Recently Lee published a comparison between the images from several different-sized cameras and there was no apparent difference. What's the point of buying a camera with a larger sensor then?Sensor size matters. Sometimes. You won't have the need for a bigger sensor until you know the technical side of it and find yourself in a lot of situations you would benefit from a camera having such a sensor.
The Sensor In a Nutshell
While there are "by-products" from the size of the light-sensitive chip, in general it's just a rectangular device that reacts to light. The size of the all-in-one captured frame depends on the size of the rectangle. The bigger the rectangular area, the more you will capture from the scene your camera is pointed at. It is like the window of your home. The bigger the window, the more you will see from the outside view. A crop sensor is smaller than a full-frame sensor. That's the main difference.
When You Won't See the Difference
If you photograph mainly in large spaces or outside using enough light, available or strobes, and have a modern camera, you probably won't feel the need to buy one with a big sensor. Big sensors are often less noisy at high ISO, but modern technology provides us with cameras where even micro four-thirds ones give us quite a good quality of image in not-enough-light situations. This is why I said the "window size" is the main difference, not pixel peeping.
Two photographs with two different sensors: full frame and crop. You can't see the difference.
Wide-Angle Lenses
Can't we use wider-angle lenses when we have a smaller sensor? Won't we see more? Think of using a wider lens is like squeezing your clothes in a small travel bag. You're getting your stuff in, but you don't want to put any of these clothes on without ironing them first. The same with wide lenses: they will give you more of the environment, but will squeeze the center part of it, because there's no place to fit the rest of the view and thus distorting the reality of distances between the objects in the frame. With a larger sensor you will be able to use a tighter lens, distances will be closer to reality, and you'll still see more than a smaller sensor.
Shallow Depth of Field
Most of you know, but let me say it once again: larger sensors don't give a shallower depth of field. A 50mm lens will give the same optical image regardless of the sensor size. It's the sensor that will "crop" part of the circular picture, given by the lens. All rectangular sensors do. If you have a tight shot on a small sensor and want to have the same with a bigger one, you have to get closer to the subject, because the "window" is bigger and it will show more than you need. This will naturally make the photograph with a shallower depth of field, because of the diminished distance from the camera to the subject.
Can You Simulate a Bigger Sensor?
Absolutely, but not always. If you photograph a tight interior you better not use a wide lens, but create a panorama out of several images. This will do good to the interior designer, to the viewers (who won't be fooled by the distorted perspective of a wide view), and to your portfolio. You can do the same for landscapes and even for portraits. It's a big tricky for the last case, but if you're budget is tight, you can get around that with some tedious post-processing work.
Shot as two separate images (left and right), because I had my back against the wall with a crop-sensor camera.
So, When Would You Need a Bigger Sensor?
In my opinion and experience, it's when you're constantly in low-light situations (although not such a strong argument) and in tight spaces and you don't want to use wide-angle lenses, but capture more of the view. If you have experienced more cases where utilizing a larger sensor was an absolute necessity, please tell us in the comments below.
I shoot real estate photography. In my early years I was on a 7D with a 10-22, which was fine for the little experience and learning I was doing on the fly. Fast forward years later and I could not imagine giving up my 16-35 or my 17mm tse for the 10-22, nor would I want to pay the price of the 11-24 to compensate for the crop factor. In this case I see a significant difference in sensor size.
That said, as I have found myself moving into more architectural work and wider angle shots are way less desirable, I have considered picking up an a6400.
Exactly. Sensor size matters for your type of work.
I remember years ago the saying...'The camera doesn't matter...'. Well...does it? Open question for all. What DOES matter is - the lens. That is for sure.
The camera doesn’t matter when you’re first starting out. The reason why many teachers say this is because many people coming into the photography industry get transfixed by gear and forget about the art.
As you progress and learn how to manage light and compose correctly the camera does indeed begin to matter. It matters a lot when you have time constraints and demanding clients. It also matters when you have clients with specific requirements.
It’s much more nuanced than this but the camera does and doesn’t matter. It depends on a number of factors.
Schrödinger's cat Lol. It’s quantum.
Schrödinger's camera
Can't say it better than that.
