Using DOF Calculations Bestly for Print Size

[Tin Can]

I have used DOF Master Calculator often, this one. https://www.dofmaster.com/dofjs.html

Probably incorrectly as even I see problems in my small enlargements.

[Dan Fromm]

The usual rule of thumb is that resolution in the final print should be at least 8 lp/mm for it to appear sharp at normal viewing distance. 8 lp/mm is usually interpreted to mean a circle of confusion with diameter 1/8 mm = 0.125 mm.

Enlarging by x makes resolution in the print = 1/x resolution in the negative. So, to enlarge 2x would require a CoC of 0.125/2 = 0.0625 mm, and so on.

IMO, possibly mistaken, 8 lp/mm in the final print isn't enough, more is better. But this view is not widely held.

And now we see why enlarging film by more than 10x isn't always a good idea if (big if) a final print that looks sharp at normal viewing distance is needed. 10x requires at least 80 lp/mm in the negative. This is not easy to attain with the best equipment, film and technique. See, for example, https://www.flickr.com/photos/nesster/4424744296/sizes/o/ and https://www.flickr.com/photos/nesster/4424744224/sizes/o/

Against this, and I have and have seen lovely prints that were fuzzy all over the frame, sharpness isn't everything.

[Pere Casals]

8 lp/mm is usually interpreted to mean a circle of confusion with diameter 1/8 mm = 0.125 mm.

I guess this is not easy to interprete.

When in graphic arts they mention 300 dpi this is 11.8 pixels per mm, so 6 lp/mm at (ideally) 100% MTF. Is this more or less than analog 8lp/mm at extintion? what if using Rayleigh criterion ?

Ctein asks: Why does a print that resolves 20 lp/mm look less sharp than one that resolves 65 lp/mm?


I think that I understand MTF graphs and what is lp/mm at extintion, but I get often confused by terminology around !

[Jac@stafford.net]

[...] Ctein asks: Why does a print that resolves 20 lp/mm look less sharp than one that resolves 65 lp/mm?

If the printing is optical through an enlarger despite the degree of enlargement, or even 1:1, there is an unavoidable loss of resolution. It is physics.

[Pere Casals]

If the printing is optical through an enlarger despite the degree of enlargement, or even 1:1, there is an unavoidable loss of resolution. It is physics.

In tests I made I found that (a good) optic enlargement does not degradate Image Quality that it's in a LF negative. The enlarger lens usually works in a better situation than the taking lens:

> Lens can usually be used at the optimal aperture.

> Focus it is critically adjusted with a grain loupe, negative can be in a glass carrier, and the easel may keep paper very flat.

> Lens is usually specialized in the particular job. My old semi-destroyed "no letter" rodagon 210 takes at least 145 lp/mm from the negative.


Think that we see the grains in the loupe when focusing (if the loupe is strong enough) the LF image in the negative is less sharp than that, so little is lost.

Then paper resolves a lot, more than the enlarged image. Paper may resolve 60 to 100 lp/mm, if the LF negative has (say) 60 lp/mm when enlarging x2 we have 30lp/mm on the easel.




or even 1:1

At 1:1 we may have a loss if we don't use the right lens, for 1:1 we would need a Duplication lens (Rodagon R or D). If we use a regular lens it won't take all information in the negative because the lens won't work well at 1:1, anyway that pitfall it won't be noticed in a print because it will still print more than what an human eye would see.

______________________



In my (limited) experience what I found is that a (good) LF optic enlargement crafts an insane amount of image quality on a paper.

First that I found is that a EL Nikon 50mm takes from film way more than a drum scanner at 9,000dpi effective (if this exists), a LF Rodagon is not very far.
194778
(my glass slide is defective in the group 7, still it shows what an insane resolving power enlarger lenses have)


194779


That the enlargement has no loss it's easy to check, just make a sandwich with the negative and a usaf 1951 glass slide, with emulsion in contact with the pattern side to keep focus, with no calculation it's evident that we see way, way finer detail in the target than in the film, showing that the enlargement does not limit image quality.

194780


IMHO to challenge the enlarger lens at work we would need an extraordinary negative, something like a CMS 20 shot made with a crazy good lens that I guess it was not sold in the commercial photography market, say a 30 kg aerial lens.

[Jim Noel]

If the printing is optical through an enlarger despite the degree of enlargement, or even 1:1, there is an unavoidable loss of resolution. It is physics.

