Home | Contact Us | FAQ | Search & Site Map | Link to Us
Sign In | Join | Other 45 Sites in Network
PhotoKB Home
Discussion Groups
Digital Photography
Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot Cameras
Film Photography
35 mmLarge FormatMedium formatDarkroomFilm and LabsOther Equipment
Photo Technique
Nature PhotographyPeople PhotographyTechnique General
General Photo Topics
General TopicsAustralian PhotographyUK Photography
DirectoryPhoto Clubs

Photo Forum / Digital Photography / DSLR Cameras / April 2007

Tip: Looking for answers? Try searching our database.

Depth of Field -- Part II

Thread view: 
Enable EMail Alerts  Start New Thread
Thread rating: 
ron - 24 Apr 2007 14:28 GMT
I have just posted the second article in a two part series on depth of
field article. The article can be found at my site at:

http://ronbigelow.com/articles/depth-of-field-2/depth-of-field-2.htm

Other articles can be found at:
http://ronbigelow.com/articles/articles.htm

Ron Bigelow
http://ronbigelow.com
Charles - 24 Apr 2007 21:55 GMT
>I have just posted the second article in a two part series on depth of
> field article. The article can be found at my site at:

Thanks Ron.
RichA - 26 Apr 2007 06:20 GMT
> >I have just posted the second article in a two part series on depth of
> > field article. The article can be found at my site at:
>
> Thanks Ron.

I wonder why modern cameras (unless there is such a thing?) don't have
a program where hyperfocal distance could be automatically set?  Might
come in handy.
ben brugman - 26 Apr 2007 21:08 GMT
>> >I have just posted the second article in a two part series on depth of
>> > field article. The article can be found at my site at:
[quoted text clipped - 4 lines]
> a program where hyperfocal distance could be automatically set?  Might
> come in handy.

Because hyper focal does not only depend on focallength and aperature, but
also on the coc and this last depends on how critical the photografer is.
(Or de viewer in the end).

The second problem is that the focus distince is not calculated, but the
distance is set using a feedback system, bit similar to what we do when
focusing manually. We do not know the distance but try to make the
picture as sharp as we can. Most camera's don't have a clue what the
focusing distance is. Some camera's do have an approximate idea, which
can be used for flashing for example. But this is not precise enough to
set to a calculated focal distance.
Can a camera (or lens) be made to have exact focal distance info, yes,
but then the lens would be far more expensive, because now we focus
on the image and compensate in that process for some of the
tolerances of the lens. When calculating a setting, we do not compensate
for the tolerances of the lens.

ben
Gisle Hannemyr - 25 Apr 2007 13:37 GMT
> I have just posted the second article in a two part series on depth of
> field article. The article can be found at my site at:
>
> http://ronbigelow.com/articles/depth-of-field-2/depth-of-field-2.htm

I'm puzzled by the formula you use for computing CoC,

   CoC = 0.25mm/(325.28mm/Diameter of the sensor).

since it is just a convoluted way of writing:

   CoC = Diameter of the sensor / 1301

Any paricular reason you choose this format?

Btw. the exact number to use for CoC is not only a function
of the diameter of the sensor, but also of how big you plan
to print and even the acuity of the person looking at the
photograph.  For this reason a lot of different values for Coc
can be found in literature.  It is probably not a good idea
to present this as if it is physical constant.
Signature

- gisle hannemyr [ gisle{at}hannemyr.no - http://hannemyr.com/photo/ ]
------------------------------------------------------------------------
      Sigma SD10, Kodak DCS460, Canon Powershot G5, Olympus 2020Z
------------------------------------------------------------------------

Tony Polson - 25 Apr 2007 14:53 GMT
>> I have just posted the second article in a two part series on depth of
>> field article. The article can be found at my site at:
[quoted text clipped - 17 lines]
>can be found in literature.  It is probably not a good idea
>to present this as if it is physical constant.

Also, the idea that it can be presented to an "accuracy" of four
significant figures is spurious.  

