Diffraction Limits on Canon 5DS of f8 ??

[PureClassA]

http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

If you scroll down this link, they have a nifty diffraction calculator. If you open the advanced tab, you can punch in actual mega pixels. Set you Circle of Confusion to be based on pixels. If you set pixels to 50.6, you will get the approximate pixel size of the Canon 5DS at 4.1 nm.

This seems to illustrate that where on a 5D MkIII, diffraction doesn't mathematically effect things until between f11 and f16 (they only use full stops on this), it shows the 5DS topping out at f8 given the much smaller pixel size.

Just thought this was interesting enough to pass along. I shot landscape at f11 with a 5D3 and 6D commonly. I'm now curious to try it out on my 5DSR. The 6D given it's fewer, bigger pixels, has an even higher diffraction limit than the 5D3. Assuming one needs a longer DOF on a given shot, fewer pixels could actually be better? Sure some of that effect gets mitigated with so much more resolution, but that's pretty interesting to consider nonetheless.

Lots of people on here smarter than I am I'd love to hear opinions on

 

[neuroanatomist]

The 'diffraction limited aperture' isn't a hard limit, it's the aperture at which diffraction begins to impact pixel-level sharpness.

 

[PureClassA]

I understand that part, I just wondered if anyone had so far had any experience with this on the new bodies. I haven't gone past f8 yet, which looked amazing.

That being said, I wonder at some point we hit 75MP or 100MP on a FF camera and diffraction limits start creeping at 5.6 to 8 ...

Have we reached a limit in resolution on FF whereby anything further starts to become effectively irrelevant unless you shoot everything at or bigger than f8?

Would landscape guys shy away or go MF at that point because by f11 - f16, diffraction has grown so much that the increased resolution has become a moot point?

 

[neuroanatomist]

TS-E.

 

[PureClassA]

I have never played with one before, but it's on my bucket list. Does it focus at two different depths to get the huge DOF? Never really understood how those worked.

TS-E.

 

[neuroanatomist]

Tilt moves the focal plane so it's no longer parallel to the sensor. It doesn't give more DoF, but allows you to 'angle' the DoF away from the camera, meaning you can have crisp focus from the legs of your tripod to the distant mountains even with a relatively wide aperture like f/5.6.

 

[PureClassA]

Well then given that Canon is the only kid on the block making those as well as they do, seems that as cameras boost resolution higher, either Nikon and Sony are gonna have to invest in that product line or Canon remains the king of wide angle and landscape for diffraction reasons. I don't shoot a lot of landscape but if I did, I'd be buying the 17 TSE for sure. Don't see the point of having 50+ MP if shooting at f11 - f16 gives you sharpness results arguably no better than a 22MP 5D3. Another reason to appreciate Canon's system

Tilt moves the focal plane so it's no longer parallel to the sensor. It doesn't give more DoF, but allows you to 'angle' the DoF away from the camera, meaning you can have crisp focus from the legs of your tripod to the distant mountains even with a relatively wide aperture like f/5.6.

 

[Zeidora]

If you set pixels to 50.6, you will get the approximate pixel size of the Canon 5DS at 4.1 nm.

Should read 4.1 µm (micrometer) not 4.1 nm (nano meter). A minor difference of three orders of magnitude (aka factor 1000x).

Just got my 5dsr in. Will be fun playing around and comparing z-stacking with MPE-65 (not on the Canon recommended lens list for 5ds ) from 1:1 to 7:1 with 1.4x TC.

 

[9VIII]

Brian at TDP shows moire being eliminated at f11.
So it's not quite f8 for full effect, but in a way I like it that way so I can easily guarantee clean Macro images.
Also, the AA filter doesn't stop aliasing completely either.

http://www.the-digital-picture.com/Reviews/Canon-EOS-5Ds-R.aspx

 

[PureClassA]

Yes... a typo. apologies.

If you set pixels to 50.6, you will get the approximate pixel size of the Canon 5DS at 4.1 nm.

Should read 4.1 µm (micrometer) not 4.1 nm (nano meter). A minor difference of three orders of magnitude (aka factor 1000x).

Just got my 5dsr in. Will be fun playing around and comparing z-stacking with MPE-65 (not on the Canon recommended lens list for 5ds ) from 1:1 to 7:1 with 1.4x TC.

 

[photon]

I found Roger's article interesting regarding this
http://www.lensrentals.com/blog/2013/03/overcoming-my-fentekaphobia

 

[Mt Spokane Photography]

Don't worry about diffraction, it will still be sharper at f/11 than with a 5D MK III. Once you get to f/16 or f/22, it might be equal, but not worse.

 

[Zeidora]

I got around to doing the MPE-65 experiments. See here for thread
http://www.canonrumors.com/forum/index.php?topic=27058.msg535038#msg535038
Bottom line, at 4:1 and effective f-stop 14 I reach a limit of getting more detail.