Why are sensors/film rectangular instead of square?

[refrigerator]

From a mathematical perspective, won't a square (also a rectangle) and a rectangle yield the same area in a lens circle? It seems it would be more appropriate to use square frame instead of a rectangular.

 

[GaborSch]

The elongated form can be used in two orientations, yielding different approaches. In average, the rectangular image would be cropped much more.

The larger image area requires more film material (price), and larger sensors (price). It is particularly interesting with sensors: while the price is roughly linear with the amount of film material (it even sinks with larger area, relatively to the unit of surface unit), the price of sensor is disproportionally higher for a larger piece, because of the rejects caused by technical imperfections.

For example the very large sensors of Dalsa, used in MFDBs, are made out of several pieces "stitched" together. This is perhaps the main component of the horrendeous MFDB prices.

--
Gabor

http://www.panopeeper.com/panorama/pano.htm

 

[spt_gb]

Two factors. Firstly people have been offered assorted square formats (see 126 for example) and have largely preferred the rectangular formats. Secondly a rectangular format can use smaller image circle (cheaper) lenses for the same longest dimension.

 

[itt]

For composition reasons most photos end up in rectangular formats, so a square sensor won't be efficient.

If you want to Circumscribe the image circle with a bigger square sensor, then there is waste in the sensor corners.

Besides some DSLRs if not all, don't cast a pure Image Circle due to baffles, lens hoods, mirror, etc

The only logic out of the rectangular shape I see is with Hexagonal shaped sensors because of waffer yield efficiency and optimal Circunscription of IC.
But of course, only for fixed lens Enthusiast Digicams.

--

 

[morninglight]

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

If I am not mistaken, there are square sensors in some "medium format" digital backs. But those are meant to preserve the investment in medium format bodies and lenses, mostly used by professionals that shoot/shot 120 format film in the 6x6 (59mm x 59mm, aka, "2 &1/4").

Similar to that (but more for the benefit of the manufacturer until the advent of so-called FF, "full format") was to preserve the R&D and decades old design of the 35mm SLR body and lenses. The DSLR is in essence a 35mm half frame (like the Pen-F in the 1960's) or 35mm full frame popular since the 1930's, with the addition of a digital sensor in place of the film.

Digital P&S & "bridge" cameras bear resemblance to the 35mm P&S & bridge cameras, except for the LCD and small sensor rather than film. The digitals have longer zooms, also. But the lense formulas are built on the old 35mm ones. Cheaper for the maker, and their is only so much you can do ergonomiclly, on a small camera body anyway.

Some would point out that the 2:1 ratio is akin to the "golden mean" and therefore, aesthetically pleasing. Other reqtangles also have different attributes.

There is a relationship between film/sensor aspect ratio, and how much of an uncropped image you can print on a given piece of paper.

The move to inkjet printing has brought the utility of some rectangle ratios to the fore. This is due to the inkjet photo-printer evolving from the inkjet document printer. For example, 8.5" x 11" (A4) sized paper. "legal size" 11"x14" paper provides the next jump up (manufacturer's already had that carriage size, etc.

Previously 8"x10" became popular with the ratios of film, i.e. 8x10 sheet (contact print), 4x5 (a 2x enlargement) in the early to mid 20th century.
--

'Good composition is only the strongest way of seeing the subject. It cannot be taught because, like all creative effort, it is a matter of personal growth. In common with other artists the photographer wants his finished print to convey to others his own response to his subject. In the fulfillment of this aim, his greatest asset is the directness of the process he employs. But this advantage can only be retained if he simplifies his equipment and technique to the minimum necessary, and keeps his approach free from all formula, art-dogma, rules, and taboos. Only then can he be free to put his photographic sight to use in discovering and revealing the nature of the world he lives in.'

Edward Weston, Camera Craft Magazine, 1930.

 

[nickoly]

The main problem is that people don't use square pictures, except a few of those idiots trying to justify their square format cameras.

If you actually do the maths, you will find that a square in a circle bounding a 4:3 format is marginally larger. The choice becomes clearer when you cut a 4:3 picture out of it. The original 4:3 is about 30% larger

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

 

[morninglight]

Oh, yeah....

