1.6 crop factor to full frame question

[tdigi]

I was comparing a 40D and a 5D and it appears to me that you are not really zoomed in closer

when using a 40D you just have a smaller field of view. Is that correct? If so how does a 100mm

lens appear as 160?

 

With 100mm on a 40D aren't you just seeing what you see at 100mm on a 5D but just less of the

frame?

 

Can someone please help explain this? I thought I understood it but I guess I don't.

[dave_holland]

Conceptually I understand your point. However don't forget that you normally look at your photos on screen or enlarged, usually to full screen size. Say you print two photos, one with a 5D full frame at 8X12 and a second with a 40D at 8X12. The frame size is the same, so the 40D subjects appears to be closer. Sure, it's more cropped. But it's also enlarged more, to create the same 8X12 print.

 

Dave

[zml]

<i>how does a 100mm lens appear as 160?

</i><br>It does not. A crop sensor registers s smaller area of the image projected by the lens. <br><br><i>With 100mm on a 40D aren't you just seeing what you see at 100mm on a 5D but just less of the frame? </i> <br>Yes.

[peter_mckone]

If you crop a film negative, and then print the cropped image on the same size paper as the original image was printed on, the cropped image will appear to be "zoomed" (ie. bigger).

 

The D40 sensor is not large enough to capture the full size image produced by ordinary lenses. So you get a crop of the true image. Then, when you view the D40 image, you see a smaller field of view (compared to a 5D), and if you view both images on the same size monitor, the D40 image appears to be "zoomed". If the D40 sensor has as many pixels as the 5D, then (almost) no image quality would be lost.

[jason_hall4]

Tommy,

 

All lens produce a image circle. When you place a lens on any camera "Crop Sensor" or "full Frame" the image circle is exactly the same. Nothing about the lens changes. The difference is that the "Crop camera" has a smaller sensor so it is looking a smaller portion of the image circle. This gives you the effect of zooming in more.

 

It is no different than taking an image from a full frame camera and croping it to see what a croped camera "sees". That is the simplest way I know to explain it.

 

How does the 100mm look at 160mm? It appears no different on either camera. Again you are just seeing a smaller portion of the image circle.

 

Where things can change is that due to the fact you are seeing less of the circle it will cause you to get futher from the subject and so now the focus distance is different. This can change the DOF you have at a given aperture vs if you had the lens on a full frame camera and would have to get get closer to have the same view in the view finder. (boy that was a mouth full).

 

If you took a photo with the 100mm macro on a 40D and then on a 5D from the same distance. then crop the 5D image to give the same view as the 40D, they would appear to be exactly the same, apart from change in resolution due to pixel pitch and other difference in sensor and camera image processing. Hope that made sense.

 

Jason

[colinsouthern]

Tommy DiGiovanni wrote: "how does a 100mm lens appear as 160?"

 

Michael Liczbanski responded" "It does not. A crop sensor registers a smaller area of the image projected by the lens."

 

There's a common misconception creeping in here. Yes - crop sensors do "throw away" part of the image circle, but it doesn't stop there. Hopefully this cut and paste from a previous post I made helps to clear up this "focal length multiplier" saga ...

 

...

 

Michael, what you wrote WOULD be true if both sensors had the same pixel spacing; Unfortunately, they don't.

 

To simplify things, consider that the "long" dimension of a 5D sensor is 35.8mm, and it has 4368 pixels - giving 122 pixels per millimetre. A 40D Sensor is 22.7 millimetres, and it has 3888 pixels - giving 171 pixels per millimetre.

 

Although you're quite correct in saying that the 40D has a reduced field of view, it's incorrect to imply that you can simply crop a FF sensor to the same size as a crop sensor, and get the same result - a small print may well look the same, but as you enlarge the image you'll see that the crop-sensor on a 40D has captured significantly more information than the 5D (ie "increased resolving power") - when you enlarge this image to the same size as one shot on a FF camera you end up with an image that has roughly (keep in mind that the 40D sensor is 2MP less than a 5D anyway) the same pixel density, but at the reduced field of view - the net result is an effective increase in focal length.

