Sensor Size

[ShaheedG]

Hello All,

I am venturing into world of DSLR and am on the verge of purchasing a new camera.

In my research, largely web based, it seems the important criteria for camera are (Very broadly):

• Purpose (What I plan to use it for)
• Budget (Camera and “accessories”)
• Technical Features (Too many to mention)
• How the camera feels in my hand
• Some sort of price value ratio.

Last bullet point not withstanding, most part the research can be done on the web. I though I had it down pat until I cam across an article in the NY Times. http://gadgetwise.blogs.nytimes.com/...sensors&st=cse

The article focuses (pardon the pun) on sensors and the importance size, but I cannot seem to find a sensor size comparison for cameras.

My questions to forum are then.
• Is the sensor size important?
• Where in the comparison tables do I find the sensor size comparison?
• Are there trade-offs involved when comparing sensor size?

Sorry for long post, I just wanted to make my questions clear.

[mjs1973]

Welcome to the forum. Here is a link to a page that shows the different sensor sizes.

To answer your question, 'Is sensor size important?' I would say yes it is BUT it depends on the shooter, and their intentions. For me, I was planning on moving to a full frame sensor, but then I took a good long look at my shooting style, and decided that the APS-C size sensor was a better fit for me, and my style of shooting. It really depends on what you are planning on using your camera for. I like to get in close to wildlife, and the narrower field of view I get from my smaller sensor helps me do that.

One area where I think sensor size is very important is image quality. The more pixels you pack into a sensor, the smaller they have to become to fit. For example, if you put 10mpxls on a full frame sensor, and 10mpxls on a much smaller 4:3 sensor, you can make the pixels on the full frame sensor much larger. The larger the sensor, the better they are at capturing light. The result is cleaner images at higher ISO's. That's not to say the smaller sensors are bad though.

Something else to consider is the size of the camera. As you stated, how the camera feels in your hand is a very important thing to consider. A camera with a larger sensor, will require a larger body. A camera with a smaller sensor can result in a smaller, lighter camera.

Also, smaller sensor are cheaper to produce, and that will play a big role into the price of the camera.

So when you look at the big picture, I don't know that sensor size by itself is a huge issue, but it does play a role in just about every aspect of choosing a camera that you mentioned in your original post. Cost, how it feels in your hand (size) image quality, purpose and price.

[Anbesol]

• Where in the comparison tables do I find the sensor size comparison?

There are essentially 4 different sizes. APS-C being the most common, and being 1.5 times smaller than a full frame 35mm sensor (with Canon's APS-C, its 1.6). There is also APS-H, a rare size - only one I can think of is the Canon EOS 1D Mark II N, its 1.3 times smaller than a 35mm full frame. Theres also the 4:3 sensor, which is a bit different since its aspect ratio changes as well as its size - but it is 2 times smaller than a 35mm sensor. Then of course, the 35mm sensor is 24mm x 36mm.

We'd all love to have full frame 35mm sensors in our camera, but that requires a much more costy investment, you'll need $3k for body only, then you'll need to spend more on lens that can match your sensor. Lens perform much better on APS sensors, because the refracted image onto an APS sensor is brought through the lens sweet spot, so mid-range lens act more like high range lens on APS sensors. Lens problems can often be irrelevant on an APS sensor (vignetting, corner resolution, etc), while they become serious problems on a full frame sensor. Expect to spend a lot more on lens for full frame than for APS.

I digress, sensor size does make a difference, but all sensor sizes have manufacturers that do really well with them.

[GB1]

ShaheedG,

Welcome. Mike did a good job explaining here, but let me also throw in a few tidbits.

Previous 35mm film cameras used a 3/2 ratio of 36mm x 24mm film. This was effectively the 'sensor size' of 35mm cameras, as the film was the sensor. When they started replacing the film with light sensors several years back, instead of using a 36 mm x 24mm sensor they went with something smaller - why, I am not sure, but it may have had something to do with cost.

What the smaller sensors effectively did was not use the full area captured by the lens, instead only using the center area. This effectively has a zooming effect.

What does this mean to photographers?

