Crop + F2.8 v Full Frame + F4

[HoppyUK]

Um. You've lost me. So I shoot a portrait at f1.4. DoF is so shallow that only the tip of the nose is in focus. I shoot at 12mp, so 4200 pixels on the long edge. Are you saying that if I crop to 800 pixels the DoF changes? Or are you saying if I resize the image to 800 pixels that the DoF changes?

Put it another way. If you take a picture on full frame and print it out A4 and view from 15in, you will perceive a certain depth of field.

If you then take that same print and chop off the sides to leave a 5x7in print, what you will have left is essentially the same as the print you would get if you shot on a crop format camera (same viewpoint, focal length and f/number) and printed the whole image out to 5x7in.

So, some people would then say, nothing has effectively changed, therefore the DoF cannot have changed.

But of course it has changed. You have a smaller print, which must then be viewed from a closer distance, ie about 8in, to restore the DoF viewing parameter. So you're now looking at the image more closely, your eye can detect more detail, so the DoF has changed accordingly.

Edit: crossed post with Andy

 

[Cistron]

Put it another way. If you take a picture on full frame and print it out A4 and view from 15in, you will perceive a certain depth of field.

If you then take that same print and chop off the sides to leave a 5x7in print, what you will have left is essentially the same as the print you would get if you shot on a crop format camera (same viewpoint, focal length and f/number) and printed the whole image out to 5x7in.

So, some people would then say, nothing has effectively changed, therefore the DoF cannot have changed.

But of course it has changed. You have a smaller print, which must then be viewed from a closer distance, ie about 8in, to restore the DoF viewing parameter. So you're now looking at the image more closely, your eye can detect more detail, so the DoF has changed accordingly.

Edit: crossed post with Andy

That is a very nice explanation, but I think I see a flaw. For once, 6MP are in theory enough to print any size and the viewer won't see any pixelation, if the right distance is kept. I now believe (after having read various theories in this thread) that the perceived increase in DoF propbaly comes from having to use shorter focal lengths on cropped sensors. I'm not sure what the optical explanation is, but it probably also has to do with the amount of detail we can discern in the regions that are out of focus, since they appear much smaller with wide-angle lenses.

 

[MomentCapture]

delete

 

[MomentCapture]

That is a very nice explanation, but I think I see a flaw. For once, 6MP are in theory enough to print any size and the viewer won't see any pixelation, if the right distance is kept. I now believe (after having read various theories in this thread) that the perceived increase in DoF propbaly comes from having to use shorter focal lengths on cropped sensors. I'm not sure what the optical explanation is, but it probably also has to do with the amount of detail we can discern in the regions that are out of focus, since they appear much smaller with wide-angle lenses.

With FF you either need to use a longer lens to get the same field of view (longer lenses have shallower DOF) or you need to get closer and the closer you get to something the less DOF you have.

The reason why there isn't much difference with wide angle lenses, is because practically everything is in focus so there isn't much difference to tell.

 

[HoppyUK]

That is a very nice explanation, but I think I see a flaw. For once, 6MP are in theory enough to print any size and the viewer won't see any pixelation, if the right distance is kept. I now believe (after having read various theories in this thread) that the perceived increase in DoF propbaly comes from having to use shorter focal lengths on cropped sensors. I'm not sure what the optical explanation is, but it probably also has to do with the amount of detail we can discern in the regions that are out of focus, since they appear much smaller with wide-angle lenses.

Yes, that's right. When people say that you get more DoF with smaller sensors, it's actually a shorthand for 'you get more DoF with smaller sensors, because you must use a shorter focal length to get the same framing'.

But IMHO the shorthand is acceptable, because it's the size of the sensor (which is your fixed starting point) that dictates everything else. That's the way we work - start with the camera, then adjust lenses to match.

The DoF principle has nothing to do with pixels and simply assumes that there is enough detail available in the system. And in practise, there always is and sensor resolution doesn't come into it.

With FF you either need to use a longer lens to get the same field of view (longer lenses have shallower DOF) or you need to get closer and the closer you get to something the less DOF you have.

The reason why there isn't much difference with wide angle lenses, is because practically everything is in focus so there isn't much difference to tell.

