Focal lenght and crop sensors

[samsnikon]

Apologies for being thick but I don't understand why the focal length increases on a crop sensor. I think I understand it intuitively but not properly if that makes sense.

Would someone mind explaining as simple as possible?

I have explored the web but cannot find something that explains it well

thanks

 

[chuckles]

You don't understand it for one simple reason - - - it doesn't increase!

It just gives the illusion of an increase because the sensor 'crops' it relative to a 'normal' full frame equivalent.

Manufactures spawned this nefarious deed to 'help' people understand the effects of any particular lens on a crop sensor. It's a situation that was a good idea once but, is now, totally out of hand. :bang:

(Sorry for the over use of quotes)

 

[onomatopoeia]

The focal length doesn't increase. "Crop" sensors are exactly that, they crop the edges from the resulting image as compared to full frame. The result is that whatever is in the centre of the frame appears closer, as if a longer lens was used. So you still get the perspective and depth of field from the lens you used rather than a lens with 1.5x the focal length, you just don't get as much of the resulting picture.

Think that's right anyway

 

[NikonSimon]

(Sorry for the over use of quotes)

No need to 'apologise'

 

[NikonSimon]

I was wondering something similar.

So, in my depth of field app on my Android phone, when it asks for the focal length, I should put in the actual focal length, and not the 'crop factor' focal length? i.e. on a nifty fifty, I should enter 50 and not 75 ish

 

[samsnikon]

so bascially to fit the image within a crop sensor you end up moving it closer - (sorry difficult to articulate as I have an image of two triangles in my mind and am imagining a full frame and cropped frame) but I think I get it!

Simon - on the phone I think you put in 75mm as the software will not compensate for a crop sensor

 

[Jakska]

Kind of samsnikon, might be easier to explain with a diagram:

The circle is the size of the image a lens (lets call it a 50mm lens as an example) makes, it has to be big enough to cover whatever format you're using otherwise you start to get vignetting around the edges. The blue rectangle represents 35mm film/full frame digital - this is what the 50mm looks like when the whole image is visible. The red rectangle represents a digital crop sensor - you're still using the same focal length; same perspective, same depth of field, same amount of barrel distortion as the 50mm originally had, but you are only using a smaller amount of what the lens can actually see to make up the photograph, and therefore seeing less/it looks like you've got a longer lens.

 

[scottthehat]

full frame =35mm so say d700/d3, put a 50mm lens on you get a 50mm lens with a field view of 50mm.

put that 50mm lens on a 1.5x crop camera(say d3000/5000/90/300)
and the 50mm lens is still 50mm but because of the crop factor it gives you the field of view of a 75mm lens on a full frame camera.
so to get the same pictures standing in the same place with d300 and a d700 you would need a 50mm on the d300 and a 75mm lens on the d700.
http://en.wikipedia.org/wiki/Crop_factor

 

[NikonSimon]

full frame =35mm so say d700/d3, put a 50mm lens on you get a 50mm lens with a field view of 50mm.

put that 50mm lens on a 1.5x crop camera(say d3000/5000/90/300)
and the 50mm lens is still 50mm but because of the crop factor it gives you the field of view of a 75mm lens on a full frame camera.
so to get the same pictures standing in the same place with d300 and a d700 you would need a 50mm on the d300 and a 75mm lens on the d700.
http://en.wikipedia.org/wiki/Crop_factor

So on my DOF app on my phone, I should put in the focal length of the lens I'm using/shot I'm taking, rather than factor in the crop factor? I'm using a D300s

e.g. enter 50mm and not 75?

 

[scottthehat]

So on my DOF app on my phone, I should put in the focal length of the lens I'm using/shot I'm taking, rather than factor in the crop factor? I'm using a D300s

e.g. enter 50mm and not 75?

The depth of field may change, depending on what conditions are compared. Shooting from the same position, with same f-number, but enlarging the image to a given reference size, will yield a reduced depth of field. On the other hand, compared to the 35 mm-equivalent focal length shooting a similarly-framed shot, the smaller camera's depth of field is greater.
so i would be entering 50mm and not 75mm.