That is the reason I keep going to larger formats and contact print everything. My 2 enlargers may be dying of old age,but not overuse.

[Jim Noel]

to add to above, the reason I started this practice many years ago was a test I made. I made an enlargement of a very good negative on glossy paper. Then I contact printed a very similar,but larger, negative on Azo. The superiority of the contact was obvious to everyone to whom I showed it.

[Pere Casals]

glossy paper. Then I contact printed a very similar,but larger, negative on Azo. The superiority of the contact was obvious to everyone to whom I showed it.

Jim, if you say that the Azo result was better I'll for sure trust that, but this is not a side by side comparison. Different papers with different tonality, different shots, different formats.

IMO it's difficult that the enlargement (if good) degradates Image Quality that's in the LF negative, another question is what Image Quality we are able craft in the negative, and regarding this we have a DOF vs diffraction trade-off that may be different depending on the print size and depending on what we want.

[ic-racer]

Owners of 8x10 enlargers know that image degradation from stopping the camera lens too far is visible in prints. However everything is a compromise. The math has already been done so, setting aperture based on focus spread works best for me. I have a mark on my focus knob, so, for example, one revolution = f32, etc.

I can't tell the difference between my contacts and 1:1 enlargements onto photographic paper, other than the contact prints have Newton rings and dust.

[Doremus Scudder]

Y'all do realize that all this has been done and discussed in depth here: https://www.largeformatphotography.info/fstop.html , don't you?

FWIW, I'll gladly live with some diffraction and a smaller degree of enlargement to get the DoF I need for a particularly tricky subject. How else do you get those shots anyway? Like ic-racer, I've worked out tables for optimum aperture and maximum enlargement based on my preferred CoC. Works just fine for me.

Doremus

[Pere Casals]

How else do you get those shots anyway?

Many times Scheimflug helps a lot !!!

If the close part of the scene is at the bottom and the far part is at the top then some tilt always helps to save some diffraction, or a lot of diffraction, or all diffraction effect in the IQ. Or we can play with defocus, selective focus is also a great resource, if one likes that.

[Corran]

I have used DOF Master Calculator often, this one. https://www.dofmaster.com/dofjs.html

Probably incorrectly as even I see problems in my small enlargements.

I think the problem with DOF calculations, especially when using pure math to make them, is that it supposes the whole depth of focus area will be the same sharpness.

Regardless of aperture, only one point in space is "in-focus" while the rest inside the DOF is supposedly "close enough." I've found DOF calculators or DOF markings on small-format lenses to often grossly underestimate the acceptable DOF in most situations (35mm lenses especially, as they usually suppose a 4x6 "jumbo print" IIRC). If a given focus point calculates to have everything in focus front-to-back, I would stop down an additional 1-2 stops to ensure this if working at the extremes, regardless of "diffraction."

I also tend to reject the "1/3 in front, 2/3 behind" focus paradigm in many situations, especially with a strong foreground element. A foreground element inside the DOF and math claims should be "acceptably sharp" to me often seems unacceptably fuzzy on the negative - so focusing a bit closer and stopping down a bit more seems more effective. YMMV.

[Tin Can]

POINT taken!

Thank you


I think the problem with DOF calculations, especially when using pure math to make them, is that it supposes the whole depth of focus area will be the same sharpness.

Regardless of aperture, only one point in space is "in-focus" while the rest inside the DOF is supposedly "close enough." I've found DOF calculators or DOF markings on small-format lenses to often grossly underestimate the acceptable DOF in most situations (35mm lenses especially, as they usually suppose a 4x6 "jumbo print" IIRC). If a given focus point calculates to have everything in focus front-to-back, I would stop down an additional 1-2 stops to ensure this if working at the extremes, regardless of "diffraction."

I also tend to reject the "1/3 in front, 2/3 behind" focus paradigm in many situations, especially with a strong foreground element. A foreground element inside the DOF and math claims should be "acceptably sharp" to me often seems unacceptably fuzzy on the negative - so focusing a bit closer and stopping down a bit more seems more effective. YMMV.

[Pere Casals]

The DoF calculator app (by Jonathan Sachs) for the smartphone I use tells the CoC in each place of the field, so we know what optic performance remains at each spot we want to consider.

194829

There is also a Windows version

http://www.dl-c.com/DoF/


One thing it was useful to me to learn to balance diffraction vs DOF was making a practice with a (cheap) laser range meter and the smartphone app, without the camera, just taking a real scene, considering different apertures and focus distances, and then checking the CoC for the different spots.

