This goes way, way back. One simple answer is that humans prefer squares and rectangles.

Our Vision and Brain Like Rectangles!

Physiologically, we humans are just plain prone to favor rectangular images. Here's a chart of a visual field mapped by NASA way back in the 1960s. The central white area being the area of overlap between the two eyes.

Thanks to our stereoscopic vision, while our vision isn't precisely rectangular, it's somewhat rectangular: a rectangle could be nicely inscribed that most sensitive visual center and waste less of it than a circle. Because of the overlap, we have about a 180 degree horizontal field of view, a 130 degree vertical field of view. Our sharp field of view is narrower but similarly shaped.

Then add to the fact that we humans largely prefer horizontal and vertical lines... that's a preference wired into our brains. Look at our architecture: rectangular windows, strong use of horizontal and vertical lines. When a science fiction show wants to depict some alien structure, it's nearly always moving away from this. A good alien is probably going to get something very organic and flowing, a bad alien is probably going to something spiky, random, or insectoid-looking. When we build buildings that radically depart from these lines, they look other-worldly, alien or futuristic, to most people.

The Art World

Humans have been painting for over 10,000 years. Prehistoric cultures made paintings on cave walls, on rocks, etc. And even though they weren't making canvas yet, these paintings were typically rectangular in nature.

Artists dating back to ancient civilizations knew about the Golden ratio, and often modeled buildings, sculptures, and paintings based on this mathematical relationship. Which leads to rectangular a rectangular painting. This wasn't simply the shape of the image as a whole, but often reflected within the image as well, as DaVinci was kind of enough to illustrate here:

So over the centuries, people grew accustomed to the idea that paintings were usually rectangular.

And then the practical: it was simply easier to stretch canvas over a rectangular frame than it was to make a canvas in some other shape. True, there were occasional round or oval paintings, but even those, most of the time, were rectangular canvas with a circular or oval matte.

Enter the Camera Obscura

There's ample evidence today that early cameras, specifically the room-sized pinhole camera dubbed the Camera Obscura, were first used by artists... long before we had devised any kind of film (do yourself a favor and see "Tim's Vermeer" when you can).

And of course, being otherwise the usual artists, they tended to paint rectangular paintings, even given the circular projection of the camera.

Eventually, we got real cameras (see also: Dave Haynie's answer to What was the first camera called?) with real film. These were initially just wooden boxes with a pinhole or lens at one end and an opening in the back. Being that it's much easier and more functional to create a wooden box than, say, a wooden sphere, the camera was a cube or prism.

Some early cameras did project a circle onto the film. But it was simply easier and more efficient to make film plates rectangular, so pretty quickly, the camera makes simply chose a lens that would cover the whole rectangular area of the photographic plate.

These wet plates gave way to dry plates, and the dry plates gave way to other tricks. A guy called George Eastman first put dry emulsion on a paper roll and sold a camera called Kodak. This camera could shoot 100 shots, then you would return it to the Eastman Company and they'd send you a new camera and a box of prints. And of course, since that roll of paper, containing negatives, was contact printed to make the prints, the resulting prints were exactly the same size as the roll. So it only made sense to use the entire area of each frame — meaning a rectangle — to maximize the size of the photo. The paper rolls were eventually replaced by film rolls, and we had film that stated pretty much the same for 100 years.

The Realities of Actual Lenses

Even if you actually wanted a circular image, you would probably not want an image that capture the very edges of the image projected by a lens. Image quality, image focus on a flat plane, illumination, linear distortions, etc. are all worse near the edges of the lens image than in the center. So, just like we cropped the visual field above to deliver the better part of our vision above, we do the same thing for our lenses.

For real world lenses, you can find an MTF (Modulation Transfer Function) that shows how the quality drops off as you move away from the center of a lens. Read more here: MTF, Resolution & Contrast .

Electronic Cameras and Videography

In 1927, Philo T. Farnsworth demonstrated an "image dissector" tube that could scan and image, reconstructed in another room on a display tube. These tubes had sensitive areas that were circular, though the electron scanning they used was rectangular -- just easier to do it that way. In fact, the rather odd measurement system we use today for most camera sensors is derived directly from the days of circular imaging tubes.