Please don't say that Rhys! It's untrue and is one of the few things we seem to agree about around this subject.

If you frame the subject the same, ie closer with a wide lens and then further away with a longer lens, DoF remains the same for all practical purposes. Lots of other things change, like perspective and field of view, but not DoF.

 

[fracster]

If you frame the subject the same, ie closer with a wide lens and then further away with a longer lens, DoF remains the same for all practical purposes. Lots of other things change, like perspective and field of view, but not DoF.

Please bear with me Richard, just need something explaining in simple terms.

Say I shot a kingfisher at F8 with my 500 F4 on a D700 and then took the same shot with my 300 F4 +1.7 TC, the DOF would be the same in both shots even though the sensor size is different?

(I know the focal length is not exactly the same, but near as dammit.)

Also, does the bokeh change with longer lenses, or am I imagining that?

Sorry for the dullard questions.

 

[HoppyUK]

Please bear with me Richard, just need something explaining in simple terms.

Say I shot a kingfisher at F8 with my 500 F4 on a D700 and then took the same shot with my 300 F4 +1.7 TC, the DOF would be the same in both shots even though the sensor size is different?

(I know the focal length is not exactly the same, but near as dammit.)

Also, does the bokeh change with longer lenses, or am I imagining that?

Sorry for the dullard questions.

When you fit a telecon, the lens spec changes. With a 1.7x, your 300 4 becomes 510mm f/6.8 in every respect. So if you shoot with both lenses at f/8, on the same camera, DoF would be the same. (You mention a different sensor size - what sensor size?)

Bokeh would also stay essentially the same, though bokeh is a combination of lots of factors - f/number, focal length, camera/subject/background distance, and also the individual characterists of the lens - optical design (spherical aberration particularly) aperture shape etc affects how out of focus areas are rendered.

 

[chrisgeary]

It's simple to try. Take any of your images - the one that you think looks the sharpest. View the image so it is displayed at 20% and look at the DoF. Now zoom it to 100%. Does the apparent DoF (i.e. the bits you think are and aren't in focus) change? Certainly does here. Same image, no processing, just displayed at a different size...

So I tried that in Lightroom. Opened an image and displayed it fully. Then zoomed in 1:1. I didn't see any difference on which areas were sharp and which were out of focus. The apparent DoF doesn't change. I really don't see how it can. When I click the shutter, the sensor sees the areas sharp and those which are OOF. This is fixed at capture time. I think we're talking about two different things because if DoF changed depending on output size, I'd have a hard time achieving the image I wanted at capture time.

 

[squishy]

So I tried that in Lightroom. Opened an image and displayed it fully. Then zoomed in 1:1. I didn't see any difference on which areas were sharp and which were out of focus. The apparent DoF doesn't change. I really don't see how it can. When I click the shutter, the sensor sees the areas sharp and those which are OOF. This is fixed at capture time. I think we're talking about two different things because if DoF changed depending on output size, I'd have a hard time achieving the image I wanted at capture time.

It does work as arad explained.
Take a shot where you've just missed the focus on something, and view it at a small size or from further away, and it'll still look sharp. However, as soon as you start to zoom in, you'll start seeing area as slightly OOF.
That's the reason everything always looks sharp on the 3 inch screen on the back of the camera, unless you zoom in to check.

The sensor doesn't see areas as out of focus and in focus. There is a plane of focus, then areas around it get further out of focus on each side. The size and distance the image is viewed at will determine how much the viewer sees as "in focus" either side of the plane of focus.

 

[HoppyUK]

So I tried that in Lightroom. Opened an image and displayed it fully. Then zoomed in 1:1. I didn't see any difference on which areas were sharp and which were out of focus. The apparent DoF doesn't change. I really don't see how it can. When I click the shutter, the sensor sees the areas sharp and those which are OOF. This is fixed at capture time. I think we're talking about two different things because if DoF changed depending on output size, I'd have a hard time achieving the image I wanted at capture time.

Chris, you're looking for subtle detail changes with a relatively crude output device, ie your PC monitor, than can't reveal much fine detail even when it's there.

