you are sticking a square peg into a round hole. Optical output is circular and the diagonal of a rectangle is the measurement which aligns with the diameter of a circle.

Unless, of course, marketing gets involved. They love to spin things faster than a motorized top.

So, yeah. Fujifilm marketing: Large Format and 70 percent larger than Full Frame (135 format). Yep. That's it right there!

Stan

 

[Stan Disbrow]

Hi,

Don't forget marketing. They have been known to spin things pretty far from the facts. And, that regardless of what engineering tells them.

I can recall several instances where the customers figured that out and then Hell rained down upon Corporate Heads from Government agencies.

Not in this case. So far, anyway.

Stan

 

[DavidMillier]

As we have diverted into this little by water into eduction, I'll take the liberty of expressing some fatherly pride. This year's A-level examination results day was on Thursday. My daughter sat A-levels in Eng lit, British and American history, classical civilisations and drama & theatre studies and her grades were AAAB, plus an A* in her EPQ research project and she has been accepted at Worcester College, Oxford University to read for a BA(Hons) in Classics. She's crazy about ancient greek and roman culture and judging by the conversations I had with the professors, so are they. We are very chuffed and she is over the moon. I don't personally know anyone who got into Oxford, so this is a big thing for us.

Congratulations to you both!

Thank you, although it's all down to her, really. Interestingly, although I tried to nudge her in the direction of STEM (which she was equally good at) she ended up studying mainly the same subjects as I did back in the day, but she is so embarrassingly much better than I ever was, it's like another world. I still have a bunch of my A-level English Lit essays and it's like they were written by a young child compared to hers. The education system and the expected standards are so much better/higher achieving than my day. My middling A-level grades actually put me in the top 25% of achievers for my cohort, but students today would think those grades dire, too embarrassing to admit to. Then again, in my father's generation, most kids left school for the factory by age 14. I think only 1% of the population went to university back then; these days it's over 40%.

 

[Gandolphi]

If you were selling floor space I would bet a room 43ft by 33ft would cost approx 70% more than one 36ft by 24 ft..

Possibly true, but not relevant here. We're not talking about the cost to manufacture a sensor.

A Hasselblad sensor is bigger than a full frame sensor,

That is not in question.

does it really matter what facts we bend to define this?

Are you saying that metrics don't matter? I've seen projects fail completely because the wrong metric was optimized.

It doesn’t matter. We all agree the larger sensor provides the means to produce a better photograph and this is not changed by deciding or changing which size measurement is used.

Back in the 60’s the calculation for determining expansion of fluid carrying pipes was changed to one which was slightly more accurate. This was the era of log tables and slide rules and the new calculation took about five times as long as the original. Still, one can’t ignore a fraction of an inch every ten feet or so, therefore the new calculation was used.



it was then pointed out, the fitters were still using Stanley tapes to measure, yellow crayon to mark and cutting with a blowtorch or hacksaw to the nearest quarter inch…

 

[JimKasson]

If you were selling floor space I would bet a room 43ft by 33ft would cost approx 70% more than one 36ft by 24 ft..

Possibly true, but not relevant here. We're not talking about the cost to manufacture a sensor.

A Hasselblad sensor is bigger than a full frame sensor,

That is not in question.

does it really matter what facts we bend to define this?

Are you saying that metrics don't matter? I've seen projects fail completely because the wrong metric was optimized.

It doesn’t matter. We all agree the larger sensor provides the means to produce a better photograph and this is not changed by deciding or changing which size measurement is used.

But the difference in size is important. Would a 37x25mm sensor provide a meaningful advantage over a FF one? I think not. Using the right size metrics, many of the improvements can be quantified.

Back in the 60’s the calculation for determining expansion of fluid carrying pipes was changed to one which was slightly more accurate. This was the era of log tables and slide rules and the new calculation took about five times as long as the original. Still, one can’t ignore a fraction of an inch every ten feet or so, therefore the new calculation was used.

it was then pointed out, the fitters were still using Stanley tapes to measure, yellow crayon to mark and cutting with a blowtorch or hacksaw to the nearest quarter inch…

The difference between area and linear metrics in this case is much larger than that.

 

[BSiler]

No way Sir! All I remember is that sign over cosign equals tangent and that the square root of 4 is 2.

Hmm, I remember something about sine and cosine.