Video tubes we fine for the early days of television, but useful for mobile imagining. So it wasn't until the invention of the Charge-Coupled Device (CCD) that electronic imaging became a consumer notion. CCDs were used first in analog electronic video cameras and a few video-camera-based analog still cameras like the Sony Mavica. But they moved into the digital world. Either way, chips are chips.

This is an image of a very early DSLR, the Kodak DCS520 from 1998. And next to it, you see a wafer of the 2Mpixel sensors from that chip. Like most integrated circuits, a CCD is made via lithography and various chemical steps in a clean room, starting with a circular wafer of silicon. These wafers are expensive, and the larger the chip, the most they cost. So you really don't want to waste any space. And so, the most efficient shape that ensures the maximum area per sensor and the most sensor per wafer is square or rectangular.

As well, the idea of a digital camera was to deliver a replacement for a film camera. In the case of the DCS-520, Kodak was using a Canon EOS camera body and adding their electronics, so this dropped right in to Canon's film SLR system of the day. Which, of course, already used a rectangular film plane, because the efficiency of rectangular images on film, as mentioned, has much the same value as that on silicon.

And while we've most to CMOS, the idea of the camera, the chip production process is similar, the idea that we're most just changing the sensor but leaving the rest of photography in place is still there, etc. So we're still rectangular.

First, it’s a known fact while lenses capture an image throughout the circumference of the glass, any degradation occurs on the outer parts of the image, hence cutting off those parts results in a sharper overall image in a rectangular format. This leads to the next question, “Then why aren’t images square?”

Another good question as it makes sense to trim off the degradation equally from all sides, which results in a square. However, any designer will tell you that a rectangle is more pleasing to the eye, since we tend to naturally see more through our own eyes in a horizontal format. This rectangle format also allows us to shoot the more powerful vertical.

Though there have been square format cameras in the past, like the Kodak 126 Instamatic camera, basically a squared 35mm format, and the more commonly known amongst professional photographers was the medium format, 2¼ square format.

I myself used the square format back in high school as a photographer. The camera of choice for my photojournalism instructor for his photographers was the Yashica Mat 120G medium format camera. The square image the camera produced was 2 ¼-inches square on each side, many times larger than a 35mm negative, allowing us ample room to crop out rectangle images in either a horizontal or vertical format.

So the more I thought about that question, I decided to come up with some answers of my own (in a humorous way) and a few other questions I’d love to know about the answers myself. So first, here are a few of my assumptions on why we capture in a rectangular format vs. a round format, though we use round lenses:

• Rectangle lenses would provide lens barrels that would be difficult to focus. Imagine turning a rectangle, focusing ring, not to mention attaching rectangle filters.

• Finding picture frames would become interesting, now we’d have different circumference size frames.

• There would be tons of paper waste because paper is in a rectangular format, so round prints would create a non-green environment.

• Round computer monitors to view photos doesn’t make sense.

• It becomes difficult to properly align round images in frames and hang them exactly the way they were captured.

• If video adopted round formats, tubeless televisions today but appear as one large tube.

• Our smart phones would become round to view the images properly too. Now can you imagine round Instagram photos instead of their current square format? Even worse, if you dropped your phone can you picture it rolling down hill?

• Billboards would become round adding a tunnel vision perception.

• If theater screens were round would we be sitting in tunnels viewing movies instead of rectangular theater rooms? Can you imagine the acoustics alone?

• The Rule of Thirds would require a revision.

Now here are some of my questions about other things in life:

• Why is bacon in the U.S. cut in rectangular strips and Canadian bacon in round pieces?

• Why of all the cold cuts bologna is round knowing it will go on a square piece of bread?

• Why does Wendy’s make square hamburger patties for their round buns?

• If you made square donuts would you have square donut holes?

I’m sure there are many other interesting, and humorous questions out there that can fascinate us with answers, but ultimately there are many logical reasons why lenses are round and capture devices as well as prints are rectangular. Since I’m not a photo scientist I can only refer to the obvious, we see in a horizontal format, cropping the image in a rectangle format leaves us the sharpest part of the image and it only makes common sense when it comes to printing.

I’m just telling it like it is, I am that I am. I hope that helps!

Thahis for the A2A.

Bottom line is people want square pictures most of the time just out of habit, and image sharpness from lenses is so horrible at the corners of the image circle that you might as well just use the center and have a better picture.

