Ad
Ad
Ad
Pages: « 1 [2]   Bottom of Page
Print
Author Topic: Focusing in the Digital Era  (Read 10374 times)
NikosR
Sr. Member
****
Offline Offline

Posts: 622


WWW
« Reply #20 on: October 25, 2006, 12:54:59 PM »
ReplyReply

Quote
Nikos, DoF calculations for a digital sensor are not particularly different from DoF calculations for film.  In both cases, as Howie likes to emphasize, DoF is a property of the optics and not the sensor (provided the sensor has enough pixels or fine enough grain to resolve the optical circle of confusion).  As you guessed, the pixel pitch provides the connection between pixels and distance for the digital sensor.

If you want equations:

CoC (print) = CoC (sensor) x Magnification

CoC (print) is the blur size on the final print.  The largest acceptable CoC is determined by factors like print size, viewing distance, and visual acuity.
CoC (sensor) is the blur size on the sensor.  This is determined by optical factors like distance to subject, lens focal length, and lens aperture of f/stop.

Magnification is the degree of enlargement needed to go from sensor to print.  What does magnification mean for digital sensors?  Let's assume one pixel on the sensor is printed as one dot on the print.  (If the digital file is resized upwards or downwards before printing, you can include this as an extra magnification factor, but I will ignore it here.)  Magnification is then the ratio of the print size to the sensor size, which is also the ratio of the size of a dot on the print to the size of a pixel on the sensor.  An example will probably make this clearer than words.

Example:
     8 inch print at 300 dots/inch from
     2400 pixels on sensor with 8 micron pixel pitch

Here are two equivalent ways to calculate magnification:

Magnification = print size / sensor size
     = (8 inch) / (2400 x 8 micron) ~ (200mm) / (19.2mm) ~ 10

Magnification = print dot size / sensor pixel size
     = (1/300 inch) / (8 micron) ~ (85 micron) / (8 micron) ~ 10

Why does the physical size of the pixel matter, not just the pixel count of the digital file?  It comes back to the point that the CoC on the sensor is a physical quantity, with actual dimensions in microns not pixels.
[a href=\"index.php?act=findpost&pid=82228\"][{POST_SNAPBACK}][/a]

Thank you Eric. You expressed quite succintly the fact that its pixel no AND pixel pitch that should determine the CoC in digital.
In film of course, assuming same emulsion, 'pixel pitch' so to speak is constant, thus we only have the size (= 'pixel no') to play with.

All this of course does not bode well with the article's in question assertion that smaller sensors support 'less enlargement' which was my original complaint.
« Last Edit: October 25, 2006, 01:33:07 PM by NikosR » Logged

Nikos
EricV
Full Member
***
Offline Offline

Posts: 136


« Reply #21 on: October 25, 2006, 01:48:35 PM »
ReplyReply

Quote
So what I said in a previous post about same sized but different Mp Number sensors having different CoC seems to be indeed true. This fact is not reflected in any DoF calculator that I know of.
[a href=\"index.php?act=findpost&pid=82231\"][{POST_SNAPBACK}][/a]

Hmm, sounds like there is still some confusion .... let me try again.

The CoC on the sensor is determined by the optics, not the sensor.  Two different sensors, with different pixel pitch (same physical size but different pixel count), will have the same CoC, expressed in physical units like microns, if the same optics are used to form the image.  This is intuitive for film cameras (film granularity has nothing to do with CoC) and the same principle applies for digital cameras (pixel density has nothing to do with CoC).  

Of course if you measure the CoC in pixels instead of microns, then pixel pitch enters the equation.  But this is a backwards way of thinking, sort of like a film photographer thinking "my film has a grain size of 8 microns and my image occupies 2400 grains of film" rather than "my image occupies 19mm of film".  When calculating CoC, the second way of thinking seems simpler, and it emphasizes that pixel pitch or film grain is irrelevant.

