How does crop factor work?

[Joetsu]

I intend to buy a m43 camera and have been reading up on how these systems double the crop factor vis-a-vis full frame cameras. I UNDERSTAND what is going on. What I can't understand is WHY the image in a m43 system is so much larger than in a full frame system. To my way of thinking, the full frame photo should simply be cropped and nothing else changed. What makes the image larger? What is the mechanism at work here?

[Colin Southern]

A couple of factors come into play here ...

1. Yes - you could just crop an image from a full-frame camera ... but you then have to enlarge what's left so that the image appears to be the same size on your monitor or in a physical print so that it APPEARS to be "more zoomed in".

2. If you crop an image from a full-frame camera you're discarding the information from a lot of pixels, whereas with crop-factor cameras, all of the pixels are contained in a smaller area -- so a crop-factor camera may have less pixels than a given full-frame camera but it may well have more EFFECTIVE pixels once the crop has taken place. So although a crop-factor camera may give you a field of view equivalent to a lens 1.5 (Nikon) or 1.6 (Canon) times the focal length of the same lens on a full frame camera, the effective amount of information captured depends on the pixel density of the sensor; sometimes a cropped image from a full-frame sensor will contain more information than a cropped camera (eg a 1Ds3 compared to a 1D3) and sometimes it'll contain less (eg a 1Ds3 compared to a 60D).

Hope this helps

[Joetsu]

A couple of factors come into play here ...

1. Yes - you could just crop an image from a full-frame camera ... but you then have to enlarge what's left so that the image appears to be the same size on your monitor or in a physical print so that it APPEARS to be "more zoomed in".

2. If you crop an image from a full-frame camera you're discarding the information from a lot of pixels, whereas with crop-factor cameras, all of the pixels are contained in a smaller area -- so a crop-factor camera may have less pixels than a given full-frame camera but it may well have more EFFECTIVE pixels once the crop has taken place. So although a crop-factor camera may give you a field of view equivalent to a lens 1.5 (Nikon) or 1.6 (Canon) times the focal length of the same lens on a full frame camera, the effective amount of information captured depends on the pixel density of the sensor; sometimes a cropped image from a full-frame sensor will contain more information than a cropped camera (eg a 1Ds3 compared to a 1D3) and sometimes it'll contain less (eg a 1Ds3 compared to a 60D).

Hope this helps

Regarding 1: I thought the "enlargement" was going in inside the camera before anything is printed or displayed externally. If this is the case, then this doesn't help.
Regarding 2: I still don't get why "more effective pixels" should lead to the zoom effect. I would have thought they would just make for more effective image that has been cropped, but is otherwise unchanged.

[Manfred M]

Regarding 1: I thought the "enlargement" was going in inside the camera before anything is printed or displayed externally. If this is the case, then this doesn't help.
Regarding 2: I still don't get why "more effective pixels" should lead to the zoom effect. I would have thought they would just make for more effective image that has been cropped, but is otherwise unchanged.

1, Actually there is no enlargement at all in camera, a given length focal length lens will always produce the same size image on the sensor (ignoring the image circle issue). A sensor simply records the light falling on it. What is happening in a crop sensor camera like the MFT, the sensor is smaller, so when you look at an image on the same sized screen as a full-frame or APS-C image, it just looks larger.

2, Effective pixels have nothing to do with the zoom effect, but they do impact image quality. If you have a 12MP MFT sensor that is 18 mm × 13.5 mm = 243 sq mm and a 12MP full frame sensor that is 24mm x 36mm = 864 sq mm. This means that the full-frame sensor has 3.56 greater surface area than the MFT one. If we were to crop the full-frame image down to MFT size, then 12 / 3.56 = 3.37 MP effective. So while the images would be the same size, a lot of data would have been thrown away on the full-frame and you would be looking at a full 12MP image from the MFT, but only a 3.37MP from the full frame. Guess which one would give you better image quality?

[Joetsu]

1, Actually there is no enlargement at all in camera, a given length focal length lens will always produce the same size image on the sensor (ignoring the image circle issue). A sensor simply records the light falling on it. What is happening in a crop sensor camera like the MFT, the sensor is smaller, so when you look at an image on the same sized screen as a full-frame or APS-C image, it just looks larger.

So let me see if I understand. By "screen" do you mean the 3-inch LCD display on the MFT camera? If so, here's why I still don't understand. If I cropped (cut away) the outer rectangle on a FF image, I would be left with the inner portion of the image without any altering of the content's size, right? Given that the sensor on the MFT is the size of this inner image of the FF, why does the content appear to be enlarged? What is the mechanism that is blowing it up to fit the 3-inch display?

I appreciate everyone's efforts to educate me. Thanks!

[Manfred M]

So let me see if I understand. By "screen" do you mean the 3-inch LCD display on the MFT camera? If so, here's why I still don't understand. If I cropped (cut away) the outer rectangle on a FF image, I would be left with the inner portion of the image without any altering of the content's size, right? Given that the sensor on the MFT is the size of this inner image of the FF, why does the content appear to be enlarged? What is the mechanism that is blowing it up to fit the 3-inch display?

I appreciate everyone's efforts to educate me. Thanks!

No I am referring to a computer screen where you view the image or even a hardcopy if you produce prints. The playback on your camera is just going to show you the image produced on your specific camera model. You need a piece of hardware (or piece of paper) that one would normally use to view the images downloaded from the camera's memory card. As an example, I have a number of cameras ranging from a very small sensor point & shoot, a cross-over camera, a crop frame camera and a full-frame camera. Even though each of these camera has a different sensor size, they look the same size when I display them on my computer screen.