 

[NikonSimon]

The depth of field may change, depending on what conditions are compared. Shooting from the same position, with same f-number, but enlarging the image to a given reference size, will yield a reduced depth of field. On the other hand, compared to the 35 mm-equivalent focal length shooting a similarly-framed shot, the smaller camera's depth of field is greater.
so i would be entering 50mm and not 75mm.

cheers. When I start using it I'll stick with the focal length I'm shooting at, and not the focal length with the crop factor, and see how it goes

 

[samsnikon]

Jakska loved the pic

sam

 

[HoppyUK]

Kind of samsnikon, might be easier to explain with a diagram:

The circle is the size of the image a lens (lets call it a 50mm lens as an example) makes, it has to be big enough to cover whatever format you're using otherwise you start to get vignetting around the edges. The blue rectangle represents 35mm film/full frame digital - this is what the 50mm looks like when the whole image is visible. The red rectangle represents a digital crop sensor - you're still using the same focal length; same perspective, same depth of field, same amount of barrel distortion as the 50mm originally had, but you are only using a smaller amount of what the lens can actually see to make up the photograph, and therefore seeing less/it looks like you've got a longer lens.

Good diagram! But a few other things do change. They're usually relative details but certainly depth of field is different (because the final image has to be enlarged more) and there will be less distortion visible, less vignetting, and potentiall slightly increased sharpness because you're using the 'sweet spot' in the centre of lens (off-set by the greater degree of enlargement). Perspective remains the same because the shooting distance hasn't changed.

But in practise it's the DoF change that's most significant. When you restore the same framing by using a shorter focal length lens on the crop format camera, the DoF changes by the crop factor, ie f/number x 1.6 (or 1.5x Nikon). So f/4 on a crop camera delivers the same DoF as f/6.4 on full frame. It's about one and a quarter stops difference.

cheers. When I start using it I'll stick with the focal length I'm shooting at, and not the focal length with the crop factor, and see how it goes

Yes, you should put in the actual focal length. The DoF calculator should have an option of full frame or crop format where you input the camera type, which accounts for all the various changes.

This is a good calculator www.dofmaster.com

 

[scottthehat]

cheers. When I start using it I'll stick with the focal length I'm shooting at, and not the focal length with the crop factor, and see how it goes

ok, its like saying that if you have a 500mm on a d300 you should have a min shutter speed of 750/sec but you only need 1/500th.

 

[NikonSimon]

Yes, you should put in the actual focal length. The DoF calculator should have an option of full frame or crop format where you input the camera type, which accounts for all the various changes.

This is a good calculator www.dofmaster.com

Chars.

Yes, it lets me choose the camera type (annoyingly I have to do it every time), so I assume when I select Nikon D300s, and enter 50mm for my nifty, it would take into consideration the crop factor etc

 

[HoppyUK]

ok, its like saying that if you have a 500mm on a d300 you should have a min shutter speed of 750/sec but you only need 1/500th.

Not quite sure what you're saying there, but it raises a good point.

When you're hand holding longer lenses, the rule of thumb is that your shutter speed should not be lower than the focal length of the lens, to avoid camera shake. That is 1/500sec for a 500mm lens.

But what matters here is not the focal length, but the effective magnification, which is of course increased on a crop camera. So the rule has to be revised to focal length x crop factor. In this case, on a Nikon cropper it would be 1/750sec.

From which you can see that actually the rule is nothing to do with focal length directly and it's just a happy coincidence of numbers which makes them match up like that on full frame cameras.

Chars.

Yes, it lets me choose the camera type (annoyingly I have to do it every time), so I assume when I select Nikon D300s, and enter 50mm for my nifty, it would take into consideration the crop factor etc

That DoFmaster link is very good for comparing things, and there's a huge amount of info on there if you go into it.

To see how DoF changes between formats, if you input a 50mm lens and your D300s at a certain distance and f/number, then change to a D3 with 75mm lens but raise the f/number by 1.25 stops (that site runs in quarter stop increments) you'll find the DoF is identical.

 

[scottthehat]

Not quite sure what you're saying there, but it raises a good point.

When you're hand holding longer lenses, the rule of thumb is that your shutter speed should not be lower than the focal length of the lens, to avoid camera shake. That is 1/500sec for a 500mm lens.

But what matters here is not the focal length, but the effective magnification, which is of course increased on a crop camera. So the rule has to be revised to focal length x crop factor. In this case, on a Nikon cropper it would be 1/750sec.

From which you can see that actually the rule is nothing to do with focal length directly and it's just a happy coincidence of numbers which makes them match up like that on full frame cameras.



That DoFmaster link is very good for comparing things, and there's a huge amount of info on there if you go into it.

To see how DoF changes between formats, if you input a 50mm lens and your D300s at a certain distance and f/number, then change to a D3 with 75mm lens but raise the f/number by 1.25 stops (that site runs in quarter stop increments) you'll find the DoF is identical.

so what your saying is that when using a 500mm on a nikon the min shutter speed at 500mm should be 750/sec then, i have always been told that this does not change.
but would love to see proof.