DoF calculators require the user to enter phisical quantities like image size.  Given just the pixel count of a digital file (like 6MB = 2000x3000 pixels) and no information on physical image size or pixel pitch, there is not enough information for the calculation.  The DoF calculator can calculate how large a CoC is formed by a particular lens, but has no way of translating this into pixels without knowing the phisical pixel size or sensor size.
Logged
dlashier
Sr. Member
****
Offline Offline

Posts: 518



WWW
« Reply #22 on: October 25, 2006, 01:59:26 PM »
ReplyReply

While I agree with the basic principle of the article, I think this issue is over-exaggerated. First, with digital landscapes and "front to back" DOF I don't usually have a problem when printing at A3 and don't find the issue much different than film days. Second, I would dispute the fact that typical viewing size was 4x6 in film days. Maybe for snapshot drugstore prints, but serious amateurs like myself and even my dad typically printed 8x10 or larger, and in fact in my family the habit was to shoot chrome and project viewing at 40 inches. Bottom line - I don't really see where digital has significantly changed DOF issues from film days except possibly for those who never used to print larger than snapshot (4x6) size.

- DL
« Last Edit: October 25, 2006, 01:59:58 PM by dlashier » Logged

Nemo
Sr. Member
****
Offline Offline

Posts: 276



« Reply #23 on: October 25, 2006, 02:47:31 PM »
ReplyReply

Quote
As I have already exlained, I fail to see the direct relationship between sensor size and CoC. Asserting that a direct relationship exists (similar to the film case) would entail that one accepts the concept of 'enlarging' the sensor dimensions to the final output to be judged (say 8x10).

Well, it is easy to see. The sensor reproduces some real detail. Lets say, a branch in a distant tree. It is reproduced in a 24x36 surface. That miniaturized reproduction of the branch must be reproduced in the print. If you want a 8x12 print you need to enlarge that reproduction of the branch 8,5 times. It has nothing to do with number of pixels. You have the same number of pixels in a A4 or A3 print, like you had the same number of grains "projected" from the film to the paper. They point is you have a reproduction, an image, projected by the lens on the focal plane (the branch of the tree). That image is capture by something (it doesn't matter if film or sensor), but the image that the lens draws has physical dimensions, and it must be enlarged. The branch or the face of a person, including all the details captured, are multiplied by 8,5 times for a 8x12 (A4) print.
Logged
NikosR
Sr. Member
****
Offline Offline

Posts: 622


WWW
« Reply #24 on: October 25, 2006, 03:05:28 PM »
ReplyReply

Quote
Hmm, sounds like there is still some confusion .... let me try again.

The CoC on the sensor is determined by the optics, not the sensor.  Two different sensors, with different pixel pitch (same physical size but different pixel count), will have the same CoC, expressed in physical units like microns, if the same optics are used to form the image.  This is intuitive for film cameras (film granularity has nothing to do with CoC) and the same principle applies for digital cameras (pixel density has nothing to do with CoC). 

Well, this is were I tend to disagree. I can see the CoC being independent of the medium's resolution, but not the CoC diameter limit that is used in DoF calculations. And so do lots of other people like Norman Koren. But, it is too late over here for me to think harder and I'm off to bed.

Another DoF discussion was not what I had in mind with my original post.
Logged

Nikos
NikosR
Sr. Member
****
Offline Offline

Posts: 622


WWW
« Reply #25 on: October 25, 2006, 03:26:50 PM »
ReplyReply

Quote
Well, it is easy to see. The sensor reproduces some real detail. Lets say, a branch in a distant tree. It is reproduced in a 24x36 surface. That miniaturized reproduction of the branch must be reproduced in the print. If you want a 8x12 print you need to enlarge that reproduction of the branch 8,5 times. It has nothing to do with number of pixels. You have the same number of pixels in a A4 or A3 print, like you had the same number of grains "projected" from the film to the paper. They point is you have a reproduction, an image, projected by the lens on the focal plane (the branch of the tree). That image is capture by something (it doesn't matter if film or sensor), but the image that the lens draws has physical dimensions, and it must be enlarged. The branch or the face of a person, including all the details captured, are multiplied by 8,5 times for a 8x12 (A4) print.
[a href=\"index.php?act=findpost&pid=82246\"][{POST_SNAPBACK}][/a]

 A larger sensor with less Mp according to your assertion will have to be 'enlarged' less but will it produce a better enlargement than a smaller sensor with more Mps?

