6D MkII, Diffraction Correction

[Marek M]

I've read about it a bit but still am unsure what the function of "Diffraction Correction" setting does in practice.

Could someone who knows explain, please?

 

[SalmoTrutta]

The diffraction error that reaches the sensor will be different for every camera - lens pairing depending on the aperture blade shape and the low pass filter setup.

I don't know if they model or measure the pattern of the error, but whichever it is they can then apply a profile that adds or subtracts what the diffraction subtracted or added for each pixel of the final image.

 

[Marek M]

The diffraction error that reaches the sensor will be different for every camera - lens pairing depending on the aperture blade shape and the low pass filter setup.

I don't know if they model or measure the pattern of the error, but whichever it is they can then apply a profile that adds or subtracts what the diffraction subtracted or added for each pixel of the final image.

What I am really asking about, is the effect on the picture. What difference will/should there be?

 

[MitchAlsup]

I've read about it a bit but still am unsure what the function of "Diffraction Correction" setting does in practice.

Could someone who knows explain, please?

Given some information about the lens and the known quantity of a given sensor, there is a sharpening function which can sharpen the diffraction out of an image. This is used to sharpen images taken at F/stops from F/8 and slower through whatever the lens can manage.

If it works, it will be subtle like a gentle hand on the sharpen controls of DPP or photoshop.

 

[Marek M]

I've read about it a bit but still am unsure what the function of "Diffraction Correction" setting does in practice.

Could someone who knows explain, please?

Given some information about the lens and the known quantity of a given sensor, there is a sharpening function which can sharpen the diffraction out of an image. This is used to sharpen images taken at F/stops from F/8 and slower through whatever the lens can manage.

If it works, it will be subtle like a gentle hand on the sharpen controls of DPP or photoshop.

Thank you for the explanation.

At the risk of sounding ignorant, I hardly ever get to f8 and above , would it then mean that keeping "Diffraction correction" on be irrelevant?

 

[Kaso]

I hardly ever get to f8 and above , would it then mean that keeping "Diffraction correction" on be irrelevant?

Just set it to OFF (factory default)

 

[apersson850]

Just set it to "ON". If you use an aperture where diffraction doesn't make much of a difference, the function will not destroy anything.

 

[brightcolours]

The diffraction error that reaches the sensor will be different for every camera - lens pairing depending on the aperture blade shape and the low pass filter setup.

Yeah, not that low pass filter set up. Just the lens, and for how they want to correct it the pixel pitch.

I don't know if they model or measure the pattern of the error, but whichever it is they can then apply a profile that adds or subtracts what the diffraction subtracted or added for each pixel of the final image.

 

[brightcolours]

I've read about it a bit but still am unsure what the function of "Diffraction Correction" setting does in practice.

Could someone who knows explain, please?

Diifraction correction is also known as "deconvolution". With mathematical models it is possible to recreate some detail that appears to be lost in soft areas, softness resulting from diffraction of light for instance.

If you shoot RAW, the setting has no influence, It is only for JPEG out of camera results. When you shoot RAW and still would like to use the deconvolution, you can use Canon's DPP software for converting the RAW.

 

[whakapu]

Diffraction limited aperture for your 6DII is f9.3. That's the aperture at which stopping down no longer produces a sharper image and begins to soften the image, but diffraction should still be negligible at f11. It shouldn't really matter whether you leave it on or off unless you want to shoot f16 or f22.

 

[brightcolours]

Diffraction limited aperture for your 6DII is f9.3.

No it isn't. How did you come by that f9.3 figure? Different wavelengths of light diffract differently. You already can measure lesser sharpness with normal light past f4 with a 6D, so also with a 6D mk II which has a smaller pixel pitch.

That's the aperture at which stopping down no longer produces a sharper image and begins to soften the image, but diffraction should still be negligible at f11. It shouldn't really matter whether you leave it on or off unless you want to shoot f16 or f22.

 

[Mako2011]

Diffraction limited aperture for your 6DII is f9.3.

No it isn't. How did you come by that f9.3 figure? Different wavelengths of light diffract differently.

The Diffraction Limit: point at which selecting a smaller aperture (all else being equal) starts to diminish sharpness by minimizing lens aberrations.

When using an online calculator (or the formula directly) to determine the Diffraction Limit...most assume light in the middle of the visible spectrum and use a wavelength of ~550 nm. Given that, the f9.3 number seems about right with a 10 inch print and close viewing distance (inside 20cm or so). Change the Viewing distance and/or print size and limit changes...of course

 

[brightcolours]

Diffraction limited aperture for your 6DII is f9.3.

No it isn't. How did you come by that f9.3 figure? Different wavelengths of light diffract differently.

The Diffraction Limit: point at which selecting a smaller aperture (all else being equal) starts to diminish sharpness by minimizing lens aberrations.

So far, no nonsense.

When using an online calculator (or the formula directly) to determine the Diffraction Limit...most assume light in the middle of the visible spectrum and use a wavelength of ~550 nm. Given that, the f9.3 number seems about right with a 10 inch print and close viewing distance (inside 20cm or so). Change the Viewing distance and/or print size and limit changes...of course

DOF is expressed with viewing distance and print size in mind.

Diffraction softening is... not.