Here's a square within a circle:

As you can see, making a square makes sense since it fits within the circle, or the optical limits of the lens. It will create this type of image:

It's uniform, it allows for image framing and it fits the confines of what early photographers felt paintings should be presented as: rectangles. Nobody made triangular or trapezoidal paintings back in the 1800s, and circular or more commonly oval frames were used for portraits.

Now, for comparison, here is a circle within a square:

Notice the wasted space and the lack of framing options. Try making a beautiful well-framed photograph working with something that will turn out this:

See how it vignettes towards the edges, and the fact that the edges are insanely blurry. Now try framing something, like the ocean.

Whereas with a square sensor you would get something like this:

Something which has a central plane, the horizon, and a vertical plane, the sun and its reflection. Now try a circular photo of the ocean:

It's kind of hard to do things with that, with the added problem that any lines you photograph will have to match up with the circle's axes to be aesthetically pleasing. This ocean's horizon doesn't, so it instinctively looks really weird (to me, at least).

So finally, why do we use rectangular film when we could be using circular film? Because when you want a circle, all you have to do is frame the camera correctly and you can either have a square or circular portrait:

So having square sensors/film sheets actively benefits picture making while circular image fields would be of no practical use.

This is a good one because I love optics. I also have a great picture of the human eyeball that helps explain stuff. Lenses are amazing. All kinds; single reflex, fresnel, just cool stuff.

What’s “fresnel”? I may get to that later. Here’s my non gross picture of an eye.

Not to worry about rods and cones on this one. The only cone you should imagine is a cone of light. See, the back of the lens (the focusing lens) is special and these yellow lines show how light enters a round lens. That light from the front of the lens is actually being “focused” into a cone shape. The “bottom”, or narrowest, part of the cone is going to shine on the “film plane”, shown here just in front of the optic nerve.

(that’s your film in a film camera or the Cee-Moss sensor: CMOS, on a digital back.)

That’s basically what happens. Now you have a circle of light that can expose the film or be picked up by a digital RGB sensor. The crazy part is that the light is always a circle (even if you make the lens square). Ever try to burn leaves with a magnifying glass?

This has something to do with a complex system of physics and photons that you don’t really need to know. But, if you’re a photographer that wants to sound smart, spin a verse and drop knowledge.

The inverse square law.

Like I said, just remember that light always always lands in a circle when it is focused for photography. So what in photography is square? I’m going to open that part with a bit about film cameras. In this respect they are similar to digital cameras. Only instead of a “film plane”, the digital system operates the CMOS. That digital sensor is basically digital “film”.

So what part of a camera is a square?

The aperture for a first off. Just notice how it makes a more squarish shape. This is the first step to shaping light. That way you get as much of the scene as you can, without losing too much around the edges. This is also much more important if you use large format film cameras with the bellows and the cloak.

The light entering your camera, bound for the film, first passes through the “aperture”, then through the “shutter” and then it sinks into the emulsion (sticky silver B&W or sticky dye layers for C41 or color film, CMOS for digital).

Image above is property of the Valve Corporation (it’s educational I don’t think they can sue me.)

The 2 Ways Camera Lens Aperture Is Used In Digital Photography

The aperture is what shapes this “light cone”, and can be adjusted to allow more or less light depending on the desired effect. Now that the light wave photons have hit something solid they can change shape. Sort of like making shadows. You can see this shape in your photos when the sun flares off the front glass making the forever all time favorite Photoshop effect… the lens flare.

But when you push the shutter you hear a click. What was that sound? In most cases it is the sound of two plastic “panels” that are spring loaded. When a film camera is “cocked” these two panels are drawn into position. When the shutter “fires” both panels slam past each other and make that distinct sound.

This is the clever bit. Each panel has a square hole in it. When the panels race by each other they allow a square opening to expose the film to the aperture light. The idea is for each part of this shutter system to allow a measured amount of time to pass as well as further “shaping” the light cone into a square.

When it comes to a DSLR camera shutter there are 3 basic mechanisms: the mirrorbox, the bottom door, and the top door. When you look through a DSLR view finder you are essentially looking through a series of mirrors that get their light directly from the lens. When you click the shutter button that system of mirrors flips upwards to allow light to pass to the sensor. This is why the viewfinder goes black for a short amount of time when taking photos.

The Science of Shutters: How Camera Shutters Work

Thats about it for DSLR and film single lens reflex. Curtain shutters make less noise and operate differently (mostly Leica models), but I haven’t seen a digital curtain system, yet.