Does it though? Computer aided design and computer aided manufacturing mean the IQ of modern cheap lenses is still really good. I can't help but think that high resolution lenses and sensors only become meaningful in edge cases, where the user is milking every last bit of performance out of them. I always think the biggest benefit is build quality.
I'll qualify my statement.....what matters is the "light-gathering device" however it's manufactured. And, yes, you have proven the point. When you said that computer aided design and manufacturing..bla bla bla..have made IQ of cheaper lenses really good, then those lenses which have NOT gone through that process and any existing lenses (and there are so many of them, old ones especially) will fall into the lesser quality lenses. Unless manufacturers want to go through the redesigning process of old lenses. Glass DOES matter. We're not talking about new glass....ANY glass.
Although I take your point, I can't help but reflect that one shoots old glass (or pinhole) and/or film because it has certain qualities.
"Sharpness is overrated" ~ Keith Carter
Edit: even by this standard, glass matters.
"Sharpness is whatever a photographer deems it to be". And yes...in the end glass does matter.
That is clearly not the case (resolution and acutance); however, whatever works for you.
In the technical sense yes..I agree. But what works for one person may not be the same choice for another.
it matters a lot :) ... in some cases more than lens ..
SO... for the same field of view at the same distance, depth of field is shallower on bigger sensors... and let be real we all crop with a lens to the look we want and often restricted where we can walk, so we zoom in or out to get the shot we want... so the same field of view, different focal lengths to get the same shot. The Field of view and how I crop with the lens is a key part of my photography, as a portrait photographer I don't want to get too close to the model with a wide angle lens due to the perspective distortion that comes with it to get the same shot I get with my medium format camera. I'm likely to be at the same distance from the model whatever camera i'm using, this is likely to end up with different looks if I was using an iPhone, micro 4/3, aps, FF, MF 43.8 x 32.9, MF 53.7×40.4mm etc etc… just different focal lengths... :-)
I'm going to just let you all have this discussion. I'm out.
- signed SPDH (Society for the Protection of Dead Horses)
Good one!!
thanks to the internet, dead horses are being flogged on a daily basis #horsestoo
It's not really about sensor performance for some. Doing events would be super difficult for me without all my customized buttons which I have committed to muscle memory. My old a6500 just didn't have enough convenience controls. Last place I would want to be during a wedding shoot, is b@lls deep in a Sony menu, while missing once in a lifetime shots.
Gear matters, be it sensor size, body design, lens quality, tripods, flash, etc, etc. These are tools and if you want the most efficient, durable, dependable equipment, it comes with a cost. If you're serious about photography or any hobby, buy the best that your budget will allow.
The bit about WA lenses is nonsense.
"Think of using a wider lens is like squeezing your clothes in a small travel bag" WTF?
A small sensor uses a "wide angle" lens insofar as it is a WA lens in relation to a larger sensor.
50mm on a 6x7cm film camera is very wide yet no one calls a 50 wide on a 35mm frame.
A a17mm lens may seem like an UWA on a FF camera but provides a 35mm AOV on a µ43 sensor. Hardly weird or exotic.
If you use a µ43 camera you choose lenses that have the AOV you need. The FL tells you what the AOV will be on that camera. I use a Canon 11-24 extensively because I need UWA AOV on FF. If I were to use it on a µ43 camera the AOV would mimic a 22-48mm lens as far as AOV.
For the same AOV you will see the same levels of distortion or lack thereof.
This only adds to the confusion about sensor size rather than clarifying.
The physics of the optical transformation in wide-angle lenses shows that you have to put more stuff on one place. Wide angle - sees more, but the canvas, where light is projected, is the same. This means detail has to be squeezed (made smaller) in order to fit on the same canvas. I'm not talking about distortion here, because there are wide-angle lenses without distortion. As it doesn't project the real perception we, as humans have (which perception is based on the optical distortion of our natural lenses - the eyes), it looks unnatural and thus doesn't present reality as is: distances look exaggerated.
Wrong, please stop peddling this nonsense.
17mm (or 17.5) m4/3 will look like 35mm on FF. Aside from DOF. End of story.
