APS-C Lens on Full Frame Camera - DoF Question

[Guest]

Advertisement (gone after registration)

9 minutes ago, jaapv said:

 

What isn't straight here?

The question was "1.4 or 2.0", so the answer has to be one of those!

[Guest]

1 hour ago, lct said:

You'll get a straight answer if i understand your question. Must be my poor english, never mind... Just to be clear: 
1. What do you mean by a 16/1.4 lens? A lens made for FF or APS-C camera?
2. What do you mean by 16mm? The actuel focal length of the lens or anything else (so-called "equivalent" things)?
3. What do you mean by f/1.4? The actual aperture of the lens or anything else (so-called "equivalent" things)?
4. Do you intend to use the same lens on both FF and APS bodies or two different lenses?

I meant the actual focal length and actual aperture - anything else I would have qualified. Just on FF. For a specific example - Sigma 16mm f1.4 (an APS-C lens) on an SL2.

Edited August 5, 2020 by Guest

[Guest]

... and I've realised I'd missed the crucial "APS-C" qualifier on the lens in my first post. Apologies. I was assuming that being in the CL etc forum would be clear enough, but with hindsight that was a bad assumption. Thanks for all your input, I think we're there now (the answer is 2.0 ).

[Guest]

40 minutes ago, jaapv said:

Light gathering is misleading nonsense; forget about it and certainly in this context. A lens is a certain speed, say 1.4, and will remain that speed regarding exposure in all circumstances, the DOF can only change if you change the sensor format (which not when using a cropped SL(2) ), change the magnification throughout the system or change subject distance and/or aperture. In other words: Use a TL lens on the SL as you would use it on a CL and you will get exactly the same results in all respects (except pixel count). Basically you are using an APSC camera and throwing away the rest of the sensor surface.

My lack of understanding here was perhaps because I don't use a CL, or indeed any APS-C camera, so saying "it works the same as the CL" doesn't help me! I am considering using an APS-C lens on a FF body though, mainly for size and cost reasons ... but I want to be clear on what DoF I'll be gaining or losing, compared to a full frame lens of the same aperture.

Edited August 5, 2020 by Guest

[jaapv]

You won't be losing or gaining any DOF, it will be the one the lens was designed for as it will be used on its native sensor size. Obviously you should not compare to a full-frame lens on a full-frame sensor; that will have a one-stop more shallow DOF (approx).  When using a DOF calculator which is a good idea, you need to use the values for APS-C.

[Guest]

Thanks. It might be obvious not to compare to a FF lens, but that's exactly what I'm doing, since I'm trying to decide between the two options. It's good to know for sure that DoF is different, and thus part of the compromise equation (24mm at f2 is still plenty fast and bright of course).

[lct]

Advertisement (gone after registration)

33 minutes ago, AndyGarton said:

I meant the actual focal length and actual aperture - anything else I would have qualified. Just on FF. For a specific example - Sigma 16mm f1.4 (an APS-C lens) on an SL2.

I have no experience with the SL2 but being an APS-C lens, the SL2 should crop and behave exactly like the CL with it. So (not so ) straight answer: Nothing will change. Same focal length (16mm), same aperture (1.4), same DoF and nothing "gathering" or "equivalent" at all .

[Guest]

That's definitely not right, I know for sure the effective focal length (field of view) changes.

Edited August 5, 2020 by Guest

[lct]

3 minutes ago, AndyGarton said:

That's definitely not right, I know for sure the effective focal length (field of view) changes.

Forget those effective things they can only confuse the matter. If the SL2 crops only a cropped portion of the sensor will be used. Then the SL2 will behave exactly like the CL.

[adan]

The rule of thumb - far better known 15 years ago when almost everything was APS-C crop - is:

Multiply actual aperture setting (i.e. f/1.4 on the dial) by the crop factor (i.e. 1.5x)

Which is f/2.1, so the previous mention of f/2.0 is "close enough for government work."

That assumes different-but-equivalent focal lengths (e.g. 16mm f/1.4 on APS-C will look like 24mm f/2.1 on 24x36).

And remembering that "DoF" is always "apparent" and will change with the amount of enlargement - something that "looks sharp" in a small print may be revealed as fuzzy in a big print, or at 100% pixel peeping on a computer screen.

What lct says is correct in that a lens projects a fixed image - that does not change, regardless of the sensor behind it (or even if there is no sensor at all - a "virtual" image projected onto thin air). The light gets bent a certain way, and that is it as far as the lens is concerned. The lens does not "know" what sensor is behind it.

But - an APS image has to be enlarged 50% more than a 24x36 image to get the same print size. I.E. to get  a 24cm x 36cm print from a 24 x 36 original, the enlargement must be 10x - and to get the same print from a 16mm x 24mm APS original requires a 15x enlargement. And thus may reveal more things that appear fuzzy (or fuzzier) due to the higher magnification required.