As far as the mechanics of light and optics, it might be easier to imagine a circle of light coming from the lens. The edges are dull and blurry so you just want the middle part. So that’s what the shutter gets. Kind of like putting a square inside a circle. The square is your film.

If you want round pictures use round film.

/Jon

The question should be the other way around… Why do we make the image rectangular, even though the lens produces a round image.

The reason is because we have two eyes. And thus we prefer to look at rectangular (Or at the very least, ellipsoidal) images, and not at round images. We see more width then we see height.

Long before the camera was invented, we have been making paintings. And those are all rectangular, not round. Round doesn’t suite our vision very well. Elipsoidal works, but rectangles are easier to frame, and we like the effect. Even when painting on a piece of wood, which could be any shape, we prefer rectangles.

So when photography was invented, even though the lens produces a round image, we made rectangular photosensitive surfaces. With roll of films, there is even more reason to make rectangles, because it makes for easy rolls of film.

With digital, rectangles are also easier to produce.

We could use squares…. but those are unpleasant to watch. Rectangles of much nicer to view, either horizontal or vertical.

So it’s a combination of ease of production, and preference to view, that we use rectangular image sensors.

Having the lens make rectangular images is no use. It make the production of the lens 10 times more expensive, for almost no benefit. The only part of the lens where we do take the image rectangle into account, is the lens hood. (The petal shape lens hoods on wide angle lenses. ) Because there is works better, and cost nothing more. But cutting a round lens into a rectangle is a huge job, for no gain. And it makes focusing much more difficult, as you wouldn't be allowed to rotate the lens.

So we make the image sensor a size that is both easy to produce, and pleasant to view. And we keep the lens circular, because that it way more easy to produce, and doesn’t do anything bad to the image behind it.

This question (in its various versions) has more to it than most answers give credit. Why do we waste over 40% of the information coming through the lens by not having a circular sensor ….

1. How do humans ‘see’ ?

There’s an important point made by Dave Haynie arguing that humans’ normal stereoscopic vision – the field of view we have in our normal outdoor existence – is about 180 degrees horizontal and 130 degrees vertical – more ‘rectangular’ than square or circular. He adds “Here's a chart of a visual field mapped by NASA way back in the 1960s. The central white area being the area of overlap between the two eyes.”

So it’s no surprise that much classical art is in a rectangular format. You will find circular portraits and even a few circular landscapes but these are rare. There’s an argument you sometimes see that humans find shapes whose dimensions have the ratio Phi (1.618, ‘the golden number’) particularly pleasing which as an ex-mathematician I find intriguing but the rationale rather dubious. https://www.goldennumber.net/art-composition-design/

2. History

Once upon a time the film at the back of the camera (and developing and printing it) was a major expense for photographers. Using it efficiently mattered: While cropping was certainly possible even in pre-zoom days many people didn’t bother and happily used the Leica 36 by 24 mm format (or one of its imitators like Kodak instamatic’s 26.5 mm square) as a given for all their prints or slides.

Then along came the SLR (later the DSLR). A mirror flapping over a circular or square sensor would have to be bigger, heavier and more expensive than one flapping over the short edge of a rectangular sensor. Result: More rectangles and fewer squares.

3. Today - cost of sensor

Those flapping mirrors in DSLRs are still going to feature for a few more years but long-term the only relevant hardware cost (apart from the lens) will be that of the sensor. Here cost is roughly proportional to sensor area – so from simple trigonometry a circular sensor that contains within it a full frame 3:2 ratio rectangle would be about 1.7 times as big and and 1.7 times as expensive (or if you prefer for a 4:3 sensor and square sensor the ratios are 1.64 and 1.57, respectively). So the extra cost of enabling your full frame camera to allow for the full circular image to be captured could be as low as 70% of the current cost of the sensor. (88% if, as some argue, the cost of a sensor is proportional to the size of the template out of which it is stamped).