Perspective distortion, exaggerated distances, etc. are by-products of the angle of view, NOT the focal length.
Please stop spreading falsehoods. If you don't believe me (and everyone else here telling you that you are wrong) - give it a try and post the results. Take the same picture from the same distance with two lenses on two different sensor-size cameras that have the same equivalent focal length (i.e. AOV). Provided they are both decently corrected lenses, all you will notice are DOF differences (with the same f-stop anyway).
So glad someone else read that and threw up a flag on the play.
After trying both, I have to say that I’m pleased with the full frame.
The only situation I found I wish I had a crop sensor was in Iceland when I wanted to get longer exposure not having found the ND filter for my 14mm Samyang AF. The sensor was too big and the slowest I could get was 1/20.
It all depends on what kind of work you're mostly doing. I've been using both and I still use a crop sensor. I do commercial portraiture when it comes to stills and full frame doesn't justify its price for my type of work.
I don't understand the argument about wide angle shots. If you use a 24 mm lens on a FF-body, but a 12 mm on a MFT-camera. Problem solved!
The whole issue is about the exaggerated distances the wide-angle lenses give (I'm not talking about the distortions). If you want to show an interior as is without lying the viewer it about its dimensions, you need to use a glass that will see it as your eyes do. In this case small interiors won't "fit" the sensor and you need to have a bigger sensor or to make a panorama.
Never heard of field-of-view have you? That's what it's all about. It has nothing to do with sensor size.
Let me put it this way. You put a 40mm on a full-frame and on a crop sensor. One "sees" more, because the light is projected onto a bigger canvas. If you put a 24mm on both full-frame and on a crop sensor it will do the same, but the image will have a 24mm look, i.e. not the way the eye sees it. If you have a situation where the architect of the interior wants an image the way the eye sees it, go and tell them that they don't understand what a "field of view" is and that you're going to use a 24mm on a crop sensor (for example) and let them deal with it. The task is about making a normal-looking shot, not about measuring distance from lens to sensor and lens' focal length.
It's about getting the job done and in this case job is done by using the "natural" looking lens on the appropriate sensor size and make a single shot of the interior OR if you don't have a big-enough sensor, make a panorama with that same lens.
This is not remotely accurate. What is a 24mm look? There is only FOV and of course, DOF changes with the actual focal length, but I have no idea what you're talking about in regards to seeing more or the "way an eye sees it." Does the eye "see" like a full-frame camera? A medium format? APS-C? I have no idea what this means, because it's wrong.
The only exception to what you're talking about would be distortion of ultra-wide angle lenses, but most modern lenses (especially m4/3) are incredibly well corrected. A very good 12mm M4/3 lens will look like a 24mm FF lens (minus DOF differences).
All that is is an argument that you'd have to use a different lens to get the same FOV on a smaller format. It isn't an advantage for full-frame at all. The look of a given focal length is 100% based on how far the optical center is from the subject. The actual focal length doesn't matter. Most cell cameras have a focal length of about 4.5mm, but the results are indistinguishable from a 35mm camera with a 28mm lens standing in the same place, albeit at a high f number.
A 24mm on APS-C will not have a "24mm look", that's insane. For example, when you match the FIELD OF VIEW: https://fstoppers.com/originals/can-you-guess-which-camera-took-which-pi...
This is a misconception. How natural or not natural a shot will look does not depend on the mm of the focal length. A 24mm on a smartphone sensor will be a long portrait lens that will give the nice portrait compressed look (not much bokeh though).
"Big sensors are often less noisy at high ISO"
Nope they are not. Lens aperture makes the difference, not sensor size.
Bright optics for lager sensors are simply easier to make, you can build an 85mm f/1.8 but it's way harder to build a 45 f/0.9, that's why M43 optics are usually actually MORE expansive than their FF counterparts.
"With a larger sensor you will be able to use a tighter lens, distances will be closer to reality, "
Kidding right? Sensor has absolutely no impact on perspective, squeezing etc, that's all down to the distance from the subject and lens distortion correction. A 12mm on M43 (ignoring the different aspect ratio for simplicity) has the exact same framing as a 24mm on FF.