The difference in perceived DoF between say f/5.6 and f/8, is slight in most pics. But the changes you're looking for are there if you look carefully. Do the print test I suggested above, outputting at 300 dpi, and with a subject where you can easily detect small shifts - maybe a ruler at close distance. That will show it.

But don't take my word for it. Input the figures into http://www.dofmaster.com/dofjs.html That's the science.

Edit: crossed post with Adam The LCD on the back of the camera test he mentions is a good and easy one - take a shot, and pretty much everything looks sharp. But zoom in to 10x and suddenly some areas are not - not at all sharp! Same principle.

 

[MomentCapture]

Please don't say that Rhys! It's untrue and is one of the few things we seem to agree about around this subject.

If you frame the subject the same, ie closer with a wide lens and then further away with a longer lens, DoF remains the same for all practical purposes. Lots of other things change, like perspective and field of view, but not DoF.

Thanks for correcting me on that, I'll check to double confirm later.

 

[akr]

If I'm understanding this all correctly, hypothetically

1) You own a Crop Camera (eg 40D) and an 85 f1.8. You would like a 50mm 1.4 for the shallow DOF and the shorter focal lengh. You can either but the 50mm 1.4, or (cost aside) you could buy a full frame camera and stick the 85 1.8 on it. At the same distance this would give roughly the same FOV and a shallowed DOF using the FF.

So if you have a crop body with say 50 f1.8, 85 1.8, 135 f2 and 200 f2.8 and you wanted more DOF could do so but getting a FF eg 5d for around £600 without having to replace your existing kit and you'd get (a very approximate) equivalent on the FF to the crop of 30 f1.4 or less, 50 f1.4 or less, 85 f1.6, 125f2. I have completed made up those F numbers simply on the basis of the DOF effect you would get on FF compared to a crop (I'm sure there is a calculation that could be applied I have just lowered them)- they clearly would remain at their original apertures for the purposes of exposure etc but simply in terms of DOF is that about right as a concept for FF v Crop?

And I assume that is why people often refer the DOF being amazing on FF.

The essence of the start of this thread was really if you want more DOF, used a crop camera instead of getting lots of more expensive lenses, that were heavier, could you simply get a FF camera and get your DOF that way (I realise there are many other factors to getting large aperture lenses).

So if you don't shoot sports, would a 24/105 f4 and 70-200f4 on a FF give you the same DOF as a 24/70f2.8 and 70-200f2.8 where the FOV was the same (I accept the short and long ends of the focal ranges would be different).

 

[HoppyUK]

Thanks for correcting me on that, I'll check to double confirm later.

Full frame, f/8

200mm at 4m, Dof = 18.2cm
100mm at 2m, DoF = 18.3cm
50mm at 1m, DoF = 18.4cm
25mm at 0.5m, DoF = 18.9cm

But massive changes in perspective and field of view! The field of view thing often gives the impression of more shallow DoF with longer lenses, becuase the background becomes relatively magnified, out of focus shapes become bigger and the background looks less cluttered and distaracting so the subject tends to stand out more. Which can look like a DoF effect, but is different.

 

[HoppyUK]

If I'm understanding this all correctly, hypothetically

1) You own a Crop Camera (eg 40D) and an 85 f1.8. You would like a 50mm 1.4 for the shallow DOF and the shorter focal lengh. You can either but the 50mm 1.4, or (cost aside) you could buy a full frame camera and stick the 85 1.8 on it. At the same distance this would give roughly the same FOV and a shallowed DOF using the FF.

So if you have a crop body with say 50 f1.8, 85 1.8, 135 f2 and 200 f2.8 and you wanted more DOF could do so but getting a FF eg 5d for around £600 without having to replace your existing kit and you'd get (a very approximate) equivalent on the FF to the crop of 30 f1.4 or less, 50 f1.4 or less, 85 f1.6, 125f2. I have completed made up those F numbers simply on the basis of the DOF effect you would get on FF compared to a crop (I'm sure there is a calculation that could be applied I have just lowered them)- they clearly would remain at their original apertures for the purposes of exposure etc but simply in terms of DOF is that about right as a concept for FF v Crop?