4. How much time do you want to spend on post processing ?

Imagine you have a camera with a circular sensor – you’ll need the software to let you choose between a circular image, a ‘landscape’ horizontal rectangle, a ‘portrait’ vertical rectangle and maybe a square too but more importantly you’ll need to learn how to use it. Or spend time in cropping your circular image after taking the picture. Most people won’t want the extra bother – in today’s market anything more than one click to take the picture and one or two more to post it online reduces market share …

5. Image quality – even circular lenses aren’t all entirely circular

If you examine lenses you’ll find some of the more expensive ones are already limited to produce a smaller image than a circle. Rectangular lens hoods and rectangular internal baffles are designed to cut out stray light that would downgrade the quality of that rectangular image. Sure, these lenses could easily be produced in a circular-sensor friendly version with circular lens hoods and without the rectangular internal baffles but then any rectangular images would be of lower quality because of the additional flare (stray light bouncing around inside the camera). You can’t have your cake and eat it.

6. Commercial reality

Technically having a circular sensor is easy and more fully utilises existing lenses than today’s rectangular ones; as sensor costs continue to decrease the bigger sensor won’t be prohibitive but most people want to end up with rectangular images most or all of the time and cropping from a circular image is a hassle so you’ll need some clever software in the camera to help you out. How do you market the idea ? Well, there’s no longer any need to rotate the camera 90 degrees for a ‘portrait’ picture. You never have to worry about level horizons when taking a picture - you fix that later just by turning the circle ! The lens mount could be the same as any existing lens mount so there’s no need to produce new lenses – though lenses with rectangular baffles wouldn’t be fully compatible and cause customer confusion. The viewfinder conundrum could be solved by adapting Leica’s rangefinder concept - again, don’t we just love doing that ! - but maybe someone has an even better solution. It may yet happen, but don’t expect it to be a big seller. http://photographyhappenings.blogspot.nl/2010/04/rokton-announces-circular-sensors.html

Historically, painted pictures and drawings that predated photographs used rectangular canvasses.

Early cameras used boxes for exposing the film, with rectangular film holders. Photosensitive film was made in sheets or strips that were cut to rectangles; circles would waste material.

The film area is thus traditionally square or rectangular. And so it has followed that digital sensors are also rectangular. Following a tv standards, digital images are arranged rows and columns, decidedly rectangular.

The image cast by a cylindrical lens is round… the image overlaps the sensor/film sensitive areas. The extra image falls on the black walls (they are black and baffled so that reflections won’t cause fogging of the film and loss of contrast) and is lost forever.

We mostly use rectangular pictures, prints, images in articles, albums and internet. Don’t think its going to change.

No matter what type of lens you use (let it be a wide angle or extreme fisheye lens of a very low focal length) ultimately the image will get formed on the sensor, which is always rectangular in shape. Or in other words while taking a photograph we convert a multi dimensional world of infinite resolution in to a 2 dimensional digital image of a fixed resolution. (depending upon the megapixel count of the sensor)

Since the time when Photographic Films were used for taking pictures, the final sensor element or the film has always been traditionally a plane/rectangular surface.

Also one major technical reason : Sensors are made using semiconductor fabrication techniques, in which they print multiple sensor circuits onto a silicon wafer. Today these wafers can be 300 mm in diameter and manufacturers are moving toward 450 mm diameter. A lot of sensors can be printed on wafers that large.Sensors are tiled onto the wafer to efficiently use the space available and in a way that makes them easy to cut apart into “dies” (or the individual sensors, in this case). The process is called dicing. The most cost effective shape for dies is rectangular. Usually a saw or scribe is used to cut the wafers in straight lines. Imagine if the dies (sensors in this case) were supposed to be round (the cutting cost will be too much and also it will lead to more wastage of the wafer material).

Here i am also attaching the sensor size comparison of various standard sizes for your reference.

Since the sensor is made rectangular, now you will think that why do we make our lenses round and not rectangular ?

Lenses made from high quality glass are generally ground using lathes. It is easier to spin, grind, and polish a round lens in these machines because there are no edges to catch on the tooling as the lens spins around. It also is consistent with the optical symmetry they are trying to achieve in the finished lens.

Lenses that are not round would generally be cut from round lenses, again a step that adds cost to the production of the lens assembly. Lenses don’t necessarily need to be round.

As i have stated above, the lens shape does not determine the image shape or quality (apart from diffraction caused by edges, which can be mitigated, and some second order aberration effects, maybe), and for the most part, every point on the lens can gather light from every point on the object and focus each point on the image plane. I’ve already pointed out that changing the shape of the lens adds cost. There really isn’t any practical reason (generally) for changing the shape.

