Jump to content
Sign in to follow this  
SJP

DOF resolution and perspective

Recommended Posts

Advertisement (gone after registration)

Best wishes for 2008 to all - it should be a good year for more fun with our new toy.

 

Over the last couple weeks I have been writing some stuff - basicallly to see what I can figure out myself. I think this may appeal to some of you. It concerns how to think about DOF resolution and perspective. Some of the stuff is pretty basic - other things I found surprising myself. For example concerning the effect of a cropped sensor.

 

Anyway - I hope you will enjoy the read, any comments & suggestions you may have are welcome. In the thread below or by PM.

 

Stephen

 

Note: to do calculations yourself just click on the picture of the excel file.

Share this post


Link to post
Share on other sites

Stephen,

 

That's rather some piece you wrote. I'll have to read it a few more times to really get to the bottom of it. But for somebody who is interested but "unschooled" on photography (like me), it gives a nice insight in what's what.

 

I had a look at the spreadsheet and tried to enter the circle of confusion figures for M8. You say it's "accepted" to be 20 ... is that not 23? I'm not sure what it "does" this small difference. The sheet is protected. Are you willing to share / fill in the 20 and 23 and post them as well?

 

Marco

Share this post


Link to post
Share on other sites

Marco

 

Thanks for your comment.

 

I agree that the COC could be 23 instead of 20 micrometer - did you find a reference for that somewhere? Maybe other forum members could comment on what they think is appropriate & I can modify the pdf.

 

The green cells in the spreadsheet are unprotected including the COC value. Changing this from 20 to 23 micrometres gives the following:

 

COC 20, 35 mm, F5.6, 5m gives: DOF 6.06 m (3.405 - 9.464 m)

COC 23, 35 mm, F5.6, 5m gives: DOF 7.56 m (3.260 - 10.818 m)

 

The unprotect code if you want to tinker further with the spreadsheet is: leica - obviously

 

Stephen

Share this post


Link to post
Share on other sites

Stephen,

 

My reference is Hyperfocal Distance and Depth of Field Calculator - DOFMaster. They use 23 for the M8 and 15 for the digilux 3. But then again, that's just 75% of 30 for the M8 and 50% of 30 for the D30.

 

There is an excellent article by Harold M. Merklinger "adjusting depth of field". He wrote it in 1991 .... and, if I understand him correctly, he says that the whole idea of 1/30 of a mm is outdated. I'm sure I'm confused enough by this subject and am reading about it just out of some weird curiousity and in the hope I can translate it back to real world situations. I'm sure I am completely out of my own depth of field:D with regard to this subject.

 

If you did not already read Merklinger, you can find it at:

 

JimDoty.com - Merklinger on Depth of Field

 

Marco

Share this post


Link to post
Share on other sites
Marco

 

Thanks for your comment.

 

I agree that the COC could be 23 instead of 20 micrometer - did you find a reference for that somewhere? Maybe other forum members could comment on what they think is appropriate & I can modify the pdf.

 

The green cells in the spreadsheet are unprotected including the COC value. Changing this from 20 to 23 micrometres gives the following:

 

COC 20, 35 mm, F5.6, 5m gives: DOF 6.06 m (3.405 - 9.464 m)

COC 23, 35 mm, F5.6, 5m gives: DOF 7.56 m (3.260 - 10.818 m)

 

The unprotect code if you want to tinker further with the spreadsheet is: leica - obviously

 

Stephen

 

Marco is right: norm for the 1.3 crop is 0.23, for APS-C is 0.20. The values are arrived at mathematically by postulating the greater enlargement needed. However, the value of the COC is the weak point in all this. It is a historical value, based on film of the thirties. I would argue that for the more precise sensors and larger prints of today a value of 0.20 FF, 0.18 for 1.3 and 0.15 for APS-C would be nearer the mark.