Even more importantly, a larger sensor with the same no. of Mp as a smaller sensor will produce a better print because the degree of 'enlargement' is less? (Noise and such issues not withstanding)

I just find it hard to buy this.
Logged

Nikos
Nemo
Sr. Member
****
Offline Offline

Posts: 276



« Reply #26 on: October 25, 2006, 03:28:07 PM »
ReplyReply

8,5 is the ratio of the diagonals of the 24x36mm frame and the 8x12 paper. Whether you want a bigger print, you can enlarge the "projected" image (by the lens) as much as you want. The image projected by the lens has a particular physical size, limited by the circle of light. Enlarging that image (capture by any medium) you distribute the capture details in a bigger surface.
The human eyer does not resolve much detail. Therefore, things that the eye cannot distinguish when printed at small sizes, are easily distinguishable when printed at bigger sizes.
The focus plane is just a plane. Anything outside that plane is out of focus. But cannot see this if the print is small enough.
Logged
Nemo
Sr. Member
****
Offline Offline

Posts: 276



« Reply #27 on: October 25, 2006, 03:31:47 PM »
ReplyReply

More pixels allows for more detail to be captured from the lens, and this allows for better quality (enlarged) pictures, but it does not affect the enlargement phenomenon. Any detail the camera has captured is preserved when transferred to paper, but depending on the size of the paper that detail is spread more or less. If you spread the details too much, the human eye will see the errors, differences in focus, etc. That is the point.
Logged
Stephen Best
Guest
« Reply #28 on: October 25, 2006, 06:41:57 PM »
ReplyReply

That depth of field markings (and hence CoC targets) for film aren't applicable to new sensor sizes/densities and the required degree of enlargement is obvious ... the same way that (arbitrary) CoC targets were different for different film sizes.

A better reference for all this is here:

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

(These Cambridge in Colour articles are excellent.) If you skip down to the Diffraction Limit Calculator halfway down the page, you can calculate both the CoC and Airy Disc for a given print resolution (dimensions, viewing distance and eyesight), sensor (dimensions, megapixels) and aperture. All this is geometry. When reproducing 3D subjects, the CoC is generally the overriding determinant for "resolution" in the print ... unless it's exceeded by the pixel size or diffraction. Note the diminishing return for greater pixel densities in real world situations where the pixel size is dwarfed by CoC/diffraction (optics).
« Last Edit: October 26, 2006, 01:15:39 AM by Stephen Best » Logged
AdrianW
Newbie
*
Offline Offline

Posts: 40


« Reply #29 on: October 26, 2006, 05:54:03 AM »
ReplyReply

Is it me or are the 100% crops of the house labelled wrongly? Surely the OOF one should be Infinity @ f/2.8, and not f/8.

To those who don't see the problem, are you using hyperfocal caculations for working out DOF, or are you just shooting @ f/16 anyway? What size are you enlarging to? If you're not using hyperfocal calculations, or making massive enlargements - it's all moot.

The article makes sense to me, and I think it raises a valid point - the crops illustrate his point nicely IMO.
Logged
jashley
Jr. Member
**
Offline Offline

Posts: 53


« Reply #30 on: October 26, 2006, 10:16:38 AM »
ReplyReply

Quote
Is it me or are the 100% crops of the house labelled wrongly? Surely the OOF one should be Infinity @ f/2.8, and not f/8.

To those who don't see the problem, are you using hyperfocal caculations for working out DOF, or are you just shooting @ f/16 anyway? What size are you enlarging to? If you're not using hyperfocal calculations, or making massive enlargements - it's all moot.