The square format (which I would prefer for a digital sensor, but no can afford the digital back

Is from the popularity of the Rolleiflex twin lens reflex camera in the mid 20th century. "120 film", 59mmx59mm. That type of camera essentially required a square negative (though 35mm and 645 adaptors or backs where made by some). It was adopted for many clones of it, then the Hasselblad, then its many clones. BTW: both still exist today, new. To preserve their customer base, Hasselblad and others offer digital back in the square, and "645" (59mmx45mm) formats. These start out at a price similar to a Nikon D3 body (Mamiya) or more than the most spendy Canon DSLR.

--

'Good composition is only the strongest way of seeing the subject. It cannot be taught because, like all creative effort, it is a matter of personal growth. In common with other artists the photographer wants his finished print to convey to others his own response to his subject. In the fulfillment of this aim, his greatest asset is the directness of the process he employs. But this advantage can only be retained if he simplifies his equipment and technique to the minimum necessary, and keeps his approach free from all formula, art-dogma, rules, and taboos. Only then can he be free to put his photographic sight to use in discovering and revealing the nature of the world he lives in.'

Edward Weston, Camera Craft Magazine, 1930.

 

[morninglight]

For composition reasons most photos end up in rectangular formats, so
a square sensor won't be efficient.

Standard chemical process photo papers existed as 4x5, 8x10, 11x14, 16x20 etc. prior to the invention of the Rolleiflex. That is because the dominate sheet films were 4x5 and 8x10. Those aspect ratios work for uncropped enlargements or contacts with those paper sizes.

The Rolleiflex (square) concept was popular among professionals (and still is) because the idea was to crop to fit the paper in the darkroom. Also, "art" photographers who stepped "down" from sheet film, would just often print square. In fact I still have an easel (device to hold a piece of photographic paper flat under an enlarger's lense) that is sized to do a full frame, uncropped square print from a square negative, on 8x10 paper.

While most of my imaging now is 100% digital, I also have two square format film cameras I use, get a good scan done, and do digital flow from that point on.

Note: By the above post, I do not claim anything I do, to be better than anything you do. Simply what I do.).

Actually, I like the aesthetic of the square image. But many photographers (film or digital) do not.

--

'Good composition is only the strongest way of seeing the subject. It cannot be taught because, like all creative effort, it is a matter of personal growth. In common with other artists the photographer wants his finished print to convey to others his own response to his subject. In the fulfillment of this aim, his greatest asset is the directness of the process he employs. But this advantage can only be retained if he simplifies his equipment and technique to the minimum necessary, and keeps his approach free from all formula, art-dogma, rules, and taboos. Only then can he be free to put his photographic sight to use in discovering and revealing the nature of the world he lives in.'

Edward Weston, Camera Craft Magazine, 1930.

 

[Joe Lavee]

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

Square sensor makes a lot of sense for the pro/avid amateur.
It offers an option to crop the image either horizontal, vertical or square.
The only reason that it isn't used in DSLR, must be the cost factor.

 

[morninglight]

The only reason that it isn't used in DSLR, must be the cost factor.

Yeah, and pretty much like I said in my previous posts. Perhaps someday, if sensor costs drop far enough, there will be a niche camera with one, at a low to mid-range (current dslr-wise) price, say 30mm x 30mm, 12mp. The individual receptors woud be nice size on that.

Maybe a digital twin lens reflex, with an 3"x3" hi-res LCD under the hood, instead of a ground glass focusing screen. Sigh.

--

'Good composition is only the strongest way of seeing the subject. It cannot be taught because, like all creative effort, it is a matter of personal growth. In common with other artists the photographer wants his finished print to convey to others his own response to his subject. In the fulfillment of this aim, his greatest asset is the directness of the process he employs. But this advantage can only be retained if he simplifies his equipment and technique to the minimum necessary, and keeps his approach free from all formula, art-dogma, rules, and taboos. Only then can he be free to put his photographic sight to use in discovering and revealing the nature of the world he lives in.'
Edward Weston, Camera Craft Magazine, 1930.