The debate should be about whether the number should be 500, 1000,
1500, 2000 or 2500.  In any case, what the figure actually means is
open to conjecture, as you say.  Whether the figure should be 1301 or
1299 or 1303 is simply not worthy of intelligent discussion.
maxsilverstar@yahoo.com - 26 Apr 2007 07:40 GMT
>> I have just posted the second article in a two part series on depth of
>> field article. The article can be found at my site at:
>>
>> http://ronbigelow.com/articles/depth-of-field-2/depth-of-field-2.htm

/snip/

>Btw. the exact number to use for CoC is not only a function
>of the diameter of the sensor, but also of how big you plan
>to print and even the acuity of the person looking at the
>photograph.  For this reason a lot of different values for Coc
>can be found in literature.  It is probably not a good idea
>to present this as if it is physical constant.

This is "the second article in a two part series on depth of field". (see above)
It is therefore unfair (and unproductive) to critique any aspect of the second
part as though it is entirely independent of the first part. More simply put,
the article must be considered as a whole.

from the first part:

"DOF refers to the sharpness of the image in its final form (such as a print).
DOF is not something that is inherent in only the camera and lens. For instance,
in the case of a print, DOF depends on how large the image is printed."

"...DOF is intractably tied to the nuances of the human visual system. What
appears sharp is determined by the human eye and brain. Thus, the DOF of an
image depends on what a viewer's eye can resolve."

Also see the section titled "Circle of Confusion" which explains what
assumptions have been made in order to create DOF tables and lens markings for
35mm film (or 'full-frame' sensors).

Better still, just read the whole first part of the article. When you do I
expect you will discover that the author has not presented CoC as a constant. I
further expect that you will agree with me that Bigelow has done an excellent
job of explaining DOF. I have probably two or three dozen books that attempt the
same task, and I doubt that more than two or three succeed as well at Bigelow
has.
cmyk - 26 Apr 2007 09:05 GMT
>>> I have just posted the second article in a two part series on depth of
>>> field article. The article can be found at my site at:
[quoted text clipped - 35 lines]
> same task, and I doubt that more than two or three succeed as well at Bigelow
> has.

That would be OK is the first article itself could be relied on. However, that article's claims, like:
. A typical assumption for a full-frame digital camera or a 35mm film camera is that the photographer will make an 8"x10" inch print
that will be viewed at about 12"
. An 8"x10" print has a diameter of 325.28mm.
. A full sized sensor is 24mm x 36mm and has a diameter of 43.27mm
. Well, the human eye can resolve about four lines per mm (up close).
. That means that the human eye can barely see a circle with a diameter of about 0.25mm. Any circle with a diameter smaller than
that will appear as a point. ... So, a 0.25 mm circle on an 8"x10" print would barely be recognizable.
make the lot worse than useless - it's outright misleading. The facts are:

1. 8"x10" prints are NOT typical In the real world, they probably account for less than 0.1% of prints. Plus the accepted
industry standard for viewing distances is 14" - and that's for 6"x4", not the larger prints that most people view from farther
away. See Kodak's Tech Report E-58 on the Print Grain Index. Quote: "The PGI method specifies a common viewing distance
of 14 inches (36 cm) for all print sizes. The distance at which small (4 x 6-inch) prints are typically viewed is 14 inches. This is
also the approximate distance at which larger prints are usually inspected for graininess." Larger prints of course, are typically
viewed from even farther away than 6"x4" prints - not closer.

2. An 8"x10" print has a DIAGONAL of 325.28mm - not a diameter of 325.28mm. Its rectangular, not circular. Likewise, a full-sized
35mm film has a DIAGONAL of 43.27mm - not a diameter of 43.27mm, and you'd have to enlarge it to 8"x12" (ie 8.47 times, not the 7.5
times the article implies) before cropping an 8"x10" print from it.

3. As for the a 0.25mm circle, and the 4lines/mm claim, that translates to a measley 100dpi. Try printing any 8"x10" photo at 100dpi
and see if its good for anything other than throwing in the bin - certainly no good for viewing from 12"!! At that distance you need
at least 300dpi for a decent print.

Kind of makes a mockery of the article's nice neat maths.

Signature

cmyk

Boskey - 26 Apr 2007 10:59 GMT
> 2. An 8"x10" print has a DIAGONAL of 325.28mm - not a diameter of
> 325.28mm. Its rectangular, not circular. Likewise, a full-sized 35mm film
> has a DIAGONAL of 43.27mm - not a diameter of 43.27mm, and you'd have to
> enlarge it to 8"x12" (ie 8.47 times, not the 7.5 times the article
> implies) before cropping an 8"x10" print from it.

