Crop Factor - A Different Approach

[boBQuincy]

I have noticed some confusion with regards to what the crop factor does to a lens, and have a new approach on how to relate to crop factors, one that may be more intuitive than what we often read.

With a wide angle lens the primary concern is field of view. Here it is convenient to redefine the focal length in 35mm/fullframe terms since that is a good way to imply what the field of view would be. If we mount a 24mm lens on a 1.6x camera we can expect it will give approximately the field of view of a 38mm. If we are looking for a really wide angle we know we will have to spring for something under 18mm.

Where this equivalency doesn't work as well is at the long end. Here field of view is not much of an issue, it is already small and 1.6x less may not be important. In many cases the lens is not as long as we would like anyway and our subject only covers a small area of the viewfinder, cropped or not. What can really be considered misinformation is equating a 200mm to a 320mm. The 200 on a crop camera only gives the field of view of a 320, but with a long lens what is really important is the magnifcation, and the crop factor does not change the magnification at all. An image of the moon projected on the sensor will be the same size with a given lens, no matter what camera it is mounted on.

So the first point is: the *only* thing that changes when moving a lens from a full frame camera to a crop camera is the field of view. The second point is that field of view is often not the most important issue when using a long lens. Magnification is generally the main reason for using a long lens and magnfication does not change with a crop camera.

To reduce confusion I prefer to consider lenses as whatever the focal length actually is, and just realize that 24mm is not very wide on a 1.6x camera.

Right or wrong, I expect to hear more about this, and that's what makes this board so good!

 

[skr8pn]

I 100% agree with you. One reason is we've had this conversation before over alchoholic beverages. But as someone else had also told me the focal length does not change at all, it's the same distance to film as it is to the sensor. It is indeed it's field of view that changes.

 

[photo_chick]

This is what I saw when I put my Rebel G next to my Rebel XT with the same shot and the same lens. Magnification seemed the same, it was what I was actually able to get in the shot that was different. I saw the most noticeable difference in close shots.

I did this when I first got my XT before I had any clue what crop factor was. I just wanted to compare the cameras and get a feel for how the XT was different. Though my comparison was far from scientific, these were my observations.

 

[handicap18]

It certainly can be a confusing subject, especially to those who are new to the digital SLR concept.

Focal length can be confusing. Its measured from the sensor/film in the body to the center of the lens. Thus when you zoom out the lens gets longer.

I think 0bli0 had the best example a few months back when he posted the picture and had the different color lines to show the field of view from a film SLR and a dSLR (I believe he used the 1.6x crop factor). The center of the picture did not change at all, however you do see less of the picture at the edges (for lack of a better term).

I don't think there is really any other way to describe an 18mm or a 24mm focal length on a dSLR then to say that is the equilivent of 27mm or 36mm (based on 1.5x crop) (just using those 2 #'s for now as we all know there are a lot more).

I think that no matter how we describe it, to someone new to the dSLR world it is going to be confusing at first and will take some time to get an understanding of it. I know I was very confused about it all at first, especially when they started making the lenses designed for digital SLR's. I see it all the time over on the nikonian's board from people who just bought or are about to buy a new dSLR.

 

[Michele]

It certainly can be a confusing subject, especially to those who are new to the digital SLR concept.



I think that no matter how we describe it, to someone new to the dSLR world it is going to be confusing at first and will take some time to get an understanding of it.

Hello, that is me!! Totally confused. Thanks for trying to clear it up, one of these days it will click.

For now, I have one quick question, how do I find out what crop factor (I think that is the right term) my camera has? Will I find it in the manual? Oh, do you mean I should read the manual?

 

[ukcatfan]

Hello, that is me!! Totally confused. Thanks for trying to clear it up, one of these days it will click.

For now, I have one quick question, how do I find out what crop factor (I think that is the right term) my camera has? Will I find it in the manual? Oh, do you mean I should read the manual?

Check this page: http://www.steves-digicams.com/cameras_digpro.html
He calls it a multiplier.

Kevin

 

[boBQuincy]

Check this page: http://www.steves-digicams.com/cameras_digpro.html
He calls it a multiplier.

