M9 easier to focus than M8

[SJP]

Advertisement (gone after registration)

Isn't the circle of confusion set by the pixel size instead of the sensor size.

Since the pixel size did not change, so did the circle of confusion stay the same?

Hans

No the circle of confusion is governed by the sensor size & to what extent you wish to examine your large size print with a microscope e.g. to count all your cat's whiskers. The bog standard CoC on which the lens barrel markings are based is 31 micron for Leica lenses as far as I can compute. That will be the same for the M9. With the M8 you needed to compensate by (very roughly) 1 stop. Most of us add 1 stop extra for reasons of pixel peeping. So for the M9 about 1 stop safety margin would be the operative procedure.

[t024484]

No the circle of confusion is governed by the sensor size & to what extent you wish to examine your large size print with a microscope e.g. to count all your cat's whiskers. The bog standard CoC on which the lens barrel markings are based is 31 micron for Leica lenses as far as I can compute. That will be the same for the M9. With the M8 you needed to compensate by (very roughly) 1 stop. Most of us add 1 stop extra for reasons of pixel peeping. So for the M9 about 1 stop safety margin would be the operative procedure.

 

What about this article,

Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

are they wrong in their assumptions?

 

Hans

[SJP]

What about this article,

Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

are they wrong in their assumptions?

 

Hans

No at a glance the article is just fine. The point is that diffraction is normally much smaller that what we need to make a sharp image (or at least something that appears sharp). For the M* the size of the circle of confusion (required to make an acceptably sharp image) is quite a bit larger than diffraction limits except at f/8 or f/11. So for largish sensors difraction is not really a factor to consider. For small sensors like in P&S, the stuff that 99.99% of the population use, diffraction is a major issue.

[lct]

For a CoC calculator, see Circles of Confusion for Digital Cameras

[jaapv]

No the circle of confusion is governed by the sensor size & to what extent you wish to examine your large size print with a microscope e.g. to count all your cat's whiskers. The bog standard CoC on which the lens barrel markings are based is 31 micron for Leica lenses as far as I can compute. That will be the same for the M9. With the M8 you needed to compensate by (very roughly) 1 stop. Most of us add 1 stop extra for reasons of pixel peeping. So for the M9 about 1 stop safety margin would be the operative procedure.

Yes- that one stop will take the difference in the appearance of DOF between film an sensor into account. We must consider, however, that the 0.3 mm is based on a final print of 6x9 cm. So as long as you don't print larger, you will be fine.

[t024484]

Yes- that one stop will take the difference in the appearance of DOF between film an sensor into account. We must consider, however, that the 0.3 mm is based on a final print of 6x9 cm. So as long as you don't print larger, you will be fine.

 

You probably mean 0.03 mm and 6*9 inch ?

[jaapv]

Advertisement (gone after registration)

Yes, 0.03 mm Thanks for correcting the typo . And no, the original figure is based on 1930-ies 6x9 cm enlargements. It is wildly optimistic for current-day photography. The supposed reason that it was never changed is that lens makers felt that the public would see any new standard as "modern lenses are worse quality".

[t024484]

For a CoC calculator, see Circles of Confusion for Digital Cameras

 

Thanks.

 

I started from the wrong side, that is with the sensor.

One shoulds start with the vision capability of men, which roughly has a CoC of 0.2mm when looking at a picture of 8*10 inch wide from a distance of 25cm.

Any detail below this figure of 0.2mm will we perceived as being sharp.

Going back from there to a 24*36 sensor this translates in approx. 37,5 lp/mm or 1800 LW/PH.

 

The M9 will easily outperforms this figure, having probably something like 2800 LW/PH, or 58 lp/mm, but are FF lenses capable of resolutions this high ?

If so, the sensor will outperform the available lenses, which is not bad.

 

Hans