At what aperture lens has the most resolving power?

[Ivo Stunga]

While doing research on films with clear base, I stumbled across the description of Adox CMS 20 II PRO, which resolves up to 800l/mm and recommends this, quote: "If you want high resolution pictures you need to open your lens aperture to one stop below maximum opening. Otherwise the lens diffraction will lower your lens resolution down to half of what this film can capture. Best lenses are F1,4 high speed high quality lenses."

https://www.adox.de/Photo/films/cms20ii-en/


Luckily I do have a Zuiko 50mm 1.4 lens and can try this out in projection. While I'll hunt the film down I wanted to hear on this more from you, as I was under the impression that lenses are best the other way around - about two stops from being max closed.

I feel contradicted. Please uncontradict me.

 

[Dan Fromm]

There is no useful rule of thumb.

 

[Nicholas Lindan]

It depends on the lens. Exceedingly high end lenses - think Hubble space telescope - work best wide open. No camera lenses available to mere mortals are made to such high standards that they will work best wide open.

In general 50mm f1.4 lenses have lower resolution than 50mm f2.0 lenses. Usually the highest resolution will be around f4.0 with a 50mm lens.

For 35mm the highest resolution lenses in a maker's line-up are the ~100mm macro lenses.

Using a #25 red filter will increase the resolving power of a lens as it removes chromatic aberration. You can try an APO lens if one is available for your camera but a #25 filter is a lot cheaper.

 

[Arthurwg]

It depends on the lens. Exceedingly high end lenses - think Hubble space telescope - work best wide open. No camera lenses available to mere mortals are made to such high standards that they will work best wide open..

I think I've read that some lenses, perhaps Hasselblad, are just as good wide open as stopped down.

 

[Paul Howell]

Just to piggy back on what has bee posted, for high resolution in 35mm think macro lens, 50MM and 100 to 105 macros are designed for optimum resolution. Most lens have highest resolving power 1 to 3 stops down depending on the age of the lens. Very few legacy lens from the 70s and 80 will resolve at 800l/mm currently the the Sigma AF 50 1.4 Art is as sharp at 1.4 as it is up and down the aperture range, still I don't know if it will resolve 800/MM. CMS 20 is a microfiche film designed for high contrast images of books and newsprint and needs a developer to match for general purpose photography. Maybe 10 years I tested my box of 50mm lens with my last few rolls of microfiche film. All of the lens I tested were MF from the 60s and 70s, did not have an Olympus, Canon or Nikon, my testing showed Konica 50mm 1.7 to have the highest resolution at around 650/MM, just a rough test as I did not have a very good microscope, with a better scope might have shown greater resolution. Next was Miranda, Pentax, and Petri. Likely the Nikon medical macro, Leica and Zeiss would have tested as good or better than the Konica 50 1.7. By the way the Konica hit the sweet spot at F8 the Miranda 50 1.4 at 5.6. All the lens I tested including Minolta, Yashica Mamiya and Kowa all were able resolve Tmax 100 at 200/MM.

 

[ic-racer]

Luckily I do have a Zuiko 50mm 1.4 lens and can try this out in projection.

You can see if your results match published tests. The lens was found to be about average at the center and better than that at the edges. Good lens if you print full frame.

 

[Ivo Stunga]

Thank you for your insights, will be fun to test this out. To elaborate my situation and what I'm hoping to achieve:

Been shooting E6 for a decade now, and DIY BW Reversal for 5 years - all for stupid fun and hobby level. Tried various BW films regardless their base, and am always impressed how 35mm projects effortlessly and fills my 2m screen - some films do it sharper, some less so, but all keep their own look and feel.
Doing research on BW films with clear base I also looked at resolving power figure, if manufacturer provides it, and most films I shot are in the range of 160-260l/mm. But there are films that can do more. Can my lenses, can my projection lens? Can I achieve mere 420l/mm at least? : D

Seeing the stupid figure of 800 got me thinking about my lenses and two use scenario's - will I see the difference in resolution developed in
1) PQ Universal - have it.
2) or Adotech IV only - must source it somewhere.


Thinking being that maybe CMS 20 II Pro reversed in PQ Universal will reach a resolution above the usual ~200l/mm figure I'm used to shooting, which would be cool! Logic behind being - reversal process in itself gives much finer grain than developed normally. Delta 3200 in PQ Universal looks like a ISO400 negative grain-wise.