Kevin

It's disappointing to see that from Steve, who has a really good website full of information.
The term "multiplier" is one of the ideas that I believe is causing much of the confusion. It may apply ok to a P&S with it's tiny sensor where a "normal" lens is 10 mm.
In a dSLR where the sensor is much closer to full frame size, the term "divider" is more correct in that we are getting less of an image than a full frame camera would.

Multiplier is what causes people to think their 200mm lens is now magically a 320mm, when in reality for everything except field of view it acts just like the 200 that it is.

 

[Furgus]

Geez boB, I thought you would have linked to the Luminous Landscape Definition of DSLR Magnification. Here let me do it for you


http://www.luminous-landscape.com/tutorials/understanding-series/dslr-mag.shtml

I hope this explains it the way you meant to explain it. LL has always been pretty good info, if not, just smack me tomorrow when you see me

 

[Master Mason]

Here was ObliO's example I agree, best explination I have heard

yellow lines full frame Pink lines 1.6 crop

 

[0bli0]

i have seen that image pop up a few times...

the reason you take the crop factor into account for the 1 over shutter speed rule, is due to the percentage of motion in relation to the size of the sensor (or film). in Bob's moon example, although the image of the moon hitting the sensor is not multiplied (compared to a 320 lens on a full frame body), the amount of motion as a fraction of the size of sensor (and image) will remain the same. what does not remain the same is a lens' ability for image resolution (i don't mean megapixels). as Bob correctly pointed out, the image of the moon is the same size on the sensor, so the sharpness of a full frame at 320 will be the same as the sharpness of a 1.6x crop at 200. when you take the crop into account, the fine detail is magnified in relation to the crop's ratio (assuming you are comparing images of the same size such as 4"x6"). in otherwords less fine.

and of course there is the whole effect of Circle of Confusion on the bokeh (long story short- the out of focus areas on a bigger sensor look better than on a smaller sensor).

 

[boBQuincy]

Here was ObliO's example I agree, best explination I have heard

yellow lines full frame Pink lines 1.6 crop

This is about the best example of crop factor that I have seen too.
But...
In some ways it doesn't address what I was thinking about long lenses, where the subject is only a small portion of the frame. This is where I see the normal explanation of crop factor falling apart. This is also where many people think they are getting a free lunch, a 320mm out of a 200mm. Are they?
Does crop factor math get even worse? Oh yes, it does!


Say we take a photo of the moon again, with our same 200mm lens. On a full frame it is maybe 1/4 of the vertical frame, on a 1.6 crop it is .4 of the frame, a much larger portion. Same number of pixels though, so it is the same size moon image.

Unless...
the 1.6 crop camera has a greater pixel density, which many do. If the 1.6x crop camera has a 1.6x greater pixel density then we have a moon image that is 1.6x as many linear pixels as the full frame image and we *do* have a 320mm equivalent lens, at least in magnification.

Afaik no 1.6 dSLR has that kind of pixel density (P&S do) but that brings up another issue, which full frame are we comparing it to? 5D? 1Ds? 1DsMkII? They all have different pixel densities.

So am I just trying to cause trouble? Nah, just getting people thinking.
Of course this is pretty much a theoretical example, but it may raise some questions, at least for those of us who always need a longer lens and end up cropping our telephoto images.

 

[photo_chick]

STOP MAKING ME THINK!!! AGHHHHHHHHHH!!!

The little hamster is gonna have a heart attack if I keep using my brain so much!

 

[MarkBarbieri]

I have to admit that I don't see any difference between cropping a wide lens or cropping a long lens. A crop is a crop to me.

My concern about focusing strictly on the crop is that there really is magnification invovled. People with smaller sensors do not, in my experience, make smaller prints. So if you take a shot on a Canon 5D at 100mm and take a shot on a Rebel XTi at 100mm and then view them side-by-side, the Rebel shot has indeed been magnified by 1.6x. The magnification occured at the point of printing/displaying rather than at the capture point, but it still occurred.

I'm not saying that there isn't a cost to relying on the crop factor for magnification. The cost comes in having either fewer or smaller pixels, both of which are bad.

If you took two pictures with the following givens:
1) Two identical cameras with different sensors
2) Both sensors were magically perfect
3) One sensor was 1.6x smaller than the other
4) Both sensors had the same number of pixels
5) You took a picture with a 320mm lens with the big sensor and a 200mm lens with the small sensor.
6) Everything in both pictures was perfectedly in focus (no circle-of-confusion, bokeh, DOF issues in this case).
7) You printed both pictures at the exact same size.