'Next to a dog, a book is a man's best friend. Inside a dog, it's too dark to read.' G. Marx

 

[refrigerator]

Um, what? Are you trying to tell me a rectangle is 30% larger fit into a circle than a square?

That's extremely incorrect.

 

[Papa Carlo]

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

Yes but from a user perspective nobody wants square pictures. Modern TV's are getting more rectangular too.

 

[Steen Bay]

The main problem is that people don't use square pictures, except a
few of those idiots trying to justify their square format cameras.

If you actually do the maths, you will find that a square in a circle
bounding a 4:3 format is marginally larger. The choice becomes
clearer when you cut a 4:3 picture out of it. The original 4:3 is
about 30% larger

Facts are that within the same 43,3 mm (FF) image circle, instead of a 24 x 36 mm sensor we could have a 30,6 x 30,6 mm sensor, with a 8,4% bigger area.

 

[Joseph S Wisniewski]

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

If I am not mistaken, there are square sensors in some "medium
format" digital backs.

It's more like there "were" square sensors. The sensors aren't in production any more, and you can get one of the few remaining square format backs at blowout prices ($10,000 for a bundle of a back, 500 series Blad, and the wide angle that you need because of the crop factor).

But those are meant to preserve the investment
in medium format bodies and lenses, mostly used by professionals that
shoot/shot 120 format film in the 6x6 (59mm x 59mm, aka, "2 &1/4").

Actually, they were meant to save money. Large sensors go up in price exponentially with area. The 36x36mm square backs cost about half what the 36x48 backs did.

With a 1.55x crop factor, that "investment" in medium format lenses didn't work "right", and you typically found yourself buying expensive wides to replace your normal, and often a different portrait lens, too.

Similar to that (but more for the benefit of the manufacturer until
the advent of so-called FF, "full format") was to preserve the R&D
and decades old design of the 35mm SLR body and lenses.

And to get a larger sensor area than the first cropped cameras, improving the low light performance and allowing for higher resolutions.

Some would point out that the 2:1 ratio is akin to the "golden mean"
and therefore, aesthetically pleasing. Other reqtangles also have
different attributes.

A 3:2 ratio is close to the golden mean. There aren't any 2:1 ratio cameras on the market.

There is a relationship between film/sensor aspect ratio, and how
much of an uncropped image you can print on a given piece of paper.

The move to inkjet printing has brought the utility of some rectangle
ratios to the fore. This is due to the inkjet photo-printer evolving
from the inkjet document printer. For example, 8.5" x 11" (A4) sized
paper. "legal size" 11"x14" paper provides the next jump up
(manufacturer's already had that carriage size, etc.

Actually, 11x17 is a very common size. 8.5x11 is very close in aspect to 8x10, so no one really notices. A4 is entirely different, that's a sqrt(2) based aspect ratio system.

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com

 

[refrigerator]

I read somewhere that widescreen TVs and Monitors are actually a poor(er) value, because they have less real estate than a normal 4:3 monitor. In it's extreme, it's like having a thin, long monitor, though the diagonal length of it would be very high.

 

[Joseph S Wisniewski]

Oh, yeah....
The square format (which I would prefer for a digital sensor, but no
can afford the digital back
Is from the popularity of the Rolleiflex twin lens reflex camera in
the mid 20th century. "120 film", 59mmx59mm. That type of camera
essentially required a square negative (though 35mm and 645 adaptors
or backs where made by some). It was adopted for many clones of it,
then the Hasselblad, then its many clones. BTW: both still exist
today, new.

Actually they don't. Rollei actually went bankrupt, and was liquidated. Parts of their portfolio and property was bought by a variety of companies, including Samsung and Franke and Heidecke.

Hasselblad did this weird European "beg the creditors for more time" thing for years instead of going bankrupt, then got bought by an investment company called CINven, and finally by Kohg Kong based heavy machinery company Shriro. There is no longer any square format R&D at Hasselblad, and all 645 design and manufacturer is farmed out to Fuji.