And I assume that is why people often refer the DOF being amazing on FF.

The essence of the start of this thread was really if you want more DOF, used a crop camera instead of getting lots of more expensive lenses, that were heavier, could you simply get a FF camera and get your DOF that way (I realise there are many other factors to getting large aperture lenses).

So if you don't shoot sports, would a 24/105 f4 and 70-200f4 on a FF give you the same DOF as a 24/70f2.8 and 70-200f2.8 where the FOV was the same (I accept the short and long ends of the focal ranges would be different).

Yes, there is a very simple calculation - it's f/number x crop factor.

When you equalise framing with the two formats, by adjusting focal length, the difference between full frame and crop is about 1.2 stops in terms of DoF, ie f/8 on full frame is about the same as f/5 on a cropper.

Your theory stands. It's just not very practical and you lose the other benefits of lower f/numbers, like low light shooting.

 

[akr]

Yes, there is a very simple calculation - it's f/number x crop factor.

When you equalise framing with the two formats, by adjusting focal length, the difference between full frame and crop is about 1.2 stops in terms of DoF, ie f/8 on full frame is about the same as f/5 on a cropper.

Your theory stands. It's just not very practical and you lose the other benefits of lower f/numbers, like low light shooting.

Ok thanks and thanks everyone for some useful posts in this thread.

 

[chrisgeary]

Well I was never very good at theory, physics or maths. I'll just accept what you're saying and be happy with my images

Confusing stuff nonetheless!

 

[MomentCapture]

Full frame, f/8

200mm at 4m, Dof = 18.2cm
100mm at 2m, DoF = 18.3cm
50mm at 1m, DoF = 18.4cm
25mm at 0.5m, DoF = 18.9cm

But massive changes in perspective and field of view! The field of view thing often gives the impression of more shallow DoF with longer lenses, becuase the background becomes relatively magnified, out of focus shapes become bigger and the background looks less cluttered and distaracting so the subject tends to stand out more. Which can look like a DoF effect, but is different.

Ah, that's what it is, thanks for clearing that up.

 

[HoppyUK]

Well I was never very good at theory, physics or maths. I'll just accept what you're saying and be happy with my images

Confusing stuff nonetheless!

Chris, DoF is a bit of an illusion, an optical trick based on the fact that the human eye can only resolve a certain amount of detail. So while there may be more detail available in a given optical system, if you can't see it, it makes no difference.

That visual limit is set by the standard size print viewed from a distance equal to the diagonal. DoF works by calculating how much of the subject is rendered at that level of detail (the size of the the circle of confusion) or smaller. In which case, under the set viewing standards, everything looks equally sharp.

Of course, it's not actually all equally sharp. Only the area on the exact plane of focus is as sharp as it possibly can be, and the image becomes gradually more blurred away from it. So if you change the viewing parameters and look more closely, what previously looked sharp to the eye is now less sharp.

HTH!

 

[chrisgeary]

Thanks for trying Richard et al. :bonk:

 

[arad85]

When I click the shutter, the sensor sees the areas sharp and those which are OOF. This is fixed at capture time.

Noooooooooo.... A lens is only perfectly in focus at one precise distance. Anything behind or in front of that is out of focus compared to the perfectly in focus area. It's how a lens works. It's physics. There isn't a concept of "in focus distance" or "out of focus distance". There is one precise distance (on axis to the lens just to simplify) that is in focus. Everything else is blurred to some degree or other.

Whether you perceive something to be in focus nearer or farther away from that exact in focus point depends on how the image is displayed. Display it really, really large and look closely and it'll all be out of focus (even the in focus stuff as the sensors will start to come into play!). Display it at the "usual" distance and some is in focus, some out of focus. Display it a long long way away, and it'll all appear to be in focus. This happens in everything you look at that is a rendered image. TV, newspapers, magazines, billboard ads, paintings - you name it.

Bottom line: DoF isn't made when the image is captured, it's made when the image is displayed and viewed.

 

[fracster]

DoF would be the same. (You mention a different sensor size - what sensor size?)

D700 is FF D300 is a 1.5 crop. But I think I have grasped it now by reading what else has been posted today.

Thanks.......