On lens: before making my purchase I read that Zuiko 50mm Auto-S 1.4 is a good resolver, but one has to hunt down a six-digit serial number to arrive at best production dates/best examples. So I did years ago and surely cannot complain about this sweet lens.
Realistically I'm not even dreaming about the full 800 and could settle with half of it just to try it out and see what it's all about.

You can see if your results match published tests. The lens was found to be about average at the center and better than that at the edges. Good lens if you print full frame.
View attachment 284349

Can you help me reading the graph?

 

[Alan Johnson]

Your aspirations seem a bit on the optimistic side:
https://web.archive.org/web/20040804113631/http://www.imx.nl/photosite/technical/highres.html

 

[RalphLambrecht]

While doing research on films with clear base, I stumbled across the description of Adox CMS 20 II PRO, which resolves up to 800l/mm and recommends this, quote: "If you want high resolution pictures you need to open your lens aperture to one stop below maximum opening. Otherwise the lens diffraction will lower your lens resolution down to half of what this film can capture. Best lenses are F1,4 high speed high quality lenses."

https://www.adox.de/Photo/films/cms20ii-en/


Luckily I do have a Zuiko 50mm 1.4 lens and can try this out in projection. While I'll hunt the film down I wanted to hear on this more from you, as I was under the impression that lenses are best the other way around - about two stops from being max closed.

I feel contradicted. Please uncontradict me.

actually,there is a sweet spot. theoretically, wide open has the highest resolving power bu, lens aberrations will reduce the resolving power.While stopping down, lens aberrations are educed but diffraction is introduced.diffraction lowers resolving power and increases further with stopping down. somewhere in the middle(at about f/8-11 for 35mm)there is an optimum compromise between aberration and diffraction. this 'sweet spot' depends on lens design and film format but f/8-11 is a good starting point.Remember that resolving power and sharpness are different and both overrated when it comes to good images. HCB said"sharpness is a bourgeoise concept" and he knew what he was talking about.

 

[Dan Fromm]

Enough of theory and fantasy. OP, read these:

https://www.flickr.com/photos/nesster/4424744296/sizes/o/

https://www.flickr.com/photos/nesster/4424744224/sizes/o/

Terrifying articles from Modern Photography magazine. They make the point that getting 100 lp/mm on film is extremely difficult.

 

[ic-racer]

Can you help me reading the graph?

The vertical bars represent the range of all the 35mm format lenses tested of that focal length (50mm). The black dot is the Zuiko. So as you see, the Zuiko is about average resolution in the center of the frame but better than most of the others near the corners.

 

[Sirius Glass]

I think I've read that some lenses, perhaps Hasselblad, are just as good wide open as stopped down.

While it may seem that way with Hasselblad lenses, there is some improvement when stopping down. For Hasselblad, look at the MTF curves to get a feel.

 

[NB23]

A good rule of thumb, and logically, would be the middle aperture setting.

This makes f4-f5.6 the sweet spot of most lenses

 

[faberryman]

I am usually more concerned about image degradation from camera shake than from using a non-optimal aperture.

 

[Les Sarile]

I did a similar test using Kodak Techpan @ ISO25 developed in Kodak Tehnidol. I found out that after I conducted my testing of the SMC Pentax-M 50mm F4 macro lens at all apertures that evaluating the results would not be so easy as none of the methods I tried could actually resolve all the detail captured on the film . . .

Resolution testing my SMC Pentax-M 50mm F4 macro lens by Les DMess, on Flickr

As you can see from the optical magnification of the big crop on the rights, none of the methods I used could actually resolve the detail captured on this film

 

[George Mann]

A good rule of thumb, and logically, would be the middle aperture setting.

This makes f4-f5.6 the sweet spot of most lenses

Perhaps, but this is only where sharpness peaks, and not where the optimal sharpness/contrast ratio occurs which I find to be far more important.

For instance, the sharpness of my Nikkor 50mm F2 peaks at f5.6, but its optimal rendering performance is at f8, where both sharpness and contrast are most uniform.

 

[George Mann]

Best lenses are F1,4 high speed high quality lenses."

Most f1.4 lenses aren't even useable until f2!