[tdigi]

So at 100mm on a 40D your not really seeing 160mm? Your just seeing 100mm with

the outside of the image cut off.

[tdigi]

Colin that makes sense. So that is what is meant by effective focal length.

[jason_hall4]

You are seeing a smaller cropped portion of the image circle projected by the lens with a 40D. A 5D is looking at a larger portion, it is infact less cropped. Properties of the lens does not change in any way.

 

Colin, lets don't muddy the water just yet. :o)

 

Jason

[kelly_flanigan1]

In movie work on just "thinks" in angular coverage: thus one is not confused about this issue. A 100mm lens could be on a 4x5" camera; a MF camera; a 35mm camera; a 1.6x crop dslr; a 16mm cine camera; or a 8mm cine camera. With each smaller format one has less angular coverage. 160mm lens on a full frame 35mm camera will have the same angular coverage as a 100mm lens on a 1.6x crop camera. A 100mm lens is still a 100mm lens no matter if you are using it on a Hassleblad; or burning ants; or using it on a 16mm cine camera.

[jason_hall4]

On the other hand.... maybe I am the one makeing things muddy. :o)

 

Jason

[g dan mitchell]

Your statement is true, but sort of an unusual way to think of this.

 

By the way, the notion of denying that a "100mm lens appears as 160mm lens"

when used on a crop body is an unfortunate extension of a truth that turns out to be

confusing and misleading when applied this way.

 

In fact a 100mm lens does "appear as" a 160mm lens when used on a crop sensor

camera. In virtually all ways - and let's not quibble about DOF, ok? - you could use a

160mm lens on a FF body and a 100mm lens on a 1.6x crop body and the image

would "appear" essentially the same.

 

Yes, the 100mm lens is still a 100mm lens on a crop sensor or a full frame (or MF or

LF) body. Yes, the 100mm lens gives you the same angle of view on a crop body

that a 160mm lens would provide on a FF body.

 

Dan

[zml]

Actually, Colin, you are wrong here: the FF and crop sensors see exactly the same image from the 100 mm lens in question (a smaller sensor will see a smaller area of course.) Now, when talking about linear enlargement of the image, pixel density and such start playing role but that is a totally different can of worms. Say, you are talking about enlarging both images to the same size. Why? I'd much prefer sticking with the same linear enlargement for both images, say 10x, eh? Yeah, the sizes will differ, but the math will be cleaner :-)

[tdigi]

Now I am even more confused. Makes me want to just go get a 5D. :}

[JDMvW]

What counts is not only the image size and its crop, but ALSO the resolution. Yes the perspective is the same, etc. but there will be as much or more data from the sensor with higher resolution. The other aspect is that on lenses made for FF, often (not always) the part of the lens coverage that is used by the crop sensor camera will be the "sweet spot" --that is the 'best' resolution part of the image cast by the lens.

 

Lenses made especially for crop-sensor cameras can also be cheaper, in theory, since they don't have to cover so large an area.

 

Other factors that enter into the equation besides crop, resolution, and sharpness include NOISE. It is still easier to make a large sensor with less noise for a given pixel by pixel sized image. That's a factor in why some of the P&S cameras with huge MP counts, don't really make such great pictures--the tiny sensor heats up and produces much noise.

[photo_dark]

Im not sure how this thread got so convoluted so quickly... Either way, the original couple of responses are correct. The lens doesn't change, none of the properties change, you just are taking a smaller piece of the image circle pie.

 

Since the crop is different, to get the same 'framing' you would have to use a different focal length.

 

Using a 100mm lens on a crop sensor does NOT mean that you get the same 'compression' effect as you would with a 160mm lens... everthing is the same, you are just seeing less of the image circle.

 

I think people get confused because the general consensus is that using a 100mm lens on a crop sensor is 'the same as' using a 160mm on a full frame, when in fact it isn't, you're just getting the same framing/crop that you would get with the 160 on a full frame.

[james_gu]

This is an interesting topic. Does anyone have samples?

