Size of Circle of Confusion for Hyperfocal Distance

Moderator: jsachs

Post Reply
tomczak
Posts: 1367
Joined: April 25th, 2009, 12:56 am
What is the make/model of your primary camera?: Fuji X-E2
Contact:

Size of Circle of Confusion for Hyperfocal Distance

Post by tomczak »

What would be a reasonable size of a circle of confusion, given the camera sensor size/res. and image use?

Fuji apparently uses 0.02mm for X cameras (16MP ?), but how did they come up with this value?
Maciej Tomczak
Phototramp.com
jsachs
Posts: 4203
Joined: January 22nd, 2009, 11:03 pm

Re: Size of Circle of Confusion for Hyperfocal Distance

Post by jsachs »

There is no "correct" value for the CoC. The conventional value used by most full frame cameras to mark DoF on lens distance scales is around 0.03mm which is based on an image appearing sharp when printed at a size of about 8x10 and viewed from a little over a foot away by someone with average vision. Note that this totally ignores the sensor resolution. If, on the other hand, you want your images to look as sharp as possible when viewed on-screen at 1:1, you need to use a somewhat smaller value for the CoC. Most of the time -- at least in landscape photography -- you need to either compromise on DoF or use focus stacking.

If you download my new DoF calculator and the accompanying manual, this topic is discussed in sickening detail and you can enter in values for the sensor size and resolution and to calculate CoC using either the sharp print or sharp image method.

www.dl-c.com/DoF
Jonathan Sachs
Digital Light & Color
tomczak
Posts: 1367
Joined: April 25th, 2009, 12:56 am
What is the make/model of your primary camera?: Fuji X-E2
Contact:

Re: Size of Circle of Confusion for Hyperfocal Distance

Post by tomczak »

Cheers!

Fuji X-E2 (and other 16MP X's) have some 0.005 mm nominal pixel pitch. So the blob of 0.02 mm diameter would cover some 16 of them? Why is that considered sharp, in one sentence?
Maciej Tomczak
Phototramp.com
jsachs
Posts: 4203
Joined: January 22nd, 2009, 11:03 pm

Re: Size of Circle of Confusion for Hyperfocal Distance

Post by jsachs »

It's bad if you want your images to look sharp when you zoom in on them or you plan print them large and view them up close. It's adequate if all you want to do is print 8x10's that you view from about a foot away.

DoF is defined as the range of subject distances for which the image is "acceptably" sharp. Ignoring diffraction, imperfect optics, subject and camera motion, etc., the CoC specifies how sharp is acceptably sharp. Generally a spread across 2 pixels is considered about the limit of sharpness for a sensor due to Bayer dither, and other factors so a spread across 8 pixels is nowhere near as sharp an image as the camera can take -- but the camera is better than it needs to be to make 8x10 enlargements. A similar situation has existed for years with film cameras as film resolutions increased and lenses got better.
Jonathan Sachs
Digital Light & Color
Dieter Mayr
Posts: 453
Joined: April 24th, 2009, 11:47 am
What is the make/model of your primary camera?: Nikon D700
Location: Salzburg / Austria

Re: Size of Circle of Confusion for Hyperfocal Distance

Post by Dieter Mayr »

Maciej,

Sorry for the late answer, i must have missed your post.
The Fuji X has, from what I found, a sensor size of 23,6 x 15,6 mm, giving a diagonal of 28,29 mm (Square root of (23,6^2 + 15,6^2))
To view the total printed image, the viewing distance is about the diagonal of the image, resulting in a viewing angle of roughly 50° on the diagonal.
The smallest detail the human eye can separate is about 2 minutes of arc resolution.
50° are 3000 minutes of arc, dividing that by the resolution of the eye 2 minutes of arc gives 1/1500 the of the diagonal as maximal viewable resolution for a image.
So these 1/1500 the of the diagonal are standard for the circle of confusion, 28,29 / 1500 gives 0,019 mm for the circle of confusion.

If the image is viewed from a closer distance and the image should still appear sharp the maximum allowed circle of confusion has to be smaller accordingly.
Dieter Mayr
Post Reply