[stevebrot]

Not because of DOF considerations, but because you don't want to be, or can't be, 50ft from your subject... Even if in theory you could do it...

Yes! The laws of perspective apply and the eye knows them very well. Even though the lens may allow a tight framing, the perspective is still one of distance.


Steve

[stevebrot]

Perhaps more simply--DOF varies with f stop and magnification--at same f stop.
So different FL lenses at different distances have same DOF if:
-- the same object in focus is the same size (same magnification)
-- and the same f stop
-- and you print/view result at the same magnification (same size/same crop)

Yes! A corollary is that if distance is kept constant and magnification decreases (shorter focal length) a crop to maintain framing will have the same DOF as with the longer lens. What's more, your last point makes an extremely important qualification related to DOF preview, both with live view and the optical viewfinder. Live view will generally exaggerate DOF as will a low magnification optical viewfinder.*

It is somewhere around this point that the practical photographer will give up trying to fine tune where the limits of acceptable focus fall in their composition.


Steve

* This is the cause of user frustration with manual focus where "...it was sharp in the viewfinder/rear LCD, but OOF on my computer screen..."

[Bui]

Thank you very much all of you for your very detailed and helpful answers. In this topic I only want to learn and be sure about the relation between DOF-Aperture-Focal length, I'm not specifically talking about other things that also determine image quality (bokeh, contrast, etc), and also I'm not intending to do lenses comparison

I think there's a lot of confusion regarding DOF on FF versus crop. The DOF of a specific lens at a certain aperature at a certain distance from the subject is going to be the same regardless of the image sensor size. Where I think people get hung up on this is if you're moving the lens forward or backwards to frame the same area of the subject to account for the difference in what you see framed with FF versus crop, you will change the DOF in the process. If you stay in one position and change lenses to have the same framing or field of view going between FF and crop, you'll change the perception of DOF because you've changed the lens, and the DOF of the different focal lengths will be different. So while the actual DOF of an object 10 feet away for a 70mm at 2.4 will always be the same regardless of the sensor size, if you start changing the distance of the lens to the subject - get closer with the 70 on FF to get the same field of view as the 70 on crop let's say, then you'll "change" the perception of DOF, and this case, it will get "narrower".

Hope I said that in a way that makes sense

So let me summarize my understanding of what you said:

- If I shoot an object 5m away with a 70mm at f2.4, I will get a certain amount of background blurry
- If I change to a 50mm, also at f2.4, and move torward the object to have the same FOV as previous one, the amount of background blurry will increase because I move closer to the object?

If that is true then I think the argument of "choosing a longer lens will give more blurry background" is not valid because I can always move closer with a shorter lens (of course this doesn't take anything of distortion, colour, contrast etc into account)

- The "FF camera gives shallower DOF than ans APSC one" is not directly true, it is infact due to change in distance between the camera and the object in order to maintain the same composition.

In that case, one of the argument I read on the Internet, as why people won't consider a M4/3 camera, because of limited DOF choice as compared with APSC, or FF, I think is also not valid, because I can always move closer, with a wider lens if I'm using a M4/3 camera, to achieve the same DOF effect, no?

Also this leads to my last question: does FF camera doesn't have any advantage over APSC regarding FOV,? Is a picture taken with a 20mm on a FF camera the same as taken with a 14mm on APSC? (If we leave resolution and ISO out of the equation?)

---------- Post added 08-16-17 at 08:12 AM ----------

What depth of field calculators do not do, is give you any idea of the rendering of the out of focus area or bokeh. This is far more important that depth of field on its own, and generally is "nicer" with longer focal lengths, because the out of focus points appear as big blobs, compared to small circles (or what ever shape your aperture is)

Also consider when shooting portraits, it is as important to control the subject to background distance and background pattern as it is to control depth of field

Thank you, this is good to know

---------- Post added 08-16-17 at 08:16 AM ----------

It took me a while to get my head around the crop versus full frame thing as far as DOF goes. Yes, the focal length of a lens goes up by 1.5 when you put it on an APS-C camera body. But you also "gain" depth of field because of the smaller image - roughly one f/stop. This means that a 50mm f/1.4 lens will take about the same pictures as a 75mm f/2 on a full frame.

This explains why cell phone cameras have such enormous depth of field - it is down to the teeny tiny sensor. It also explains why a medium format camera is not easy to focus manually - bigger image, less DOF.

Another APS-C vs. 35mm example - a 200mm f/2.8 becomes a 300mm f/4 and not a 300mm f/2.8. As far as DOF goes anyway. The extra light from a faster aperture is still yours to take advantage of. Hope I have it right and that this helps.

Thank you. And to be sure, the change you mentioned is based on the condition that I don't change my position, right?

[clickclick]

So let me summarize my understanding of what you said:

- If I shoot an object 5m away with a 70mm at f2.4, I will get a certain amount of background blurry
- If I change to a 50mm, also at f2.4, and move torward the object to have the same FOV as previous one, the amount of background blurry will increase because I move closer to the object?