 

[dinners]

It's also a good diagram for appreciating why a full frame camera deserves better glass which is sharp around the edges. Cropped sensors use just the sweet spot in the centre of the lens so don't make use of the edges.

 

[DorsetDude]

It's also a good diagram for appreciating why a full frame camera deserves better glass which is sharp around the edges. Cropped sensors use just the sweet spot in the centre of the lens so don't make use of the edges.

What about (for canon) EF and EF-S lenses? The EF-S ones I believe are designed to work with the cropped sensors and wouldnt work on a full frame camera. But the EF lenses, like my 70-300mm would work on both type camera bodies. With the cropped cameras you get the inner red box as shown above but on the 5DMK2 say, you get the full blue outer box view.

If that is true the EF-S lenses focal length doesnt need to be tweaked by 1.6 to give a relative new figure? Am I right?

Sorry if thats not clear!

 

[dinners]

......the EF-S lenses focal length doesnt need to be tweaked by 1.6 to give a relative new figure? Am I right?

Sorry if thats not clear!

EF-S only differ from EF in the mount and can only be attached to cropped bodies.

 

[HoppyUK]

so what your saying is that when using a 500mm on a nikon the min shutter speed at 500mm should be 750/sec then, i have always been told that this does not change.
but would love to see proof.

Yes, that's true, it's the 'effective' focal length that is the key figure, but actually, it's just a shorthand for the magnification - that's the thing which really matters.

Folks assume that since the focal length doesn't change when you change formats, that the hand-holding rule isn't affected. But it is, and as I said above, the fact that shutter speed and focal length just happen to work out the same with full frame, is pure coincidence.

By the same token, the hand-holding rule also changes when you enlarge an image, because you enlarge the amount of blur within it. It's a very rough rule, and actually I think it's pretty marginal. It should certainly be regarded as a minimim standard, as while a shorter shutter speed obviously reduces the affect of camera shake, it doesn't eliminate it. People should do their own tests to see where their individual limits lie. It also varies a lot according to how you hold the camera, leaning against a wall, down on one knee and stretching uncomfortably, elbows giving good support, that sort of thing.

To take an extreme example, you wouldn't try and hand-hold a compact with a 100mm lens at 1/100sec. An average compact has about a 5x crop factor and 100mm is a massive lens on one of those, about 500mm equivalent in terms of field of view, relative to full frame. So 1/500sec minimum.

Then there's IS of course Absolutely fantastic invention IMHO

 

[StewartR]

A couple of years ago I created a set of visualisations of these issues. I think they're still relevant and they are here.

 

[scottthehat]

Yes, that's true, it's the 'effective' focal length that is the key figure, but actually, it's just a shorthand for the magnification - that's the thing which really matters.

Folks assume that since the focal length doesn't change when you change formats, that the hand-holding rule isn't affected. But it is, and as I said above, the fact that shutter speed and focal length just happen to work out the same with full frame, is pure coincidence.

By the same token, the hand-holding rule also changes when you enlarge an image, because you enlarge the amount of blur within it. It's a very rough rule, and actually I think it's pretty marginal. It should certainly be regarded as a minimim standard, as while a shorter shutter speed obviously reduces the affect of camera shake, it doesn't eliminate it. People should do their own tests to see where their individual limits lie. It also varies a lot according to how you hold the camera, leaning against a wall, down on one knee and stretching uncomfortably, elbows giving good support, that sort of thing.

To take an extreme example, you wouldn't try and hand-hold a compact with a 100mm lens at 1/100sec. An average compact has about a 5x crop factor and 100mm is a massive lens on one of those, about 500mm equivalent in terms of field of view, relative to full frame. So 1/500sec minimum.

Then there's IS of course Absolutely fantastic invention IMHO

thanks richard so when im holding my 500mm (750mm in 35mm terms) and get shots sharp at 1/100th sec i should be well pleased.

 

[HoppyUK]

thanks richard so when im holding my 500mm (750mm in 35mm terms) and get shots sharp at 1/100th sec i should be well pleased.

Yes, very well pleased. It's about consistency - could you do that six times on the bounce? Probably not, not without IS anyway, but a good trick when you're marginal on shutter speed is to take several shots and there's a good chance one will be sharp.

And good holding technique makes a big difference - left hand cupped under the lens takes all the weight, left elbow pressed against the side of your chest to form a triangular brace, firm but relaxed, then breathe out and sqeeze the shutter. Compare that with holding the camera at arms length, watching the LCD compact-style - flippin hopeless!

The other thing of course is subject movement, and there's nothing you can do about that, with or without IS. It's purely down to shutter speed.