One more reason: The light gathering capability of any surface is largely governed by the exposed area, whereas some of the optical quality goes down (or it is more expensive to correct to the same level) with the maximum dimension. A circle minimizes the maximum dimension for a particular area.

Hope this helps.

Its easy to cast and polish glass in a circle for lenses, then they need to be coaxially aligned as opposed to being aligned in 3 dimensions for rectangular glass.

OTOH, films and sensors are easier to lay out in rectangles and rectangular arrays. Hence the images are made rectangular. The lenses cast an image circle but the sensor always fits in the circle

Simply?

My eyes (and the lenses in them) are round, yet when I sit in my living room and look out of the window, I see a rectangular view of the outside world.

There has to be an edge to everything - it’s called framing.

Humans were designed to see things that ensured survival. Most of our adversaries were ground-based creatures. Our optical field of view is circular but our brain is most sensitive to things that lie in the horizontal plane (lions, tigers, and bears) because that is where most of our threats come from with a relative few coming from above. Our normal ‘frame’ of reference is a rectangular view of our world.

The reason camera manufacturers initially settled on the horizontally-based, rectangle-shaped format is that its appearance is the most natural for us to perceive our real and photographed world.

The sensor that captures the picture is rectangular.

The rectangular central part in the above image is sensor. You can visualize that a conical beam of light falls on the inner circular part of the camera shown in the picture. Since sensor is housed in the central rectangular part of the image plane, the output image is rectangular in shape. Had the sensor been circular, the image would have been circular too.

I can't imagine a camera lens being anything other than circular. As a lens, the image will always be circular even if the glass elements aren't. So lenses have stayed circular out of feasibility: round glass elements are easier to make; they allow for better movement of elements for zooming and focusing; and aperture diaphragms are more easily housed in tubular lenses.

The real question would be: why are image sensors not round? And indeed, some very old cameras like the early Kodak box cameras shot circular images. I suppose, as film became more popularly available in reels/rolls, it was only natural that images were shot in rectangles and squares to use more film space. Similarly, round images probably didn't lend themselves well to printmaking. In any case, rectangular/square images became the standard, and digital simply continues the tradition. For digital cameras, particularly DSLRs, rectangular 35mm film format is still very much the measure by which digital sensors are compared to.

Why is the camera lens circular but the image sensor is rectangle in shape?

Sensors are rectangular by convention, in view of the generally customary state of picture media.

Be that as it may, there is an innovation/business choice that drives them to be rectangular, moreover. Sensors are rectangular in light of the fact that they are made utilizing semiconductor manufacture methods. These strategies call for "printing" various sensor circuits onto a silicon wafer. Today these wafers can be 300 mm in distance across and producers are advancing toward 450 mm width (see here). A ton of sensors can be imprinted on wafers that enormous.

Sensors are tiled onto the wafer to proficiently utilize the space accessible and such that makes them simple to cut separated into "kicks the bucket" (or the individual sensors, for this situation). The procedure is called dicing. The most practical shape for bites the dust is rectangular. Generally a saw or recorder is utilized to cut the wafers in straight lines. Envision if the bites the dust (sensors for this situation) should be cycle (an inefficient and exorbitant utilization of the material) or hexagonal (effective utilization of the material however the cuts are not straight over the entire wafer). (See here for more information.)

B) Lenses made of top notch glass are for the most part ground utilizing machines. (This can be found in this video. Watch around the 7:00 moment mark specifically. Apologies, it's in Japanese, yet the video is captivating and uncovering.) It is simpler to turn, pound, and clean a round focal point in these machines on the grounds that there are no edges to get on the tooling as the focal point turns around. It likewise is reliable with the optical evenness they are attempting to accomplish in the completed focal point.

Focal points that are not round would for the most part be cut from round focal points, a stage that adds cost to the creation of the focal point gathering. Focal points don't should be round. For the good of heaven, most eyeglasses are not round! At the point when your eyeglasses are made, you should know that the focal point creator isn't loading a focal point for each state of eyeglass outline. He's slicing or crushing round focal points to fit the casing.

When the focal point producer has his round focal points, what might rouse him to cut it into an alternate shape? The same number of individuals have called attention to in different gatherings, the focal point shape doesn't decide the picture shape or quality (aside from diffraction brought about by edges, which can be alleviated, and some second request variation impacts, perhaps), and generally, every point on the focal point can accumulate light from each point on the item and spotlight each point on the picture plane. I've just called attention to that changing the state of the focal point includes cost. There truly isn't any viable explanation (by and large) for changing the shape.