Both pictures would be identical.

And just to re-raise this old argument...if you handheld both cameras and shook them an identical amount, the visible affect of the camera shake would be identical despite the different focal lengths.

 

[photo_chick]

It seems like this is all saying that you can't really zoom in any closer with a 1:6 crop sensor that you could with a FF sensor... But that it might appear that way because of the way the 1:6 sensor captures the image????? JUst because the moon physically takes up more space in the frame does not mean that it is actually magnified more...

Ok.. am I getting the idea here??? You guys have me so confuzzled, which admittedly is an easy thing to do.

 

[MarkBarbieri]

JUst because the moon physically takes up more space in the frame does not mean that it is actually magnified more...

Yes and no. It is not magnified more on the sensor. The image of the moon is the exact same size on your sensor, regardless of how large your sensor is. However, if you print both pictures the same size, you are magnifying the image from the smaller sensor more than you are magnifying the image from the larger sensor. So by the time it is printed, the picture from the cropped sensor is actually magnified more.

 

[boBQuincy]

Yes and no. It is not magnified more on the sensor. The image of the moon is the exact same size on your sensor, regardless of how large your sensor is. However, if you print both pictures the same size, you are magnifying the image from the smaller sensor more than you are magnifying the image from the larger sensor. So by the time it is printed, the picture from the cropped sensor is actually magnified more.

Yes, but... magnifying the image after it is captured is digital zoom. If we want optical zoom, the only true zoom, it must be at the time of image capture.
Can the crop sensor give us that? Sort of.

Your assumptions in your previous post appear entirely correct to me, and they may be the same assumptions used by most people when discussing crop sensors. Where they don't seem to hold is that most crop sensors do not have the same number of pixels, nor the same pixel pitch, as full frame sensors.
The other factor that must be applied before the assumptions do not hold is that the image size must be less than the full crop sensor. When that happens the crop is no longer the determining factor, we have sensor size to spare so it doesn't matter how much larger the sensor is, it isn't used.

So now we have a small portion of the sensor being used, and the crop sensor may have more pixels in that portion than the full frame sensor does. In this case we are actually getting more optical zoom, the free lunch.

 

[fitzperry]

Yes and no. It is not magnified more on the sensor. The image of the moon is the exact same size on your sensor, regardless of how large your sensor is. However, if you print both pictures the same size, you are magnifying the image from the smaller sensor more than you are magnifying the image from the larger sensor. So by the time it is printed, the picture from the cropped sensor is actually magnified more.

But isn't that sort of analogous to digital zoom? The optics of the lens are the same. It's just the print that's being enlarged?

 

[photo_chick]

But isn't that sort of analogous to digital zoom? The optics of the lens are the same. It's just the print that's being enlarged?

That is want I was thinking.

I think I am going to stop trying to figure it out and just continue to make prints in sensor ignorance!!!!!!

 

[Master Mason]

i have seen that image pop up a few times...

Hey when your good your good And I hate to try and reinvent the wheel when you have done such a supurb job. But you will notice I did and do give proper credit.

 

[Groucho]

It's disappointing to see that from Steve, who has a really good website full of information.
The term "multiplier" is one of the ideas that I believe is causing much of the confusion. It may apply ok to a P&S with it's tiny sensor where a "normal" lens is 10 mm.
In a dSLR where the sensor is much closer to full frame size, the term "divider" is more correct in that we are getting less of an image than a full frame camera would.

Well, to be fair, math terms are pretty flexible on this stuff, aren't they? I think the multiplier is just used to judge the result of the crop factor, not the crop factor itself. Ie, a 100mm lens on a 1.5 camera will have approx. the same field of view as a 150mm lens on a 35mm... or 160mm for a 1.6 camera... or 200mm for a 2.0 camera. It's not the same as the longer lens, but it's probably the easiest way for mere mortals to quickly gauge relative field of views.

My preferred way of explaining the crop factor is to explain it as an actual crop. Imagine you took a picture with your film camera, made a print, then cropped off a bit of it on all sides. Voila, there's your DSLR image. Simple and easy-to-understand IMHO.