 

[Paul Howell]

I found that when I did tests that all the lens I tested out resolved Tmax 100, all the lens I tested were better than 200l/mm. I did use a heavy tripod, cable release, lighting was open shade on my patio, film was microfish film I bought in the 90s from EFFE (Spell?) who stated it was 800L/MM. I developed my film as high contrast mircofish as I wanted to count the lines. Used a student's microscope, bright light worked out ok, used lowest magnification. Maybe false results as I did not use the correct microscope. In terms of the Modern Photo article Dan has, I recall that in the 70s Modern tested HC Control another microfish film paired with a low contrast developer which they got more than 100L/MM. The Konia 50 1.7 tested best at F11, the others at 5.6, maybe the difference is that the Konica is an 8 element lens while the others were 5 or 7?

 

[Paul Howell]

Most f1.4 lenses aren't even useable until f2!

That's true, I had a Konica 1.2 and later a Nikon 1.4 for available light, all other times used a slower 1.7 or 1.9 or 2.

 

[Pioneer]

F8 and be there.

Flash helps.

or a tripod.

The best lens I have ever owned is the SMC Pentax M 50m f2. It was small, very inexpensive and every time I pressed the shutter I got all the resolution I needed for a good print. I only used the aperture to adjust the depth of field.

I am now very fond of the Leitz Elmar 50mm f3.5 for the same reasons.

 

[Alan Edward Klein]

My 35mm underwater Nikkor for my Nikonos IVa was the sharpest lens I had, better than my Nikormat 50mm f/2.0 lens. I think it was an f/2.5. I have no ideas what it resolved. But it was sharp.
Scuba 2 by Alan Klein, on Flickr

 

[Paul Howell]

The Sigma 50 1.4 art is as sharp at 1.4 as at 5.6 or 8 or 16, the Nikon AF 1.4 G is similarly as sharp. I have a Sigma 50 1.4 (non art) and a 50 2.8 macro in Minolta A mount, both seem very good wide open, I bought these after I did my testing so no real data.

 

[Ivo Stunga]

Your aspirations seem a bit on the optimistic side:
https://web.archive.org/web/20040804113631/http://www.imx.nl/photosite/technical/highres.html

That might be true and that's a lot to swallow right that article.

Enough of theory and fantasy. OP, read these:
https://www.flickr.com/photos/nesster/4424744296/sizes/o/
https://www.flickr.com/photos/nesster/4424744224/sizes/o/

Thank you! Taken together with this - I figure that in reality I should focus on the inherent contrast in the scene and camera shake to saturate them "mere" 160l/mm films. Let alone 260+

Just did some measurements and my projection setup is:
> Meridian PC 45mm f/2.8 MC lens
> 42mm image diagonal of Gepe 7013 Slide mount with metal mask
> ~2,5m throw distance (measured from slide) that projects ~2,13m diagonally. This gives me what, ~59x magnification right?
> 2,2m viewing distance. Measurements are approximate.
It would also be nice to know exactly what this projection lens can do - is it a limiting factor in this quest for sharper 1,20x1,8m projection or not? Guess I can ignore these figures in the future for sure

The vertical bars represent the range of all the 35mm format lenses tested of that focal length (50mm). The black dot is the Zuiko. So as you see, the Zuiko is about average resolution in the center of the frame but better than most of the others near the corners.

Thanks!

A good rule of thumb, and logically, would be the middle aperture setting.
This makes f4-f5.6 the sweet spot of most lenses

That's what I got from links above too, thanks!

 

[wiltw]

I think I've read that some lenses, perhaps Hasselblad, are just as good wide open as stopped down.

This pair of MTF curves for 90mm f/3.5 lens taken from Hasselblad brochure

...shows damn fine performance wide open, very similar to performance stopped down to f/8

OTOH, the Hassy 80mm does not fare as well when wide open vs. f/5.6

...proving yet again, "It depends!"

 

[AgX]

Can you help me reading the graph?

The dots in this graph (post #6) indicate the value yielded by the very lens.
The collumns indicate the range of values of all tested lenses.

Likely the most common believe is that there is one certain aperture for a certain lens where reduction of lens abberations and rise of diffraction are at equlibrium concerning their effect on imge quality. And that this aperture yields best results at center and samewhat lesser results off center, though there still being the best.

This way of making a graphic though shows well, that this not necessarily so. As for this lens for most subjects one rather would close the aperture one stop more then being the optimum for the center.

But things can be driven even further and not even incliuded in that graphic. As one not only may differenciate between center and off-center, but also between radial and saggital structures..

Whether this fine differantiation is academic or yields significant improvements is to you. But I think it is at least good to know about this, when doing own testing.