The lens may be circular but the sensor in the camera that actually captures the picture (i.e. converts the light energy to electrical signal) is square.

The Lenses are circular

But the sensor inside is always square so that the image produced can match the shape of display devices (LCD, CRT)

The lenses are made circular to provide minimum distortion of the light beams (that constitute the picture) which are passing through it.
Also, it's easier to manufacture (i.e., spin, grind and polish) the round lenses as they have no corners.

Because the sensor is rectangular!

(Image from canonrumors.com)
(Pictures are rectangular, not square)

The shape of the lenses comes from the manufacturing. Normal lenses are spherical lenses (that makes them easier to produce) therefore they are round. You could cut out rectangles from them If you wanted to.
The shape of the sensor also comes from the manufacturing. There is more than one sensor on a silicone wafer and at the end they cut the wafer in pieces. That is simplest when this pieces are rectangular.
What the lens does is that it focuses one point of your object on one point on the sensor. The shape of the lens and the sensor do not have to correspond. You could have a rectangular lens and a triangular sensor. That would be strange but it would work.

The image indeed is circular. However the sensor or film on which it is focused is rectangular and hence the shape. The central part of the focused image is the sharpest and hence the rectangular sensor. Bigger sensor sizes require bigger and heavier lenses. They tend to be very expensive too.

For eg. U can compare the size of 50mm f1.4 fx vs DX lens. Or the equivalent 35mm(1.5x) DX lens.

Also, the image on the periphery is not too sharp. Due to lens defects/ optical defects and is easier to ignore it then make a perfect lens.

PS: imagine a reel of circular negatives. Circular shutters, special apparatus to develop the circular negative and last the circular prints. Our eyes are designed to see rectangles better. Else, we won't be buying wide screen home theatre systems.

Lens does projects the circular image. But due to the rectangular shape of the sensor (which record the image data) makes it rectangular.

The rectangle on the image shows the secsor area.

image credit : Photography Life

Square images from a circular lenses have been around for at least the last 100 years. Lenses focus light in a cone. When the cone gets very small, it becomes “the circle of confusion,” which is the point of focus. Different sizes of films or sensors require different sizes of circles of confusion to look sharp to us.

Many cones are required to create the given lens’ round “Aspect ratio.” The film or sensor must fit within the aspect ratio. Thus, you have a square or rectangular shape INSIDE OF a CIRCULAR shape.

Sensors are different from film in that they have photo sites which are sort of like pits. Light is best if it enters vertically. This requires lenses having “telecentric” design, meaning the focus happens up in the lens and exits the lens as straight columns of focused light, best for sensors. Zeiss’ lenses, even wide angle ones, use a telecentric design best for sensors. Regular ‘old fashioned’ wide angles without telecentricity are not so good for sensors because they project their focused rays at an angle, only good for very flat film.

Imagine round images. Now imagine the film required to make them. It would be a a strip with a series of circular images and large areas of unexposed film. It seems a wasteful way to make film. Before film the medium was mostly glass. Round glass rolls around. That’s inconvenient. By the time film came around people were accustomed to rectangular images. There was no reason to make round images with the waste. By the time we got to digital rectangular ruled the world. Not only was the shape standardized but even the sizes. Now, round pictures look odd and I no always wonder what was left out of the corners.

As with most things in life, just follow the money.

Circular picture frames are far more expensive to manufacture both in labor and materials. A rectangular frame is four straight sticks fastened at the corners.

Ever try to cut a perfect circle with a scissor? Circular film or paper requires an expensive die to cut efficiently, leaving a significant amount of waste. Rectangular shapes can be easily sliced off a master roll with a cutting wheel and straight cutter leaving little waste.

Using rectangular shapes on a roll of film maximized image size (and thus quality) and lowered waste.

My fellow seniors will remember that TV tubes were originally round, yet the picture was rectangular. Transmitting an rectangular array is far less complicated than a circular array. Even in today’s digital world, rectangular TVs and monitors cost less to manufacture.

And, as several answers have pointed out, glass lenses are easier to manufacture and operate when they are round. So round lenses cost less. The light they “waste” above and beside the rectangular film/sensor doesn’t cost a thing.