Nobody is talking about distortion here, nor about sensor affecting the physics of the optics. The sensor FORCES you to use certain optics in certain situations. If you read it carefully, if you want to capture an image on a small interior the way your eyes see it, you need to either make a panorama (simulating a bigger sensor) or use a bigger sensor. The wide lens (without any distortion) will not present a reality although you may have the same objects in frame as on a bigger sensor with a more "normal" focal length.
That's not what you are saying in the article.
You are literally saying that "they will give you more of the environment, but will squeeze the center part of it, because there's no place to fit the rest of the view and thus DISTORTING THE REALITY OF DISTANCES between the objects in the frame".
Small sensor forces you to use a certain optic as much as a bigger sensor forces you to use a certain optic to achieve a desired FOV at a given distance. There is simply NO WAY to tell from the geometry and perspective of the image alone which combination of sensor size an focal lenght was used.
I've added a test image I shot a while ago, can you tell roughly at which focal lenght this was shot and what sensor size have I used?
"With a larger sensor you will be able to use a tighter lens, distances will be closer to reality, " What does that even mean? What distances? What reality? Are you suggesting that you get less perspective distortion by using tele lenses? Cuz that's not how it works.
The context of the article speaks about wide-angle lenses. This is a perfect example why context is important and how taking something out of context can be twisted the way you want it. Tighter lenses in that context means tighter lenses, not long lenses. The context also talks about "close to reality." Having said that, close to reality depends on what you're currently using as a sensor and what your eyes see. If you're on a 8x10 camera you will use a "tele lens," yes. If you're on a standard 1.6 crop sensor you will use a different one. The goal is to make it close to reality.
What you show as an image is an example that you can deceive the viewer perfectly, because sometimes it's hard to guess if that's the reality or not. This is why I gave an example with an interior which usually people can visit and judge by themselves if they saw a false advertisement or saw an image identical to what their eyes see when they are on that location.
Still don't explain what "close to reality" means, and my eyes definitely don't see stuff like in your image, whatever...
And again I repeat, you spefifically said that a larger sensor "will give you distances will be closer to reality" which I've explained why is false.
"Absolutely, but not always. If you photograph a tight interior you better not use a wide lens, but create a panorama out of several images. This will do good to the interior designer, to the viewers (who won't be fooled by the distorted perspective of a wide view), and to your portfolio."
You do know that making a panorama with a telephoto (or longer focal, call it however you like and don't move the subject to semantics) lens doesn't change perspective, "distances" (whatever that means) compared to a wide angle right?
"Close to reality" means that you bring someone from the street with you, you photograph the interior and ask them if the dimensions on the photograph look similar to the dimensions they see with their eyes.
The idea in this example is not about using or not using larger sensor or what field of view a larger sensor gives you, but making an image that represents the reality to anybody who looks at the image and the real location. And that's only for the example I gave with the interior where if you want to avoid wide-angle lenses you need to have a bigger canvas to draw onto, i.e. a bigger sensor. In other cases it may not matter what sensor you're using if you can easily do your job with what you currently have and you match the client's requirements.
The lens changes your perception for reality. Ask someone from the street if they think that photograph with a 10mm (distortion corrected) image is the same with what they see with heir own eyes. No way.
" if you want to avoid wide-angle lenses you need to have a bigger canvas to draw onto,"
No! That's not how it works! What you say doesn't make any sense!
You get the same perspective shooting at a given distance regardless of sensor size and REGARDLESS OF THE LENS you use.
Perspective is completely independent from both sensor size and lens use, and shooting a panorama with (let's say as an example) a 50mm to cover the angle of a 12mm lens won't give you a different perspective or a more "realistic" one, it will simply give you more resolution.
I've asked you a question which you didn't answer, what camera and lens combination do you think I've used to take that avenue shot?
Literally everything you're talking about is a property of field of view, not focal length.
The realism (or lack of) you're talking about is true of wide angle lenses, but has nothing to do with focal length. It is entirely about FOV and perspective distortion.