- If you like wide angle lenses and photos, you are unfortunately losing a lot. These small sensor cameras come with lens factor values to let you know the 'zoom' effect (I forget the official term). Canon's sensors are 1.5 and Nikon's about 1.6. This means that with Canon, if you had a 24mm lens, it is now the equivalent of a 36mm.

- If you are a wildlife or sports photographer, you may love the zoom effect. You can now reach out a lot farther w/o having to buy very expensive lenses!

- The center of the image is always superior in quality than the sides. Some bad lenses have a lot of light dropoff in the corners, for example, and it is usually sharper in the middle. Since the smaller sensors only use the middle area with normal lenses (more on that below), you are effectively getting the 'sweet spot' all the time - a plus.

- Manufacturers are now producing lenses that do not even produce a light area that exceeds the smaller sensors' area. This is cheaper for them. What does it mean? Not much if you use the smaller sensor cameras. But if you try using these new lenses on an older 35mm camera or on a full-frame sensor camera, you will not get the entire viewfinder area in your pix (!!!). So these lenses are only for C sensor cameras.

That's about it. I have a Nikon D200 with smaller sensor, and personally I don't like losing my wide angles But I do a lot of landscape work. Thinking about getting a Nikon D700 FF camera these days........

G

[ShaheedG]

Thanks a lot guys. I surfed various forums and after a few days selected this one to join and post my question on. True to form ,you guys have answered my question from all angles.

Regards

[Franglais]

I know I'm late in posting to this thread but I lost track of it. I wanted to reply because nobody seems to have expressed much enthusiasm for the APS-C sensor size.

Having to cover a smaller image circle (than 24x36) allows the lens designer to do more daring things. Lenses like Canon and Nikon's 18-200 offer small size, good image quality, relatively low cost and enormous zoom range - there is simply no equivalent in lenses covering 24x36 format

You're missing a really wide-angle lens? There are a number of very wide wide angles at about about $500. My Tamron 11-18 is very light and the quality is enough for my needs.

A 24x36 sensor brings you about an extra stop of usable speed i.e.12800 ISO instead of 6400 ISO for a D700/D300. Not the sort of thing you need every day. For me the most important advantage is that the highlights on a larger sensor burn out more gracefully than on a smaller one, but between APS-C and 24x36 the difference is not very great.

BTW Nikon has a "correction factor" of 1.5 and Canon has a "correction factor" of 1.6 and not the other way round. The Nikon sensor is slightly larger than the Canon one.

[GB1]

Looks like you're right Charles, at least for Nikon:

http://www.nikonians.org/nikon/nikko...mm/review.html

[Sushigaijin]

Something not addressed is that sensor size effects depth of field. The higher the "crop" or "correction" factor, the more depth of field you get for a given F/stop. so f/2.8 on a full-frame sensor has less depth of field than f/2.8 on an aps-c or 4/3 sensor, always (everything else being equal). If your applications need high depth of field, like in macro photography or long telephoto, a crop sensor might be preferable to a larger sensor.

So while Anbesol generalized that "We'd all love to have full frame 35mm sensors in our camera," it's simply not always true. Personally I'd loathe to give up the extra depth of field and correction factor of the 4/3 format. For me, those extras outweigh the cost of a slightly "noisier" image. What Anbesol really intended to say is that we'd all like our cameras sensor to produce images as high-quality as full frame 35mm sensors, and he'd be right about that!

[Franglais]

Something not addressed is that sensor size effects depth of field. The higher the "crop" or "correction" factor, the more depth of field you get for a given F/stop. so f/2.8 on a full-frame sensor has less depth of field than f/2.8 on an aps-c or 4/3 sensor, always (everything else being equal). If your applications need high depth of field, like in macro photography or long telephoto, a crop sensor might be preferable to a larger sensor.

Take a look at the next thread in this forum and folks have concluded the opposite - for a given focal length and aperture, the depth-of-field is exactly the same whatever the size of the sensor. It's in the optical properties of the lens.

However even that is not quite true. Depth-of-field is the difference between "in-focus" and "out-of-focus" which is a progressive thing. The more you enlarge an image the earlier you notice that it looks out-of-focus. As you enlarge an APS-C image more than a 24x36 one it looks like it has less depth-of-field.