The article makes sense to me, and I think it raises a valid point - the crops illustrate his point nicely IMO.
[a href=\"index.php?act=findpost&pid=82332\"][{POST_SNAPBACK}][/a]
Logged
jashley
Jr. Member
**
Offline Offline

Posts: 53


« Reply #31 on: October 26, 2006, 10:27:40 AM »
ReplyReply

Quote
Is it me or are the 100% crops of the house labelled wrongly? Surely the OOF one should be Infinity @ f/2.8, and not f/8.

To those who don't see the problem, are you using hyperfocal caculations for working out DOF, or are you just shooting @ f/16 anyway? What size are you enlarging to? If you're not using hyperfocal calculations, or making massive enlargements - it's all moot.

The article makes sense to me, and I think it raises a valid point - the crops illustrate his point nicely IMO.
[a href=\"index.php?act=findpost&pid=82332\"][{POST_SNAPBACK}][/a]

Oops, a little fat-fingering there.

The point I was trying to make before is that I haven't changed my approach to focusing at all since going digital, and it still "works", whether I'm going hyperfocal or not.  I'm able to enlarge more because of the absence of film grain, but I haven't found DOF issues to be a limiting factor.  For me (and my 1Ds) this just hasn't been an issue.  Perhaps it's an issue with MF but not 35mm digital?
Logged
EricV
Full Member
***
Offline Offline

Posts: 136


« Reply #32 on: October 26, 2006, 11:10:20 AM »
ReplyReply

Quote
Is it me or are the 100% crops of the house labelled wrongly? Surely the OOF one should be Infinity @ f/2.8, and not f/8.
Read the text again carefully.  The shots were all taken at f/8; only focus was varied.  The image labeled f/8 was taken at the closest focus setting.
Logged
NikosR
Sr. Member
****
Offline Offline

Posts: 622


WWW
« Reply #33 on: October 26, 2006, 11:22:08 AM »
ReplyReply

Quote
More pixels allows for more detail to be captured from the lens, and this allows for better quality (enlarged) pictures, but it does not affect the enlargement phenomenon. Any detail the camera has captured is preserved when transferred to paper, but depending on the size of the paper that detail is spread more or less. If you spread the details too much, the human eye will see the errors, differences in focus, etc. That is the point.
[a href=\"index.php?act=findpost&pid=82262\"][{POST_SNAPBACK}][/a]

I'm sorry. It sounds like you're contradicting yourself (better quality (enlarged) pictures but it does not affect the enlagement phenomenon).

Seems to me you're speaking in analog terms about a digital issue. The details are spread, yes, but only based on the Mp resolution and not based on any sensor dimension magnification.

It is like saying that a larger (physically) CD will provide more sound resolution and detail....

If what you're saying is true then the APS vs FF debate would not exist since it would be obvious that larger sensor = better quality (and not because of noise issues etc). Of course, there is a limit in pixel density from an optical point of view and this has to do with lens resolution and diffraction. But this is another story).

I can't follow your thinking. Is it just me who thinks that 'enlargement' in the digital world is not directly related with sensor size because the captured image does not have physical dimensions?
« Last Edit: October 26, 2006, 11:31:43 AM by NikosR » Logged

Nikos
Jack Flesher
Sr. Member
****
Offline Offline

Posts: 2595



WWW
« Reply #34 on: October 26, 2006, 11:46:28 AM »
ReplyReply

Quote
While I agree with the basic principle of the article, I think this issue is over-exaggerated. First, with digital landscapes and "front to back" DOF I don't usually have a problem when printing at A3 and don't find the issue much different than film days. Second, I would dispute the fact that typical viewing size was 4x6 in film days. Maybe for snapshot drugstore prints, but serious amateurs like myself and even my dad typically printed 8x10 or larger, and in fact in my family the habit was to shoot chrome and project viewing at 40 inches. Bottom line - I don't really see where digital has significantly changed DOF issues from film days except possibly for those who never used to print larger than snapshot (4x6) size.

- DL
[a href=\"index.php?act=findpost&pid=82237\"][{POST_SNAPBACK}][/a]

Precisely Don!