We can easily observe that diffraction softening starts to limit sharpness past f4 by looking at MTF test data, with a ~20mp FF sensor.

These MTF tests are done with a normal "white" light spectrum and using a normal RBG imaging sensor with Bayer CFA.

 

[Mako2011]

Diffraction limited aperture for your 6DII is f9.3.

No it isn't. How did you come by that f9.3 figure? Different wavelengths of light diffract differently.

The Diffraction Limit: point at which selecting a smaller aperture (all else being equal) starts to diminish sharpness by minimizing lens aberrations.

So far, no nonsense.

When using an online calculator (or the formula directly) to determine the Diffraction Limit...most assume light in the middle of the visible spectrum and use a wavelength of ~550 nm. Given that, the f9.3 number seems about right with a 10 inch print and close viewing distance (inside 20cm or so). Change the Viewing distance and/or print size and limit changes...of course

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

These MTF tests are done with a normal "white" light spectrum and using a normal RBG imaging sensor with Bayer CFA.

You asked were the "Diffraction limited aperture for your 6DII is f9.3" came from. I explained it as it relates to the normal Diffraction Limit Calculation most use. MTF tests are a bit apples to oranges regards the math as they use a different set of calculations and variables to derive the MTF

 

[brightcolours]

Diffraction limited aperture for your 6DII is f9.3.

No it isn't. How did you come by that f9.3 figure? Different wavelengths of light diffract differently.

The Diffraction Limit: point at which selecting a smaller aperture (all else being equal) starts to diminish sharpness by minimizing lens aberrations.

So far, no nonsense.

When using an online calculator (or the formula directly) to determine the Diffraction Limit...most assume light in the middle of the visible spectrum and use a wavelength of ~550 nm. Given that, the f9.3 number seems about right with a 10 inch print and close viewing distance (inside 20cm or so). Change the Viewing distance and/or print size and limit changes...of course

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

No, that is not true. CoC related to DOF indeed is viewing distance and print size dependent.

The CoC related to diffraction has nothing to do with the CoC of DOF calculations. Tow different circle of confusion definitions.

These MTF tests are done with a normal "white" light spectrum and using a normal RBG imaging sensor with Bayer CFA.

You asked were the "Diffraction limited aperture for your 6DII is f9.3" came from.

And explained that that is incorrect.

I explained it as it relates to the normal Diffraction Limit Calculation most use. MTF tests are a bit apples to oranges regards the math as they use a different set of calculations and variables to derive the MTF

 

[Mako2011]

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

No, that is not true. CoC related to DOF indeed is viewing distance and print size dependent.

Look at the formulas for Diffraction Limit at ~550 nm. You will find no DoF variable

Examples:

diffraction limit depends on the pinhole radius, ro. It is defined by the angle Δθ = t0λ/2πro where t0 = 3.832

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

 

[brightcolours]

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

No, that is not true. CoC related to DOF indeed is viewing distance and print size dependent.

Look at the formulas for Diffraction Limit at ~550 nm. You will find no DoF variable

You are the one confusing DOF and diffraction, not me.

Examples:

diffraction limit depends on the pinhole radius, ro. It is defined by the angle Δθ = t0λ/2πro where t0 = 3.832

 

[Mako2011]

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

No, that is not true. CoC related to DOF indeed is viewing distance and print size dependent.

Look at the formulas for Diffraction Limit at ~550 nm. You will find no DoF variable

You are the one confusing DOF and diffraction, not me.

I never brought up DoF and as you can see...it is not a variable in the calculation of Diffraction Limit

Examples:

diffraction limit depends on the pinhole radius, ro. It is defined by the angle Δθ = t0λ/2πro where t0 = 3.832

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

 

[brightcolours]

DOF is expressed with viewing distance and print size in mind.

DoF is not a variable in the Diffraction Limit calculation...but viewing distance and print size are as it relates to CoC

No, that is not true. CoC related to DOF indeed is viewing distance and print size dependent.

Look at the formulas for Diffraction Limit at ~550 nm. You will find no DoF variable

You are the one confusing DOF and diffraction, not me.

I never brought up DoF and as you can see...it is not a variable in the calculation of Diffraction Limit

You brought up print size and viewing distance, which ARE about DOF and not about diffraction.

So, yes, you brought up DOF indirectly, and I tried to explain to you that print size and viewing distance are DOF related, not diffraction related.

So, you confuse DOF and diffraction here.

Examples:

diffraction limit depends on the pinhole radius, ro. It is defined by the angle Δθ = t0λ/2πro where t0 = 3.832

 

[Mako2011]

I never brought up DoF and as you can see...it is not a variable in the calculation of Diffraction Limit

You brought up print size and viewing distance,

Yes as they are variables in the Diffraction Limit calculation

which ARE about DOF and not about diffraction.

and DoF is not a variable in that calculation. Apples and oranges. As an example, subject distance is not needed to calculate the Diffraction Limit

Examples:

diffraction limit depends on the pinhole radius, ro. It is defined by the angle Δθ = t0λ/2πro where t0 = 3.832

we are speaking about how the Diffraction Limit is derived (f9,3 in this case)...not how the DoF number is derived. As an example, subject distance is not needed to calculate the Diffraction Limit

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)