This is an interesting exercise because one learns to balance optical performance in the plane of focus vs the rest in the DOF.

[Tin Can]

Pere, need exact name of app iOS

[Doremus Scudder]

Many times Scheimflug helps a lot !!!

If the close part of the scene is at the bottom and the far part is at the top then some tilt always helps to save some diffraction, or a lot of diffraction, or all diffraction effect in the IQ. Or we can play with defocus, selective focus is also a great resource, if one likes that.

Pere,

After 30+ years of LF, I'm adept at using movements to position of the plane of sharp focus in order to optimize my aperture. I even wrote an article on the subject for the now-defunct View Camera magazine. Still, there are times when, even after applying optimal movements, stopping down into the realm of diffraction degradation is the only solution to get the necessary depth-of-field for a particular scene (think very near/far with lots of verticals... you get it). In cases like that, e.g., when I'm forced to use f/45 with a 4x5 camera, it's good to know in advance just how large one can print before diffraction becomes too distracting.

As far as defocus goes... it's not really my style.

Best,

Doremus

[Corran]

Of course it depends on what / where you photograph, but things rarely fit inside the perfect Scheimpflug "wedge" at f/22 for me.

[Doremus Scudder]

I think the problem with DOF calculations, especially when using pure math to make them, is that it supposes the whole depth of focus area will be the same sharpness.

Regardless of aperture, only one point in space is "in-focus" while the rest inside the DOF is supposedly "close enough." I've found DOF calculators or DOF markings on small-format lenses to often grossly underestimate the acceptable DOF in most situations (35mm lenses especially, as they usually suppose a 4x6 "jumbo print" IIRC). If a given focus point calculates to have everything in focus front-to-back, I would stop down an additional 1-2 stops to ensure this if working at the extremes, regardless of "diffraction."

I also tend to reject the "1/3 in front, 2/3 behind" focus paradigm in many situations, especially with a strong foreground element. A foreground element inside the DOF and math claims should be "acceptably sharp" to me often seems unacceptably fuzzy on the negative - so focusing a bit closer and stopping down a bit more seems more effective. YMMV.

Corran,

You're exactly right. One has to do test for one's own acceptable circle of confusion and do the math based on that. Using some one else's standards, especially the overly-optimistic ones from many manufacturers, just doesn't cut it. After one has done one's homework, the system is really reliable.


"1/3 in front, 2/3 behind" is just wrong except at one, rather close-to-the-camera distance. At longer distances, 1/2 / 1/2 is more like it. Finding good "near / far" points and splitting the focus spread (based on your own standards) automatically adjusts for the different distances in front of and behind the plane of sharp focus. When using tilt/swing movements, the situation becomes more complicated; "near" and "far" can be "down" and "up," or "right" and "left." Visualizing where the plane of sharp focus is in a scene helps immensely when choosing focus points.



Of course it depends on what / where you photograph, but things rarely fit inside the perfect Scheimpflug "wedge" at f/22 for me.

Me neither! I think my most used aperture for 4x5 is f/32. It seems to me, that any scene that has significant near/far depth will require stopping down past the optimal aperture if you want everything sharp (why do you think the West Coast School chose the name "f/64" ?)

Best,

Doremus

[Pere Casals]

After 30+ years of LF, I'm adept at using movements to position of the plane of sharp focus in order to optimize my aperture. I even wrote an article on the subject for the now-defunct View Camera magazine. Still, there are times when, even after applying optimal movements, stopping down into the realm of diffraction degradation is the only solution to get the necessary depth-of-field for a particular scene (think very near/far with lots of verticals... you get it). In cases like that, e.g., when I'm forced to use f/45 with a 4x5 camera, it's good to know in advance just how large one can print before diffraction becomes too distracting.


Doremus, of course... but let me propose next: if we have to shot at f/45 then diffraction limits our image quality to 35lp/mm at extintion, so a better alternative is using a MF back (or camera) with a shorter focal that it would allow an slower film for the same shutter speed, we save film without any IQ loss...


IMHO sometimes in a great LF shot we see how the photographer searched the right framing and angle to make shine the view camera taking, perhaps we may leave some kind of scenes for the smartphones...



Pere, need exact name of app iOS

Randy, sorry, this one is only released for windows and for android. Best bet would be taking an old/used/cheap android phone for that.