[colinsouthern]

"Actually, Colin, you are wrong here"

 

Actually Michael, I'm right :)

 

I'm not saying that the focal length actually changes - it doesn't. BUT (and it's a big but) - the net effect is - with regard to focal length, not DOF - is that it's the same as if the focal length had increased.

 

The big problem is that usually people only consider one side of the equasion (the reduced image circle).

 

Let me put it another way - if I get a 5D with a FF sensor - get out my roll of duct tape - and mask off a portion of the sensor so that what's visable has the same dimensions as the sensor of a 40D. When I take a shot with the 5D and print it out as a 6 x 4 (and I do the same with the 40D) what's printed shows me the same field of view as the 40D, but I've got all this black stuff around the edges on the 5D shot where the light couldn't hit the sensor because of the tape. Black stuff doesn't look good so I cut it off - but now I have a smaller picture (but same FOV) from the 5D than what I got out of the 40D. What-to-do. I know - I'll stick the smaller 5D print on the nearest (and theoretically perfect) enlarging photocopier and "zoom" it up to the same size as the, 6 x 4 that I got out of my 40D. So now the two shots are the same right? WRONG? They now have the same FOV, but the enlarged shot from the 5D won't contain anywhere nearly the same detail (because of the reduced pixel density). If this is acceptable then that's fine - BUT - we're comparing apples with apples here - we don't just want to crop a FF shot to the same FOV - we want the same picture quality - and guess what - there's only 1 way to get it - and that's to use a lens on the FF camera that's 1.6x longer so that it resolves the same detail, but spreads it across the ENTIRE sensor so that that detail isn't lost.

 

So - in summary - although the ACTUAL focal length hasn't changed, on a FF camera you'd need a lens 1.6x longer to give you the same resolving power for that FOV.

 

Try it - grab a 5D and take a photo of a number plate - take one with the same lens on a 40D - zoom them both in - you'll have a lot more detail. Only way to get that same detail from the 5D is to use a lens 1.6x as long.

 

DOF is a differnet issue, but we're not talking about that.

 

Cheers,

 

Colin

 

* And for the benefit of the para-legal fraternity - yes, I know the 5D is 10MP, and the 40D is 8MP so yes, that does increase the pixel density somewhat - none the less, a 5D sensor cropped to 40D size is STILL only about 5MP.

[kelly_flanigan1]

If that 100mm lens is on my 4x5 digital scan back' its active area is 7x10cm. A 2.4x3.6cm section thus is a fraction of the 35megapixels; ie a lazy 4.3 megapixels for a full frame 35mm. This is way less than a 5D's resolution; but the 24x36mm section has the same angular coverage.

[g dan mitchell]

"Using a 100mm lens on a crop sensor does NOT mean that you get the same

'compression' effect as you would with a 160mm lens... everthing is the same, you are just

seeing less of the image circle."

 

If by this you mean to imply that there would be a different "compression effect" from a

100mm lens on a 1.6x cropped sensor camera in comparison to a 160mm lens on a FF

camera...

 

... you are just plain wrong.

 

On the basis of relationships among objects in the frame, you cannot tell the difference

between the 100mm/crop shot and the 160mm/FF shot.

 

Dan

[colinsouthern]

I think we need to form a tag team here Dan!

[freelance]

I have made a picture with a macro, using the same target and two cameras, a 5D and a 40D. After cropping the image I wanted to get, with the 5d I get a crop of 1659 x 1014 and with the 40D another of 2463 x1409. It is obvious that I have more pixels to "see" or print the object with the crop factor. No matter the quality of these pixels.

 

You can see the pictures at the end of the set:

 

[freelance]

The lens was the same in both cases

[chrisjb]

Hmm all sorts of ideas, If you could do a contact print from esch sensor, (never mattered how tight the grain with film) to make the 40d the same size would that be enlarged x1.6? and as there isn`t a 100mm an EF 50mm compared to EFS 50mm (kit lens maybe) how much difference to you see knowing both have different image circls? should be none? :)

[chrisjb]

that would be easier to see if sensors placed on some printed image :)