Share this post


Link to post
Share on other sites
...norm for the 1.3 crop is 0.23, for APS-C is 0.20. The values are arrived at mathematically by postukating the greater enlargement needed. However, the value of the COC is the weak point in all this. It is a historical value, based on film of the thirties...

You mean 0.023mm and 0.02mm i guess Jaap.

I did not try with the M8 but usual CoC values work fine for me with full frame (0.03mm) as well as APS-C (0.02mm) cameras.

They're still used by modern DoF converters if i'm not wrong.

Share this post


Link to post
Share on other sites

Advertisement (gone after registration)

Yes-because they are used by the camera industry as an universal standard. The needed value of the COC for photographic purposes depends on the total enlargement from subject to eye.

Share this post


Link to post
Share on other sites

Lct,

 

The thing that Merklinger brings forward, is that the basis of all this (35mm with a COF of 0,030mm) is outdated and was already in 1991. This standard was set up somewhere beginning last century and lenses and output (be it print or in our age screen) have a much higher/better resolution.

 

He goes sofar that he thinks that in some situations 0,030 is OK, but in other situations you might have to go to a much smaller number.

 

Really, read his texts or even just the first one, it's even funny and very down to earth through out! Stephen's text is, in a way, much more technical (but ads to it with some very nice graphs and explanations).

 

Marco

 

Oh what joy ... I'm discussing something technical ... don't start asking real questions

Share this post


Link to post
Share on other sites

I hear you Marco but the usual CoC values are perfectly up-to-date for me. I've been using them for 30 years with film and they still work with my APS-C digital cameras.

Share this post


Link to post
Share on other sites

"Oh what joy ... I'm discussing something technical ... don't start asking real questions"

 

Marco

 

Thanks for the Harold Merklinger articles link - indeed a different point of view that requires some further thought.

 

Concerning using 30x0.75 = 23 micron COC for an M8 this looks like a safe estimate when comparing prints of the same size. The sensor however has pixel size of 6.8 micron so the COC limit would be in the order of 14 micron - I think 2x pixel size is called for. So meke your choice depending on what you need & expect. Or you can flip a coin

 

I will include this in a revised version of the file in due course once I have sorted it out for myself.

 

Anyway - let's not forget that taking pictures is what it is all about! First learn about the technique and then forget it, or at least most of the time.

 

Stephen

Share this post


Link to post
Share on other sites

Despite enormous succes of this highly scientific contribution to the forum my (former) provider managed to foul things up in a major way and my website has evaporated into cyberspace. Fear ye not I have a trick up my sleeve:

 

http://spicken59.googlepages.com/DOFresolutionandperspective.pdf

 

Note the embedded excel file also changed location:

http://spicken59.googlepages.com/resolutionanddof.xls

Share this post


Link to post
Share on other sites
Lct,

 

The thing that Merklinger brings forward, is that the basis of all this (35mm with a COF of 0,030mm) is outdated and was already in 1991. This standard was set up somewhere beginning last century and lenses and output (be it print or in our age screen) have a much higher/better resolution....

 

[/i]

 

 

The best I can recall from when I studied this (35 years ago) is that depth of field starts from the resolution required on an 8x10 print viewed at a standard viewing distance (equal to the print diagonal) and then is calculated backwards for various film or sensor formats and lenses.

 

It has nothing to do with modern vs. old lenses or the maximum resolving capability of the gear in any way. It is simply a theoretical calculation. There is certainly no guarantee that every lens/sensor combo will provide the resolution required for the "calculated" depth of field.

 

The concept was to find the largest circle of confusion that was acceptable for most viewers (average eyesight) when looking at the print. Any smaller circle of confusion on the print would not show added detail at that viewing distance.

 

As human eyesight has not evolved in the past century, whatever methods were employed to determine this would still be valid today regardless of the print size as long as the print is viewed from the correct distance. But that is not what most people do when they enlarge a photo to 100% on their monitor and then press their nose up against the screen. So if anyone wants to create new dof tables for a 24inch LCD viewed at 6 inches, then go for it.