OK, let's just look at this from an optical standpoint with the lens and subject and aperture all at the same location/same setting. If you were shooting on the K1 in crop mode with the 70, an APS-C lens, the optical viewfinder shows a black outline of the crop image size, but overall, you see the full frame, and depending on the image circle of the lens, you may or may not have vignetting. You can change the mode of the K1 from crop to full frame. When you do this, the black outline goes away. Now, did anything come into or go out of focus for doing that? No. That's because optically, we made no change. Let's say we focused on the subject's eyes, and the tip of the nose was out of focus and the back of the ear lobes were out of focus. Did changing it from crop to full frame mode suddenly make these go in or out of focus? No. It just changed how much of the image circle was captured. To continue with this thought, let's assume the 70's image circle was big enough to fill the full frame (don't think it is, but let's pretend for our example), and you had it on a K1, but you wanted the equivalent framing of being on a crop body to fill the full frame view, you'd have to get closer to the subject to do so. When you move closer to a subject with any lens, the depth of field will get shallower. Again, this has nothing to do with sensor size, but just optics. The closer a lens is to the subject, the thinner the plane of focus. So moving closer to the subject to have the exact same framing will make the DoF shallower or thinner because of the optics of the lens, not the sensor size. Now maybe the cheeks and chin aren't as crisp as the eyes you were focusing on, but when further back, they might have looked pretty good. Now, this is where we get into another knot. As was mentioned right after my first post, Some view DoF as what was defined as sharp on an 8x10 print from 35mm film. This is a subjective application of measurement. It was decided that a point that still looked to have a certain level of fineness at a certain level of englargement as being in focus would be the front/back of your depth of field. Is that point really as sharp as the exact plane of focus? No. Will making the print bigger reveal the focus change more obviously? Yes. So when you start looking at different size prints from different formats, you start introducing other subjective perceptions of what was in focus or out of focus. The focus of the light converging on the plane of film or sensor hasn't changed by printing it bigger or smaller, but what the human eye can see and what any one person might define as acceptably sharp or in focus is affected, and here's where I think we go back down the rabbit hole and get all wrapped around the axle on DoF changing when changing from crop to FF. And finally, you mention changing from a 70 to a 50 to maintain a field of view. Well, changing the optics will change the DoF - longer lenses have a shallower range of focus in front/behind the exact plane of focus than wider lenses, so swapping the 50 for the 70 will bring in a whole new set of optical characteristics that are unrelated to the sensor size, but again, how much you enlarge it etc. will determine what you perceive as in focus and the DoF.

Hope this is making sense.

---

Last edited by clickclick; 08-16-2017 at 03:14 PM.

[MadMathMind]

- If I shoot an object 5m away with a 70mm at f2.4, I will get a certain amount of background blurry
- If I change to a 50mm, also at f2.4, and move torward the object to have the same FOV as previous one, the amount of background blurry will increase because I move closer to the object?

If that is true then I think the argument of "choosing a longer lens will give more blurry background" is not valid because I can always move closer with a shorter lens (of course this doesn't take anything of distortion, colour, contrast etc into account)

No, the 70mm will have a stronger blur. The strength of the blur depends on both focusing distance and focal length, but the focal length has a stronger effect over small changes in distance. It's subtle (and I used to have a really good example but deleted it from my Flickr), but the 70mm will be slightly stronger in blur than the 50mm.

If you use a short focal length, you'll have to get a lot closer but the blur will not be nearly as strong. Here's a non-perfect example, both shots at f/2.8:

At 31mm:


At 77mm:


Mari's position in the frame is about the same for both of these, but obviously I am a lot closer to her than I am with the 31 than the 77. Some differences:
- Everything in front of her is in good focus in the 31mm shot. You can see some slight blurring with the 77 in the forest ground
- There's some definition in the waves in the 31mm shot; you can see lines in the waves in the 31mm shot and in the 77 shot, these are far less defined
- The biggest differences are in DoF and right behind her. The sand behind her is mostly blurred out in the 77mm shot and in the 31, you can see well-defined footprints and tracks in the sand

This isn't a perfect comparison because the angles are different and there was no attempt made to get equal magnification (just happened to have two similar shots), but it does show you how focusing closer isn't quite enough. You need a long focal length and have to focus closely.

[clickclick]

No, the 70mm will have a stronger blur. The strength of the blur depends on both focusing distance and focal length, but the focal length has a stronger effect over small changes in distance. It's subtle (and I used to have a really good example but deleted it from my Flickr), but the 70mm will be slightly stronger in blur than the 50mm.

and just to continue the thought, this ties into my post above - the 70 as a longer lens has a shallower depth of field than the 50, so the "blur", the areas going out of focus, will be more pronounced sooner, and as you move further or closer from/to your subject, you will increase or decrease the depth of field, respectively.

I think what makes this challenging is the number of variables involved and tracking how changing just one component can have a trickle down effect that's greater than what is first anticipated.

[mikesbike]

Absolutely true that when going for blurring the background, the game changes depending on the distance of that background behind the subject, as well as the FL being used, and distance to the subject. Under some circumstances, a background will appear very blurred with an aperture of say f/4.5, while under other circumstances it will take f/2.4 or wider, with the subject filling the frame to the same degree. The older prime lenses of wide angle to mid FL or short tele, invariably had a very useful DOF scale right on the lens. Some still do.

Then there's a matter of refinement. Just how much blur is what you are looking for? Sometimes you want enough blur to make your subject stand out, but still have the background quite clearly recognizable. This would establish an environment that may be important to the context of your shot. On the other hand, you may want to make the background virtually disappear.

Compared to film use of olden days, practicing with a DSLR to get a good feel for results under varying conditions, and with various equipment, is much easier.

---

Last edited by mikesbike; 08-17-2017 at 09:48 PM.

[titrisol]

In the old days, my favorite portrat (head-shoulders) was a 105/2.8 for studio and a 200/4 for outdoors.
Yes, both gave me nice blurry backgrounds

I think the DOF - distance - focal length is the same as the focusing scale; in a wide angle (31mm) the distance marks were closer than in a 105mm and thus the separation of fore and background is easier.
FWIW, using wide angles was an easy wy to have a lot of things in focus