The other factor is in-camera sharpening. The camera will make an area that is slightly out-of-focus sharper so it looks in-focus. The cut-off between in-focus and out-of-focus becomes more abrupt and less natural. Generally the larger the sensor the less sharpening the camera does.

[Anbesol]

Something not addressed is that sensor size effects depth of field. The higher the "crop" or "correction" factor, the more depth of field you get for a given F/stop. so f/2.8 on a full-frame sensor has less depth of field than f/2.8 on an aps-c or 4/3 sensor, always (everything else being equal). If your applications need high depth of field, like in macro photography or long telephoto, a crop sensor might be preferable to a larger sensor.

So while Anbesol generalized that "We'd all love to have full frame 35mm sensors in our camera," it's simply not always true. Personally I'd loathe to give up the extra depth of field and correction factor of the 4/3 format. For me, those extras outweigh the cost of a slightly "noisier" image. What Anbesol really intended to say is that we'd all like our cameras sensor to produce images as high-quality as full frame 35mm sensors, and he'd be right about that!

Actually 'APS lens' is somewhat of a misnomer - it simply translates to smaller cuts of optics (or short back focus with canon), throw it on a full frame and you'll vignette around the edges at wide and corner problems are worsened. Essentially you cut out sweet spot in favor of smaller form factor/lighter lens with "APS designed lens". But f/2.8 dof is the same on both APS and full frame. The difference is crop factor, not optical construction, or maybe thats what you meant and I misread.

no-less, when I said 'we'd all love to have a full frame camera', i was speaking in a mystical, pipe-dream sort of way. ;)

[Sushigaijin]

The smaller the sensor the more DoF for a given aperture (at output level).

That's a factor of sensor size. The only other things that effect DoF are magnification and aperture. In the case of very wide lenses, focal distance can factor in as well - it's important to think of focal distance as different than magnification IF the lens is approaching hyperfocal distance.

F/2.8 has the same DoF for any focal length at the same magnification on the same sensor, but that changes drastically when you move to a smaller sensor. A 4/3 DSLR has roughly two stops more DoF at any given aperture (at the same magnification) than a 35mm full frame camera.

Of course, that's in relation to each other. Absolute depth of field on the same sensor is irrelevant because a 4/3 lens would never cover the sensor of a 35mm camera. This post also explains one reason why it is very hard to get short DoF on the tiny sensors in compact cameras.

[OldClicker]

Is the 'APS' term a carry over from the Advanced Photo System 24mm film cartridges? - TF

[Dougjgreen]

Is the 'APS' term a carry over from the Advanced Photo System 24mm film cartridges? - TF

Yes it is.

[Franglais]

The smaller the sensor the more DoF for a given aperture (at output level).

That's a factor of sensor size. The only other things that effect DoF are magnification and aperture. In the case of very wide lenses, focal distance can factor in as well - it's important to think of focal distance as different than magnification IF the lens is approaching hyperfocal distance.
...

OK I see what you mean for the first statement. If you take the exact same picture at the same aperture on a camera with a big sensor and a camera with a small sensor the image from the small sensor will have more depth of field because the focal length of the lens is shorter.

I don't understand the second statement. What is "focal distance"? How does the hyperfocal distance come into it?

[Sushigaijin]

OK I see what you mean for the first statement. If you take the exact same picture at the same aperture on a camera with a big sensor and a camera with a small sensor the image from the small sensor will have more depth of field because the focal length of the lens is shorter.

I don't understand the second statement. What is "focal distance"? How does the hyperfocal distance come into it?

Focal distance in the distance from the sensor to the sharpest point of focus.

Lenses all behave similarly at the same F/stop except when nearing hyperfocal distance. So a 400mm at f/4 has the same depth of field as a 50mm at f/4, assuming the magnification is the same.

or,

As long as the magnification level doesn't bring the focal distance close to the hyperfocal distance, f/4's depth of field is always the same.

Since a 100mm and a 400mm both have very far away hyperfocal distances, they will have the same (virtually) depth of field at any given f/stop at the same subject magnification. Make a quarter the exact same size at the exact same place in focus, and the depth of field will be the same at the same aperture on the same sensor, even with lenses of different focal lengths.