Well, that's what the article is about: the size of the canvas light is projected. In the example with the small interiors I've never said I'm changing focal length, but using "normal" focal length where a small sensor is not enough to capture everything in one shot. Glass is glass and it's an autonomous tool. Try to make a photo of a bookcase from a 30-degree angle with a wide lens (e.g. 16) and then make a panorama with a 50 from the same spot. The ratio between the height of objects closer to the camera and those further away is different on both lenses. This is what the lens do and this is how we call a lens "50mm" or "16mm." We're not using "50mm focal length," because this is an abstraction. We're using a tube with lenses that is calculated to have a 50mm focal length. The books say many things about what truly would happen to light and everything, but practice says it all.
This is why when you are on a movie and you have few extras to form a crowd you film them with a long lens, not with a wide-angle lens. This is why you frequently use a wide-angle lens when you are running in low mode with a camera stabilizer, because it makes it look fast because of the perception this particular lens creates. Imagine a DOP tells his assistant to give him a 24, because he likes the optical distortion it creates and the assistant tells him that this is not caused by the focal length, but by the field of view and distortion the optics create. That guy will probably be on the run for a different job.
In the article you literally talk about smaller sensors and wider lenses and how it's better to use a bigger sensor with a longer lens. If you said "larger sensors give you a wider field of view FOR A GIVEN PARTICULAR LENS" that would be true. But that's not what you said. You said we can't just use a wider lens with a smaller sensor - but that's not true, we can.
I literally work on movies for a living. Yes, if a DOP said that and his assistant said that back, that would be snarky and rude because the DP probably knows that already and we use shorthand when speaking. But you weren't using shorthand. You've argued both in the article and in your comments that there is some kind of difference between wide angle lenses on different formats that is somehow unrelated to the AOV.
And if that DP was using that 24mm lens on, say, a camera with a 2x crop, he would be severely disappointed with the lack of distortion. Many films shot on super 16mm film use a 12mm lens as their "normal lens" - because it looks like roughly a 35mm FF lens.
I see your point. The issue is with tests I've personally made and it's probably the lens' imperfections (although I corrected the distortion), but a panorama with a "normal" lens always gives me something that my eyes see. I agree that mathematically it can be proven it's the same, but experience shows that the lenses we are using (I'm not using cheap ones) show it the way I explained.
Paolo, it's easier to make better quality pixels if they're bigger. Think about a smart phone sensor vs anything else. Current sensor tech tempers that logic to an extent, but certainly not entirely. Noise isn't at all just about the aperture.
People are making far too many sweeping statements in these comments...
Never said there aren't advantages in using bigger sensor: more resolution (both because you can have more mpx and because the optics are working on lower MTF to resolve the same detail compared to smaller sensors), more dynamic range, better lenses choice, better lenses for low light (you just don't get f/0.6 or 0.7 on M43 equivalent to 1.2 & 1.4), less epxansive lenses for equivalent luminosity are all advantages, but claiming the perspective changes because you are using a longer focal lenght on a bigger sensor is just WRONG.
You guys are confusing and conflicting at the same time :-/ (referring to the blind test and then this). Anyhow, nice read, it's not like it makes any difference to me (still gonna buy gears)...
Not at all. Although sometimes fellow writers write an article with their the opposite point of view (because we're different and we all have opinions), in this case with Lee and this article of mine, there are no conflicts. In my type of work sensor size doesn't matter and I agree with Lee, but if I were photographing interiors or I find myself constantly up against the wall, I would buy a bigger sensor if I want to avoid wide-angle lenses. In this article I say that sensor doesn't really matter to me, but matters to others and I show examples why.
That's what my experience shows. Unless you photograph a wall standing right in front of it (when you won't see a noticeable difference), if you photograph a non-blank wall from the side the ratio between the height of objects closer to the camera and objects further away from the camera is different for wide-angle lens and one with a "normal" focal length. When I create a panorama with the "normal" lens the difference is obvious although the distortion is corrected on the wide-angle lens. That's what my experience and tests show. I've never tried that on a real cinema lens though (not a stills glas inside a cinema lens housing).
If "theory shows that there should be no optical difference," that doesn't mean the tools we have are perfectly aligned to the theory. And if they aren't this is what matters, not what paper and pencil say.