Share this post


Link to post
Share on other sites

If it works it works folks. Try the DoF markings of the nearest faster f stop of your lenses i.e. f/5.6 when you choose f/8 for metering by example. Works fine with my APS-C digicams AFAIC.

Share this post


Link to post
Share on other sites

1 The green cells in the excel file are unprotected and can be changed to the values you want to use. To unprotect the whole file the secret password is leica (see one of the previous posts).

2 for the M8 use the actual focal length (as that does not change) but use a COC value of 23 micrometers instead of 31 micrometers for 'film'.

 

The pdf file should also provide you with the information you need.

 

Succes!

Stephen

Share this post


Link to post
Share on other sites

Hi,

could the spreadsheet be used in order to predict the appropriate Si when working with different wavelengths (and then figure out where to focus the lense using the current scale written on it ?)

I am using lenses without IR focusing marks for IR photo with the M8 and I am guestimating the focus by reviewing the picture on the LCD (and I start to know by how much I should shift the focus) but being able to use the spreadsheet for that would be quite fun.

 

Can it also be used to compute the distances at which an object will be in focus while using macro extension rings ? (It looks like all the info is there in the spreadsheet I assume that playing with SO/Si is what I need to do for that).

 

Thanks

Share this post


Link to post
Share on other sites

Hmmmm.

 

Interesting question. I think you need to know the IR focal length of the lens to do that (so 50mm visible is say 55mm for IR - just as a guess). The IR focal length presumably depends on the specific lens (glass & disperion curve) you are using so IR Noctilux could be diffrent from IR Summicron 50mm. And this also depends on the actual IR wavelength that is used (i.e. filter dependent?).

 

Does anyone have any suggestions? Are there Leica lenses with a IR focus line on the barrel? Then you could backcalculate the appropriate focal length.

 

I have no experience with (intentional) IR photography with the M8 so this sounds like something the old &-or experienced rangefinders could help with.

 

Stephen

Share this post


Link to post
Share on other sites

Dear Professor,

Your overview (pdf file) and spreadsheet are very interesting and I want to understand more than I know.

In Analysis 1: constant perspective, page 12 of pdf file, you write: "Replacing the 35 mm film camera with a 50% cropped digital sensor at f=25 mm..." From where comes f=25 mm? In the same case i.e. 50 mm lens but now with the M8, how much % is cropped the the digital sensor of the M8 and at f=? (For me crop is 1/0.75 = 1.333)

Thank you.

Best regards.

Michel

Share this post


Link to post
Share on other sites

Dear Michel,

The factor 0.5 (50% crop) was used as a 'easy example' in that case a 25mm lens on a 50% cropped sensor is 'the same" as a 50mm lens on 35mm film (the 35 mm film format being defined as full frame i.e. uncropped). For the M8 with a crop factor of 0.75

film to M8 x 0.75

35 => 26 (35x0.75 = 26.25)

50 => 37 (50x0.75 = 37.5)

70 => 53 (70x0.75 = 52.5)

135 => 100 (90x0.75 = 101)

 

Other way round

M8 to film x 1.33

28 => 37 (28x1.33=37.25)

35 => 47 (35x1.33=46.55)

50 => 67 (50x1.33=66.5)

70 => 93 (70x1.33=93.1)

90 => 120 (90x1.33=119.7)

 

I hope this helps

regards

Stephen

Share this post


Link to post
Share on other sites
Dear Michel,

The I hope this helps

regards

Stephen

 

Thank you, but what I need to understand is what are right figures must I use in your xls file for the following example?

 

M8 + Summicron 75/2 at f/4 for a distance of 10 m

 

f = ? mm [75 or 75/0.75 or 75*0.75]

f/# = 4 f/value

COC = 20 µm

 

So+Si = 10 m

 

Best regards

Michel

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue., Read more about our Privacy Policy