Well wait, that doesn't sound right - a 24mm lens has WAY more depth of field than a 400mm at f/4.

Of course. But at the same magnification level, the 24mm is approaching or exceeding the hyperfocal distance. At f/6.3 and 24mm the hyperfocal distance is INCHES from the lens...too close for a 400mm to even focus. But if you're shooting at say, 1:3 magnification with a 24mm and a 28mm, and the object is at the same magnification, f/4 will be very similar in depth of field. You could never tell the difference without actually crunching the equation on a calculator.

The nitty gritty is that in MOST situations, focal length doesn't have anything to do with depth of field. Only magnification, focal distance, sensor size and aperture control depth of field. It's an important consideration when stepping up to a smaller sensor because you've changed one of the basic determinants permanently. I can never have the same shallow DoF as f/1.4 on a full frame - I would need an f/0.2 lens (or something like that). I do, however, get twice as much depth of field at f/2 than a full frame does.

I shoot a lot of macros, so that's a good thing.

[Loupey]

...It's an important consideration when stepping up to a smaller sensor because you've changed one of the basic determinants permanently. I can never have the same shallow DoF as f/1.4 on a full frame - I would need an f/0.2 lens (or something like that). I do, however, get twice as much depth of field at f/2 than a full frame does.

I shoot a lot of macros, so that's a good thing.

Any format smaller than full-frame (1.6x, 1.5, 2x, even PS) can still be fully replicated by a full-frame sensor as long as one knows how much cropping it will take - that is, how much smaller to make the subject magnification at the imaging plane.

With today's mega MP full-frame cameras, I bet that the level of cropping necessary to achieve the 1.6x and 1.5x (and possibly even the 2x) would still keep it on par, image quality wise, with those dedicated formats.

My "old" 5D at 12mp still produces cleaner images than my "new" 50D at 15mp. Of course the number of pixels isn't the deciding factor here.

So the real disadvantages of going as large as you can (i.e. full-frame) are the size of the equipment and price.

[OldClicker]

Any format smaller than full-frame (1.6x, 1.5, 2x, even PS) can still be fully replicated by a full-frame sensor as long as one knows how much cropping it will take - that is, how much smaller to make the subject magnification at the imaging plane.

With today's mega MP full-frame cameras, I bet that the level of cropping necessary to achieve the 1.6x and 1.5x (and possibly even the 2x) would still keep it on par, image quality wise, with those dedicated formats.

My "old" 5D at 12mp still produces cleaner images than my "new" 50D at 15mp. Of course the number of pixels isn't the deciding factor here.

So the real disadvantages of going as large as you can (i.e. full-frame) are the size of the equipment and price.

I have an xx mm prime on my 1.5x camera and stand at 10 feet and my subject fills the frame. So with the same lens at 10 feet with FF camera, if I crop to again fill the frame, I will have the same DoF? - TF

[Loupey]

I have an xx mm prime on my 1.5x camera and stand at 10 feet and my subject fills the frame. So with the same lens at 10 feet with FF camera, if I crop to again fill the frame, I will have the same DoF? - TF

Yes - the same lens shot at the same distance will produce the same DOF. How can it not? The lens doesn't care what it's attached to.

[Loupey]

OK, look - here is a two minute example I just shot with what I have at the moment at my office.

To exaggerate the DOF effect, I shot both of these at the maximum zoom (105mm) at the maximum aperture (f/4) at the closest focusing distance (the front of the lens is at the zero mark). Same ISOs, same lens.

First set - uncropped from both cameras.

Second set - cropped to the same "apparent" subject size.


EXIF data attached if you want to check.

[Anbesol]

Sushi, I see what you're saying about magnification affecting DoF, but that still would be a difference of field of view, rather than of the diameter of the optical construction or aperture blades. For example, with canon, an EF and EF-S lens both with the same range, paired to a 50D will both produce the same DoF, in spite of the fact that the EF-S was specifically designed for APS. When paired with a full frame, however, the EF will produce less DoF in respect to the field of view, but still produce the same DoF at the exact mm range of the lens (with full field of view).

get twice as much depth of field at f/2 than a full frame does

Exactly, on a 2x crop sensor.