The effects of Viewing Distance, Enlargement Factor and CoC on DoF has not changed at all with the introduction of digital. What has changed is our ability to easily discern the differences that were always there; we now have "actual pixel" view with a simple mouse-click. In film days, that level of inspection required viewing negatives through a medium-power microscope...
Logged

howiesmith
Guest
« Reply #35 on: October 26, 2006, 12:31:44 PM »
ReplyReply

The article ended with:

"Today’s big print culture means acceptable depth of field has shrunk dramatically. The old notions about hyperfocal focusing and depth of field scales based on a thirty micron circle of confusion, are redundant if you’re aiming at high quality A3 or larger prints"

The DoF scales based on 30 microns are not redundant, they are not applicable to very large or A3 prints.  Maybe the problem isn't that sience has changed, but the need to be more precise and accurate in its application has changed.  DoF scales are based on a standard print size viewed at a standard distance.  I don't recall A3 being one of those standard sizes.
« Last Edit: October 26, 2006, 12:45:12 PM by howiesmith » Logged
Rob C
Sr. Member
****
Offline Offline

Posts: 12213


« Reply #36 on: October 26, 2006, 02:37:22 PM »
ReplyReply

Quote
The article ended with:

"Today’s big print culture means acceptable depth of field has shrunk dramatically. The old notions about hyperfocal focusing and depth of field scales based on a thirty micron circle of confusion, are redundant if you’re aiming at high quality A3 or larger prints"

The DoF scales based on 30 microns are not redundant, they are not applicable to very large or A3 prints.  Maybe the problem isn't that sience has changed, but the need to be more precise and accurate in its application has changed.  DoF scales are based on a standard print size viewed at a standard distance.  I don't recall A3 being one of those standard sizes.
[a href=\"index.php?act=findpost&pid=82388\"][{POST_SNAPBACK}][/a]

Hi folks

I think that there is a lot of confusion going down - or at least, I'm getting more confused than ever. As I understrand it, this is totally an optics consideration based on the fact that all lenses of all focal length give exactly the same depth of field when used at the same aperture and with the same subject magnification. To illustrate this: if you focus a 135mm lens on a 35mm format camera at, say, five feet, you will get a pretty large head-shot. Now, using shorter or longer lenses at the same aperture to get a similarly sized image will still give you the same depth of field, only the area of background or foreground will have changed.

In 35mm format terms that 135mm lens is starting to be longish: in 4x5 format terms it is a short normal.

Printing that head shot from either format to whatever common size of head will not show a difference in depth of field, only in extra picture areas covered or lost.

N'est ce pas?

Ciao - Rob C
Logged

jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #37 on: October 26, 2006, 04:16:22 PM »
ReplyReply

Quote
Of course, there is a limit in pixel density from an optical point of view and this has to do with lens resolution and diffraction. But this is another story).
... and with the wavelengths of visible light.

Darn physics, always getting in the way!
Logged

Jan
Nemo
Sr. Member
****
Offline Offline

Posts: 276



« Reply #38 on: October 27, 2006, 10:00:16 AM »
ReplyReply

Quote
historic concepts of depth of field are inappropriate if enlarging to the capacity of today’s inkjet printers

It is due to two factors:

1) DoF marks are designed for A4 prints (aprox) and many of us make A3 prints.

2) Digital sensors are smaller than 35mm frames (in general).

That is all.
Logged
dlashier
Sr. Member
****
Offline Offline

Posts: 518



WWW
« Reply #39 on: October 27, 2006, 01:56:50 PM »
ReplyReply

Quote
It is due to two factors:

1) DoF marks are designed for A4 prints (aprox) and many of us make A3 prints.

2) Digital sensors are smaller than 35mm frames (in general).

That is all.
[a href=\"index.php?act=findpost&pid=82520\"][{POST_SNAPBACK}][/a]
Interestingly these two factors are offsetting somewhat minimizing the original point of the article. This is particularly true for P/S digital whose users are more likely the former 4x6 printers.

- DL
« Last Edit: October 27, 2006, 02:18:08 PM by dlashier » Logged

Pages: « 1 [2]   Top of Page
Print
Jump to:  

Ad
Ad
Ad