Have you even considered that the DIAGONAL and the diameter could be equal?
The four points of the rectangle are located on the circle diameter line and
have opposite sides equal in length.
cmyk - 26 Apr 2007 11:24 GMT
Have you ever considered using a dictionary? Mine doesn't apply diameters to rectangles ...

Signature

cmyk

Matt Clara - 27 Apr 2007 22:18 GMT
> Have you ever considered using a dictionary? Mine doesn't apply diameters
> to rectangles ...

No, but his reasoning is correct.  I created this image in a cad program
quickly.  Each unit is a millimeter.  The red and the blue lines both
measure out to exactly 22.361mm in length.
http://www.mattclara.com/misc/misc/diameter.gif

--
www.mattclara.com
cmyk - 27 Apr 2007 22:25 GMT
>> Have you ever considered using a dictionary? Mine doesn't apply diameters to rectangles ...
>
[quoted text clipped - 4 lines]
> --
> www.mattclara.com

I think your missing the real point. Sure, you can create an image circle that has the same diameter as the film's diagonal, but
that doesn't give the film, or the rectangular image printed from it, a diameter.

In any event, the diameter/diagonal issue is a minor one compared to the other flaws I've pointed out, which are fatal to the whole
article's credibility.

Signature

cmyk

Tony Polson - 27 Apr 2007 22:52 GMT
>I think your missing the real point. Sure, you can create an image circle that has the same diameter as the film's diagonal, but
>that doesn't give the film, or the rectangular image printed from it, a diameter.

A lens provides a circle of illumination whose diameter needs to be
equal to, or greater than the diagonal of the rectangular format of
the film or sensor.  I think that was Matt's point, and if so, it was
well made.  

Where the two are equal, the terms "diameter" and "diagonal" are in
effect synonymous.  Given that, arguing from a pedantic point of view
about the different meanings of the two terms seems to me to be a very
good example of the term "missing the point".
cmyk - 28 Apr 2007 00:33 GMT
>>I think your missing the real point. Sure, you can create an image circle that has the same diameter as the film's diagonal, but
>>that doesn't give the film, or the rectangular image printed from it, a diameter.
[quoted text clipped - 8 lines]
> about the different meanings of the two terms seems to me to be a very
> good example of the term "missing the point".

Without placing too fine a point on it, it is Ron Bigelow's article that claimed to give "a technical definition of DOF" and "To
understand DOF, it is necessary to understand the circle of confusion". It behoves him, therefore to get his terms and, more
importantly, facts right.
Signature

cmyk

Tony Polson - 28 Apr 2007 20:12 GMT
>>>I think your missing the real point. Sure, you can create an image circle that has the same diameter as the film's diagonal, but
>>>that doesn't give the film, or the rectangular image printed from it, a diameter.
[quoted text clipped - 12 lines]
>understand DOF, it is necessary to understand the circle of confusion". It behoves him, therefore to get his terms and, more
>importantly, facts right.

I don't disagree, but with respect, that was not the point being
discussed.
Gisle Hannemyr - 27 Apr 2007 06:35 GMT
> This is "the second article in a two part series on depth of
> field". (see above) It is therefore unfair (and unproductive) to
> critique any aspect of the second part as though it is entirely
> independent of the first part. More simply put, the article must be
> considered as a whole.

That is a fair opinion, but that is not how people, in general, read
when looking up information on the web.  (There is a lot of research
on this, see for example, Jacob Nielsens books on Web usability).

When creating for a particular medium, I think the author need to
consider how a particular audience "reads" that medium.  In this
particular case, a brief comment with a hyperlink referring the
reader to part I for the definitions of CoC and DoF would have
been suffiscient.

As others has noted, there is some issues with part I as well.
However, I see no point in making further comments if Ron B. is not
reading the groups he posts in and is not interesting in suggestions
on how to improve his site.
Signature

- gisle hannemyr [ gisle{at}hannemyr.no - http://hannemyr.com/photo/ ]
------------------------------------------------------------------------
      Sigma SD10, Kodak DCS460, Canon Powershot G5, Olympus 2020Z
------------------------------------------------------------------------

Roger N. Clark (change username to rnclark) - 27 Apr 2007 03:06 GMT
> I have just posted the second article in a two part series on depth of
> field article. The article can be found at my site at:
[quoted text clipped - 6 lines]
> Ron Bigelow
> http://ronbigelow.com

Preface: I think very highly of Ron's articles and I have learned
quite a bit from them.  So I mean this post to point out
errors so Ron can fix them.  However, I have never seen Ron
actually respond to any post so I'm not sure he reads these
(he has not made other corrections suggested in the rec.photo
newsgroups in the past).  It would be nice if Ron wants to post here,
that he would engage in the discussion too.  To only post
is just spamming.