The screen on your camera is keyed to reproduce images shot on your camera to a useable size for review. It's really not a particularly good tool to use for looking at your images; it is far too small. Yes, your camera screen can give you an enlargement of part of your image, if you want to zoom in, but this is not what I am referring to.

[pnodrog]

If you can go into a camera shop and get a Nikon D800 or any FF camera with crop modes built in demonstrated to you and have a good play with it. It should all become perfectly clear. The fitting of any type of image to suit a particular size screen is a result of software and/or electronics not photography. Actually if you visit a TV sales shop you will see walls of the same image at different sizes.

[Joetsu]

So you're saying it's the PC software that takes the cropped image and enlarges it--i.e. gives it the zoom effect?

[Colin Southern]

So you're saying it's the PC software that takes the cropped image and enlarges it--i.e. gives it the zoom effect?

Yes.

More specifically, it just displays it at a size that fills your monitor; in most cases only 1 in every 3 or 4 pixels is displayed though because most full-resolution images contain many more pixels than can be displayed at once on a monitor.

[pnodrog]

The image is in a file that contains all the information. This information is simply used by the software to display it on your monitor. As Colin points out you will probably not use all of the pixels from the image file to display the image on your monitor. If you have a 2mega pixel screen and an 8mega pixel file the images could even be considered to be reduced not enlarged. All 1920x1080 monitors have the same number of pixels but they are simply larger on a bigger monitor.

If you are using the image file for printing there are a whole lot of other factors to consider. I suggest you do not think of the different sizes as having anything to do with a zoom effect.

[Joetsu]

Yes.

More specifically, it just displays it at a size that fills your monitor; in most cases only 1 in every 3 or 4 pixels is displayed though because most full-resolution images contain many more pixels than can be displayed at once on a monitor.

Got it. Thanks very much.

[Mike Buckley]

To follow up on the previous three posts, you can display a 1920 x 1080 image on your computer monitor and you can display it on a large television. The file will be displayed much larger on your television because the software built into the television is configured to display it at the size of your television.

[Hazeb1]

Aloha All,

Just so this is all clear to me, when people refer to the "extra reach" of a crop sensor camera, there really is no actual optical effect (bonus), it's only related to how the image is displayed. (ie 200mm is still 200mm ff or aps-c)

Am I correct?

[Colin Southern]

Aloha All,

Just so this is all clear to me, when people refer to the "extra reach" of a crop sensor camera, there really is no actual optical effect (bonus), it's only related to how the image is displayed. (ie 200mm is still 200mm ff or aps-c)

Am I correct?

Hi Warren,

Correct.

What's most important though is the amount of information contained in the image; if you have a 18MP full frame camera and a 18MP crop factor camera then ALL of the sensor pixels will be sitting ONLY under the cropped portion of the image circle on the crop-factor camera ...

... so that area will contain a LOT more information than will the same (cropped) area of the sensor of the full-frame camera.

You'll hear that crop-factor cameras essentially give you a 1.6x (or 1.5x for Nikon) focal length multiplication. In practice is true for all intents and purposes, but only in terms of field of view. In terms of actual information captured, it depends on the pixel density of 2 respective sensors.

eg If I had my Canon 1Ds3 next to a colleague with a Canon 1D3 (1.3 crop factor) then things may SEEM 1/3 or so closer for him, but if you do the maths, if I crop my 21MP sensor to the 1D3 sensor size then I'm left with about 13MP to his 10MP - so even though he has a crop-factor camera, I'll have more detail in the cropped portion of the image.

On the other hand, if I shoot with a 12MP 5D and he shoots with a 10MP 1D3, he will have more detail (or a psudo focal length increase) because I'd be left with far less pixels in the cropped area.

Hope that makes sense!

[Manfred M]

Colin is absolutely bang on with his explanation. With photography, as with pretty well anything thing else, what you get are a set of tradeoffs the impact the final product. A smaller sensor tends to have upsides (lower cost for the camera and smaller, more compact design) and downsides (lower overall performance than the larger sensor cameras)

While in general, this provides some gains if you are shooting a telephoto lens, you tend to lose out if you are shooting wide angle shots. You will find that getting a fast ultra-wide angle lens for your MFT is going to be challanging. While I can get an f/2.8 14-24mm lens for my full frame camera, the best you can do is a moderately fast f/4 7-14mm lens.

The next issue is image quality. WIth your crop factor of 2, you will have to enlarge the image twice as much as you would on a full-frame camera, and any image defects, from edge softness in the lens to image noise, would be much more apparent in a crop factor camera.

The third downside is image noise. To pack more pixels into a smaller sensor, the size of individual photoreceptors is smaller, This mean that they are less sensitive to the light hitting them, versus a larger photoreceptor. This means the sensor has to apply more gain to the signal, i.e. amplify the signal, and this results in more digital noise.

If you go to a site like http://www.dxomark.com/ , you will find that colour depth, dynamic range and low light performance tend to be significantly lower on smaller sensor cameras than ones with larger sensors.

.

[Hazeb1]

Thanks guys!!! I actually feel smarter now

This is why I love this forum!!!

[Colin Southern]

Thanks guys!!! I actually feel smarter now

This is why I love this forum!!!

Glad we could help Warren