The rule of thumb takes that into account.

A little experience with the many various film formats makes this intuitive - eventually.

A 90mm lens is: a "normal" field of view on a 6x6 camera, and a long lens on a 35mm Leica, and a wide-angle on a 4x5 view camera.

Put a view-camera 90mm lens on a 35mm camera and it produces the same image as on the 4x5 - just cropped to a narrower view and a smaller part of the image. And it will be the same image produced by a "native" 90mm designed for that 35mm camera.

Edited August 5, 2020 by adan

[jaapv]

14 minutes ago, AndyGarton said:

That's definitely not right, I know for sure the effective focal length (field of view) changes.

The focal length is a lens property and not depending on anything in the camera, including the sensor size. The concept of "effective focal length" is intended for framing purposes only.

[lct]

10 minutes ago, adan said:

[...] What lct says is correct in that [...]

Don't count me in that much please. We're not comparing an FF camera to an APS-C one here but two APS-C cameras if the SL2 crops as i suspected above. Then there is zero "equivalence", only a total and perfect draw .

[jaapv]

And please don't conflate all kinds (including Internet doubletalk)  of concepts that have very little to nothing to do with the original question.

[adan]

Well, what a particular camera chooses to do regarding format has nothing to do with optical theory.

The Canon EF-S "APS lenses" can't even be mounted onto the 24x36 EF bodies, whereas Nikon FX cameras will allow their DX (APS) lenses to be mounted, and usually allow one to either auto-crop - or not. A menu choice.

Edited August 5, 2020 by adan

[lct]

20 minutes ago, adan said:

Well, what a particular camera chooses to do regarding format has nothing to do with optical theory.

The Canon EF-S "APS lenses" can't even be mounted onto the 24x36 EF bodies, whereas Nikon FX cameras will allow their DX (APS) lenses to be mounted, and usually allow one to either auto-crop - or not. A menu choice.

I beg to differ. If the camera crops it uses only a portion of the sensor. Then the latter will have the same area as that of a native APS-C camera. The CoC value will then be the same and if aperture and focal length are also the same, DoF is a perfect draw.

[Guest]

To be honest I'm not that interested in optical theory - to say that a 16mm APS-C lens is still a 16mm lens when mounted on a FF body (or indeed an APS-C camera like the CL) isn't helpful to me, because all I care about is the results when taking pictures. Because I've shot 35mm format cameras for many years, that's the format I "think" in. That 16mm lens is a 24mm lens as far as framing is concerned, and that's what matters to me.

[jaapv]

10 minutes ago, lct said:

I beg to differ. If the camera crops it uses only a portion of the sensor. Then the latter will have the same area as that of a native APS-C camera. The CoC value will then be the same and if aperture and focal length are also the same, DoF is a perfect draw.

Exactly!

 

[jaapv]

8 minutes ago, AndyGarton said:

because all I care about is the results when taking pictures

That is precisely the reason that you should not burden yourself with the theoretical DOF values. DOF  is a subjective artistic tool that connects or separates layers in a composition. It is dependent on contrast, busy or simple forms, perception of photographer and viewer and, ultimately on print size and viewing distance.  The values assigned mathematically are irrelevant, the way that the photographer uses it paramount. Do you think that Rembrandt measured the precise size and elasticity of a brush he needed for a particular part of his paintings? I bet not, he chose out of his experience, and he was a technique artist. So just use the lens as you find it with the knowledge of your technique.

[jankap]

I would recommend the smartphone app called "Photopills", this consists of an almost complete database of cameras. Almost, because my Tessina and my Lytro Illum are missing. But focal reducer and tele-extender can be chosen.

DOF is completely based on mathematics. There still is the flavor of what is sharp of course, but in this case this is not important.

M10 (24mp) 50mm at f/11 and distance 7meter gives a DOF of 3,60 to 117 meter

M10 (24mp) 75mm at f/11 and distance 7meter gives a DOF of 5,94 to 12 meter

CL (24mp) 50mm at f/11 and distance 7meter gives a DOF of 4,32 to18,38 meter

NEX-6 (16mp) 50mm at f/11 and distance 7meter gives a DOF of 4,33 to18,23 meter

The 50mm focal length, because my Elmar LTM is on my camera at the moment.

 

Edited August 5, 2020 by jankap

[Advertisement]

Hallo,

Look here: Leica L-Mount

[Guest]

I don't agree to be honest Jaap, but that's ok! After forty years of using 35mm "full frame" lenses, I have muscle memory for what f1.4 DoF is like on various focal lengths, and my current lens selection reflects that. I shoot wide open most of the time, so a "genuine" f1.4 would be very nice to have, but of course f2.0 is still good, and a fine compromise when you consider the other benefits (size, weight, cost).

Edited August 5, 2020 by Guest