400 Lines per Millimeter... what does this actually mean

[Rob Skeoch]

Hi guys,

I was looking up something on the Zeiss M 25mm 2.8 on wikipedia.
It says "Zeiss made a milestone in resolution with this lens. 400 Lines per Millimeter."

What does this actually mean?

Is it true? How do we know?

So does this mean this is the only lens to reach 400 LPM?

Does this make the 25mm the sharpest lens in photography?

Of course on a personel note my pictures with this lens might be sharper.... but still boring.

-Rob

 

[BWKate]

I have a Zeiss T* Distagon 25mm 2.8 lens for my Contax 139 35mm camera. Is it similar to the lens you are talking about?

 

[nyx]

I have a Zeiss T* Distagon 25mm 2.8 lens for my Contax 139 35mm camera. Is it similar to the lens you are talking about?

No, this is Biogon design for Leica M bayonet mount.

 

[wirehead]

Well, depends on your point of view.

If you take any other lens designed for the 35mm format and compare it to the lens in question, you will find out that if you photograph resolution charts, it will out-resolve any other 35mm format lens. This is what it means.

However, you will likely still get a higher resolution image out of many 4x5 or medium format setups, because actual resolution just isn't that simple.

Also, the resolution of the lens in an ideal test case doesn't necessarily cover how a lens actually performs in the real world under real photography cases. It could suffer from flare or bad bokeh. Or your particular camera could have the lens mount off in one way or another. Or your particular lens could be slightly misaligned. Or you could be shooting 20asa microfilm and suffer from camera shake. Or your negative holder could be off by a fraction of a mm. Or your enlarger lens could be the weakest link. Etc.

 

[Dan Fromm]

Once upon a time the Modern Photography magazine that was published an article on getting 100 lp/mm on film. Short answer, with a high contrast target, very slow, sharp film, meticulous technique, a very good lens and some luck it can be done. It can't be done consistently.

This old story's relevance to Zeiss' claim of getting 400 lp/mm on film is that in most photographic situations resolutions on film anywhere near that high are unattainable. Until convinced otherwise, I'm going to believe that the real meaning of Zeiss claim is that owners of their wonder lens will have braggin' rights. Also that any capable photographer using perfectly ordinary LF equipment will be able to produce better large prints than the braggarts.

 

[naaldvoerder]

I don't think the abillity to buy this lens is much to brag about, making it is however....

I think the test were done on gigabit film, which apperently is highly resolving itself

Jaap Jan

 

[Jim Jones]

400 lp/mm is certainly high for consumer grade 35mm cameras. A single number doesn't tell all the story, though. Does the lens attain this resolution only on-axis? A Nikkor 20mm f/3.5 I used decades ago was extremely sharp in the center of the field, but a few magnitudes less sharp in the corners. Is it determined by measuring the aerial image or (less likely) on film? Is it determined in white light or monochrome light? If monochrome, what color? What is the contrast of the target? Remember, wikipedia articles may not be reliable. Even I (perish the thought!) could submit one.

I don't know what the resolution requirements for integrated circuit masks are, but they must be higher than 400 lp/mm.

 

[rfshootist]

Hi guys,

I was looking up something on the Zeiss M 25mm 2.8 on wikipedia.
It says "Zeiss made a milestone in resolution with this lens. 400 Lines per Millimeter."

What does this actually mean?

-Rob

Actually nothing. At least not in practical life.
Only if photographing test patterns is your hobby, shooting TechPan on a tripod and developing it to 100% contrast.
Zeiss feeds those nerds who TALK about lenses instead of using them. In other words the majority in the crowd of the "serious" amateur photogs !

Regards,
bertram

 

[Dorian Gray]

Resolution alone is not a very useful metric for comparing lens sharpness. MTF is vastly more important. All normal photographic films (excluding microfilms, etc.) have an MTF response of zero at 400 lp/mm. Likewise, the aerial image of the Zeiss lens itself probably has an MTF very close to zero at 400 lp/mm. Consider that 400 lp/mm is around the diffraction limited resolution at f/4. So the lens had to produce this figure somewhere between wide open and f/4. It's very likely that at these f-stops the lens has substantial aberrations. Wavefront error alone would cause pretty low contrast response at these f-stops.

