DSLRUsers.com

There's lots of talk about full frame (ie 35mm) sensors (eg like some high end Canon dSLRs) and the more standard sensor found in most digital SLRs. Usually you'll see people talk about a crop factor of 1.5 which effectively means the image taken by a "standard" dSLR sensor has a smaller field of view (FOV) than the same image taken using a full frame sensor. Thus a 70mm lens on a standard dSLR will have the apparent FOV of a 105mm lens on a full frame (150 = 70 * 1.5). Clear as mud. Let's put this a different way. To get the same apparent image size a standard sensor needs to be a lot further from the subject (since it has a smaller image frame) than a full frame dSLR sensor. See the pic below. We set up a Nikon D2x and a Canon 1DS MkII and were attempting to get approximately the same sized image in the camera. The picture shows where the cameras needed to be to achieve this. The D2X is on the left, the 1DS is in the middle (being used to take a shot) and the subject is at the right hand edge. In this case the standard sensor has to be 1.5 times further back in order to fit the same image in it's frame.

Very interesting Peter.
Thankfully you spared us the resultant shot of the back of Leigh's head from the D2X.

To me being further back to fit the same size image in the frame means you can fill more of the image in the frame if you shoot from the same distance as the full frame sensor. Is that an advantage where you may not be able to get as close as you want to the subject? I thought it would be under cetain circumstanses.

Craig

Yes it will have the same apparent FOV but it will not have the same resolution...
For most of us this will not matter as we are more than happy with our results..

CraigVTR,

Yes, that's correct. That's part of what the crop factor is all about.

Big V wrote:Yes it will have the same apparent FOV but it will not have the same resolution...

Well, let's see. In this case we're comparing a 16MP Canon FF sensor with a 12MP Nikon DX sensor.

Without doing any special math on this, I'd say that, if you're looking at not matching the FoV, there's very little difference in the actual resolution achieved.

For most of us this will not matter as we are more than happy with our results..

Yes, and especially so as in many cases the images don't ever get beyond being viewed on a monitor.

Logically, because the full-frame sensor is twice as big as the APS-sized one, we should be comparing the output of a Canon 20D to this 16 megapixel sensor.

If the D2x had a full frame sensor at 12mp, it's APS-cousin would be the D70/100. If it had double the 12mp at what would be the assumed double size anyway, a total of 24mp, it would be a logical conversion.

Actually Leigh, the area of a Nikon DX sensor is 1/(1.5*1.5) times that of a full-frame sensor, which equates to 1/2.25. So if the pixels were the same size on a full-frame sensor as on a 6Mp D70, there would be 13.5 Mp.

On an "APS-C" EOS the area is 1/(1.6*1.6) times that of a FF sensor: 1/2.56.
Thus a FF sensor with 20D-sized pixels would have 8.2*2.56 == 21 Mp.

Not that this discussion really gets us anywhere...

I have a couple of questions about this.

I am guessing that as the image is effectively cropped by the smaller sensor, the sensor must lie in the same plane as where the film would in an equivalent camera - the sensor is the same distance behind the back of the lens that was the case with film.

If this is the case, does that mean that light from the peripheral area of the glass in the lens actually focusses onto the area around the circumference of the sensor? ie the cropping that occurs is the light that enters through the periphery of the lens, lands outsides the sensor and is therefore redundant.

Does this make any sense? Do you think my reasoning is correct?

Mitchell wrote:If this is the case, does that mean that light from the peripheral area of the glass in the lens actually focusses onto the area around the circumference of the sensor? ie the cropping that occurs is the light that enters through the periphery of the lens, lands outsides the sensor and is therefore redundant.

Does this make any sense? Do you think my reasoning is correct?

Yes, exactly.

And it's why a DX lens, designed with a smaller image circle, won't work on a full frame camera.