So matching a rectangular picture with a rectangular lens would appear the more elegant a solution, but more expensive to execute. Now on to the obvious follow-up question, why do round wheels go on rectangular cars?

The outer part of the image circle gets more blurred and distorted the further you get to the edge. A square image inside the circle will have a good percentage in the bad area. A rectangle can have more area. This image shows a 44mm image circle, the white rectangle shows a 24x36mm image frame. The black square is the ideal format on 35mm film, but as you can see its very small 24x24mm at most. Rectangular images make the most of the image circle with any format of film. If you had a 36x36 sensor your image circle would need to be much bigger and lenses would be correspondingly bigger

I assume you are talking about DSLR cameras
The picture is produced by a digital sensor which is inside it and this is the reason the picture comes rectangle. The circular part of camera which is the lens has no part in developing the picture, it just used to get the light rays on the image sensor.
Now why is lens circular ?
It is circular because of the following reasons

1. DSLR needs different lenses for different purposes, so it is easier to make a circular mount of a fixed size than the rectangle and one has to lock the camera and lens, which happens when you rotate it.
2. It is easier to grind and fix the circular camera lens since there is no rectangular edges to cut.
3. one has to rotate the lens ring in order to zoom in and out.

Film could be round - there's no real limit there.

Digital sensors are generally rectangular because pixels are laid out in grid like manner. It would be extremely difficult arrange and process circular arrays of pixels, and it would waste space, both on your screen, and on the layer of silicon it's being made from.

Lenses are generally round - there's a few non-moving front element lenses that are square. One major reason is the lens elements have to move. They move by turning in a helicoid structure. In many lenses even the front or rear elements rotate. In those cases the lens must be round, with the diameter enough to accommodate the corner-corner coverage of your sensor.

Here's another reason. Within the lens you can have some total internal reflection occurring and light bouncing around a bit. If you were to cut lenses to just fit the image, then you introduce large flat surfaces for these rays to bounce off of. It can introduce from very strange artifacts when bright spots of light sources are introduced.

Lenses are circular in shape because this makes focusing and adjusting focal length (on zoom lenses) much simpler. This is because focusing and zooming requires changing the distance between the various pieces of glass inside the lens.

Doing this is achieved by the lenses moving a screw-like fashion, which is why you’ll see photographers and videographers twisting various parts of the lens to change the focus or focal length.

The sensor, on the other hand, is rectangular because in the early days, using a rectangular image maximized the area being used on a film strip which was important since film is costly.

I believe it really comes down to what is most efficient to build.

Lens surfaces are (mostly) spherical sections because such shapes can easily be used to focus light to make an image and it is easier to grind and polish a surface to be a perfect spherical section — to have constant curvature — than other shapes (parabolas, etc., which are called aspherical surfaces). These lens elements don’t need to be made from circular blanks, but it is natural to shape them that way because then there are no corners to cause issues during grinding and polishing. The image naturally produced by a circular lens is circular, not rectangular.

However, the substrate for photosensitive materials (wet plates, film, and even image sensors) is cut from flat sheets of material. To avoid wasting that material, you want to cut those sheets into a pattern that can be repeated to cover the material with no gaps or overlaps: you want a tessellation. The simplest shape to cut that can tessellate a sheet of material is a rectangle. That shape is also easier to build a frame around than a curved shape (e.g., a circle); a rectangular frame can be built from just four straight pieces of wood with almost no waste.

The circular image produced by a typical lens isn’t necessarily a circle with a sharp edge, but generally fades (darkens) and becomes less-well-focused as you move further from the center. Thus, the lens selected for use with a particular rectangular image format is one that projects a relatively bright and well-focused (well corrected) image with a diameter at least as large as the diagonal of the rectangle. For a rectangular image (shown outlined in red below), the coverage circle of the lens might look like:

You could have a circular picture. Kodak’s first film cameras had it. But because the lens is also lenticular in shape seen side on, the picture it produces wont be in focus at the edges and it will probably be a little darker. This is due to the properties of the lens, in addition the lens will have a stronger prismatic effect so the colours will not all be in focus. These effects are called spherical aberration, vignetting and chromatic aberration. They can be corrected with extra lenses made from different glass and shaped differently but cutting the image down will remove most of these aberrations. Rectangular is the most common shape.