To preserve their customer base, Hasselblad and others
offer digital back in the square,

Actually, the 35mm square Hasselblad square back is a leftover Imacon, it was sold mainly to 645 users (where the crop factor is only 1.37 instead of 1.55 like it is on a square Blad) as a low cost alternative to 36x48mm 645 backs. They didn't sell well, so after Shriro bought Imacon and merged Hasselblad into Imacon, they discontinued the 36mm square back. They're on sale for a good price (and a bundled wide for the price of a normal) right now, but when they're gone, they're gone.

As far as I know, there are no "others" offering a square back. I believe Leaf and Jenoptik have already disposed of their remaining supplies of 36mm square sensors.

and "645" (59mmx45mm) formats.
These start out at a price similar to a Nikon D3 body (Mamiya) or
more than the most spendy Canon DSLR.

The discontinued square Blad back currently at clearance price is about twice the price of a D3.

The discontinued Mamiya 645 back also is at a similar $10,000 price, but that is a back that failed in the marketplace and is ranked substantially below

645 MF backs that are current and respected for their quality start around $20K, about 4x the price of a D3.

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com

 

[nickoly]

Um, what? Are you trying to tell me a rectangle is 30% larger fit
into a circle than a square?

That's extremely incorrect.

Read it again, more slowly this time and with an open mind. What I said was:

The choice becomes clearer when you cut a 4:3 picture out of it. The original 4:3 is about 30% larger.

 

[Joseph S Wisniewski]

The main problem is that people don't use square pictures, except a
few of those idiots trying to justify their square format cameras.

If you actually do the maths, you will find that a square in a circle
bounding a 4:3 format is marginally larger. The choice becomes
clearer when you cut a 4:3 picture out of it. The original 4:3 is
about 30% larger

Facts are that within the same 43,3 mm (FF) image circle, instead of
a 24 x 36 mm sensor we could have a 30,6 x 30,6 mm sensor, with a
8,4% bigger area.

Facts are that a 30.6mm sensor requires 46mm of "back focus" (clear space between the rear element of the lens and the focal plane). You've got about 5mm for the shutter, and 43.3mm to accommodate the swinging SLR mirror. But lenses built for existing 35mm SLRs only have 39mm of back focus, because their frame is 24mm high, not 30.6.

Facts are that a 50mm f1.8 normal with a 39mm back focus is a nearly symmetric double Gauss design that is both high performance and low cost. But a 50mm f1.8 normal with a 46mm back focus has to be built as a retrofocus lens, a wide angle design that is larger that a double Gauss, and will either cost more for the advanced design needed to compensate for aberrations added by the retrofocus design, or perform much poorer than the double Gauss. This is also true of the wide angle lenses.

Facts are that my Nikon 300mm f2.8 and 200mm f4 macro have built in 24x36mm rectangular baffles and can't project a 30.6mm square image, and my 14mm has a built in "petal" lens hood that similarly only allows 24x36mm images. These baffles and hoods are not removable. Facts

Facts are that although the area of the mirror for the 30.6mm square is only 8.4% larger than for a 24x36mm, the angular moment is 2.64 times higher (30.6/24)^4 so there's over twice the mirror slap vibration, longer mirror blackout times, and considerably more noise.

Facts are that I have to crop a lot more of a square than a rectangle for my work, so that 8.4% bigger area is a total waste.

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com

 

[Joe Lavee]

From a mathematical perspective, won't a square (also a rectangle)
and a rectangle yield the same area in a lens circle? It seems it
would be more appropriate to use square frame instead of a
rectangular.

Yes but from a user perspective nobody wants square pictures. Modern
TV's are getting more rectangular too.

Wrong. Some compositions call for a square cropping.

 

[nickoly]

that when you cut a 4:3 picture out of 30.6 x 30.6 you get an even smaller result. Now tell me that's a good idea.

Facts are that within the same 43,3 mm (FF) image circle, instead of
a 24 x 36 mm sensor we could have a 30,6 x 30,6 mm sensor, with a
8,4% bigger area.