There are several significant issues with this article.
As other have noted, the 0.25 mm spot "would barely be recognizable"
is definitely not correct.  At 12-inches, a 0.25 mm circle
would be 2.8 arc-minutes in diameter.  The human eye
resolves two objects slightly less than 1 arc-minute.
This can be seen in Figure 2.6, the curve labeled:
critical visual angle at: http://www.clarkvision.com/visastro/omva1
The critical visual angle (CVA) was derived by Blackwell:

Blackwell, R.H., Contrast Thresholds of the Human Eye, Journal of the
Optical Society of America, v36, p624-643, 1946.

and the data are still valid today.  That CVA in good office
light is around 0.6 arc-minutes.  That is 4.7 times smaller than
Ron's value on his web page, and that is a fundamental error so
the web page needs to be redone.

Another significant error is the discussion on sensor size on the
depth of field.  It is incorrect.  Here is a discussion of why:

 The Depth-of-Field Myth and Digital Cameras
 http://www.clarkvision.com/photoinfo/dof_myth

Roger
ink - 27 Apr 2007 07:39 GMT
>> I have just posted the second article in a two part series on depth of
>> field article. The article can be found at my site at:
[quoted text clipped - 15 lines]
> that he would engage in the discussion too.  To only post
> is just spamming.
<SNIP>

Roger,

he replies if e-mailed via the link on his site - I think it might be a good
idea to copy/paste your issues into an e-mail there.

Cheers,
ink
Paul E - 27 Apr 2007 23:17 GMT
>> I have just posted the second article in a two part series on depth of
>> field article. The article can be found at my site at:
[quoted text clipped - 40 lines]
>
> Roger

I emailed Ron about the flaws in Part I, before Part II was published, alerting him to the kinds of flaws as have been highlighted
in this thread. I pointed him to industry articles (eg Zeiss camera Lens News # 1 & 2 and the Kodak paper cmyk referred to) that set
the rationale behind these in perspective and directly contradicted his claims as to what the industry standards are.

I even sent him a calculator I've developed that can calculate the CoC and DoF for any combination of film format, lens focal
length/aperture, print size, viewing distance and required sharpness.

Ron's reply indicated he wasn't interested in making any changes, regardless.

Cheers
Signature

Paul E

mark.thomas.7@gmail.com - 28 Apr 2007 02:47 GMT
On Apr 27, 12:06 pm, "Roger N. Clark (change username to rnclark)"
<usern...@qwest.net> wrote:
> Preface: I think very highly of Ron's articles and I have learned
> quite a bit from them.  So I mean this post to point out
> errors so Ron can fix them.  However, I have never seen Ron
> actually respond to any post so I'm not sure he reads these
> (he has not made other corrections suggested in the rec.photo
> newsgroups in the past).
I agree - generally Ron's stuff is excellent.  In his defence, I have
had a few conversations with him both on and off groups.  He is quite
reasonable, but tends to 'stick to his guns'!

That said, I haven't really perused this article, as my mind refuses
to engage with DOF calculations except as a mathematically interesting
pastime.  To me, such calc's are not very useful in my usual workflow,
and I prefer to just get to know my lens behaviour as much as
possible, then.. 'go with the force, Luke'....
Tony Polson - 28 Apr 2007 20:14 GMT
>I agree - generally Ron's stuff is excellent.  In his defence, I have
>had a few conversations with him both on and off groups.  He is quite
>reasonable, but tends to 'stick to his guns'!

I think everyone who takes issue with Ron's advice should email him
asking for a 100% refund of the price they paid him for it.
 
Sign In
Join
My Latest Posts
My Monitored Threads
My Blog
My Photo Gallery
My Profile
My Homepage

Start New Thread
Enable EMail Alerts
Rate this Thread



©2008 Advenet LLC   Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.