If the diameter of the optics affected DoF differently, you'd get wider DoF with EF-S than EF - but the point of the EF-S is for its super wide angles, and for its smaller and lighter construction.

[Loupey]

And so the reason why P&S cameras have these deep DOFs - even with macros - is because they use these ridiculous focal lengths (7mm ~ 21mm) matched to their tiny imaging sensors.

So, again, anyone can reproduce those effects if one really wanted to go around shooting the world with a fisheye or ultra wide angle lens all the time and then crop out a tiny portion of that image.

[Sushigaijin]

Sushi, I see what you're saying about magnification affecting DoF, but that still would be a difference of field of view, rather than of the diameter of the optical construction or aperture blades. For example, with canon, an EF and EF-S lens both with the same range, paired to a 50D will both produce the same DoF, in spite of the fact that the EF-S was specifically designed for APS. When paired with a full frame, however, the EF will produce less DoF in respect to the field of view, but still produce the same DoF at the exact mm range of the lens (with full field of view).


Exactly, on a 2x crop sensor.

If the diameter of the optics affected DoF differently, you'd get wider DoF with EF-S than EF - but the point of the EF-S is for its super wide angles, and for its smaller and lighter construction.

I don't think I mentioned aperture blades or optic diameter, so I'm not totally sure I understand your statement.

But the second part I totally agree with - with the same lens, I get twice as much DoF on a 2x "crop" sensor than on a full frame 35mm sensor.

The flip side is that I can NEVER get the same DoF as a super fast full frame camera - the 4/3 system, and APS systems, cannot have the same short DoF as a full frame at f/1.4, because they are all smaller sensors. So if you need MORE depth of field, a crop sensor might be a better option than a full frame, is all I'm sayin'.

Or phrased differently, the very limited depth of field that you get with a full frame sensor might be a negative to some people, where the greater depth of field with a crop sensor might be a positive. Or the opposite.

Now take Loupey's shots and don't crop them - make the magnification the same in-camera, and the results will be very different. The 50d will have a lot more depth of field at the same aperture because of the smaller sensor.

[Loupey]

Now take Loupey's shots and don't crop them - make the magnification the same in-camera, and the results will be very different. The 50d will have a lot more depth of field at the same aperture because of the smaller sensor.

I agee with that statement. There are only three ways to maintain the "apparent" magnification with a full-frame camera: 1) get closer, 2) use a longer focal length lens, or 3) a combination of getting closer and longer lens. (OK, there is a 4th way but it really doesn't apply here).

But doing any one of these three changes of the rules of the game. Each one of the three reduces DOF and so it is not the size the sensor itself that causes the shallower DOF, rather it is the compensating technique employed.

And in getting closer, the perspective changes and the entire comparison between sensor sizes falls apart.

[Loupey]

Another way to think about this is with macro terminology.

Macro is always expressed as subject magnification at the film/imaging plane. It is the same regardless of the format used. A 1:1 macro lens produces life-size images whether it is with a 1.6x crop camera, full-frame, or medium format.

So my 180mm macro, at full macro, will produce the same life size image with the same DOF and the same perspective on both my 5D and 50D. It only looks larger/closer on the 50D because we're comparing the smaller uncropped image to a larger uncropped image of the 5D. Because the 5D image "shows more", the angle of view is larger. But I can assure you that the lens did its job and produced the same image circle and the same magnification with the same DOF in both cases.

[Loupey]

One last thought experiment from me This time, instead of keeping the focal length of the lens fixed, change it but fix the shooting distance.

In this case, think of scenario in which your camera-to-subject distance is fixed - like shooting a football game from the sidelines.

You are there with a 50mm f/1.8 and the person next to you has a 500mm. You both shoot the same scenes from the same distances using identical exposure settings. If you were to crop your 50mm images at 10x they would look just like those shot with the 500mm - the same spacial relationship between the players and the background. The only thing that would be different (aside from the obvious degradation in IQ) is the DOF. The cropped image would look as one would expect from a cropped image.

So a format that touted a 10x crop factor (in which a 50mm becomes effectively a 500mm) would produce this apparently highly magnified image with a deep DOF. But definitely not the same image as the SI guy with his 1Ds because of the DOF only.