So even if the lens resolved 400 lp/mm in the aerial image, it might have lower contrast response at 20 lp/mm than another lens incapable of resolving 400 lp/mm. As the contrast transfer at 20 lp/mm is actually useful for determining how sharp a photograph will look (on 35 mm film at least), the 400 lp/mm figure seems to be a bit of red herring: useful insomuch as it sounds good to the Zeiss marketing department, but I bet the engineers are none to happy about the claim.

This is not to say that the lens isn't sharp. In fact, the MTF charts (PDF here) are outstanding. But you'll notice the lens is sharper (or has better contrast transfer at 10, 20 and 40 cycles/mm) at f/5.6 than at f/2.8, proving the pointlessness of the 400 lp/mm claim: at f/5.6 the lens is theoretically incapable of resolving 400 lp/mm, yet performs better in the real world at this aperture than wide open.

 

[Roger Hicks]

Actually nothing. At least not in practical life.

Exactly. No useful film can resolve it; no camera can hold the film flat enough, repeatably enough, to record it, even if the film could.

Yes, it's a super-sharp lens. You should be able to get 100-125 lp/mm on the film quite often -- and there are damned few lenses that can do that.

Otherwise, this is Wikipedia at its best: the blind leading the blind, by blinding them with pseudo-science and 'factoids'. A 'factoid' is a statement that purports to be factual and may even be defensible (as this is for aerial resolution), but is neither adequately qualified nor actually demonstrable.

Cheers,

R.

 

[Dorian Gray]

Roger, I'd be the last to defend the academic rigour of Wikipedia, but in this case I vaguely remembered Zeiss themselves making this claim. So I went and Googled it, and lo and behold, Zeiss actually have the audacity to claim they captured 400 lp/mm on film! Here you go. Apparently achieved with some kind of copy film processed in an unusual manner, a combination which allegedly delivers 800 lp/mm at a subject contrast ratio of 1000:1 (see PDF datasheet on this page). Never mind that the characteristic curves of that lot would probably give one vertigo!

Still quite irrelevant in the real world, if true at all.

PS. Loved Perfect Exposure. Actually have a signed hardback among the handful of photography books I own. It must be as strange for you to know your signature is in loads of strangers' bookshelves as it is for me to bump into the author on a forum... Assuming you're not another Roger Hicks, that is!

 

[jd callow]

kind of like the 'wind chill factor.'

 

[richard ide]

I have an enlarging lens which resolves over 400 lp/mm according to my microscope and allowances for not using a proper resolution test target. I have a Kodak film which resolves 800 lp/mm. In theory I should be able to resolve 300+ lp/mm on film. Hopefully I can do my tests in the next couple of weeks and post the results when I am done.

 

[Roger Hicks]

, Zeiss actually have the audacity to claim they captured 400 lp/mm on film! Here you go. Apparently achieved with some kind of copy film processed in an unusual manner...

I stand corrected. I'd quite forgotten the claim until you reminded me, because it was so useless. I did cover my arse by saying 'no useful film' and 'no camera could hold it flat enough'; the 'float' in a 35mm camera film gate is what limits sharpness. But no doubt Zeiss built a test rig...

I did remember their claim to get 200 lp/mm in camera, but they admitted you had to use 'focus bracketing' to get it, i.e. refocusing slightly for several shots. Apparently for a given film and camera this will allow you to see 200 lp/mm sometimes -- but again, you have to ask why, and what it means.

Thanks for the kind words about the book; if you take a look at www.rogerandfrances.com you'll see what I'm increasingly doing nowadays instead of books.

Cheers,

R.

 

[bjorke]

What does this actually mean?

That someone has forgotten waht it means to make pictures.

Isn't the limit for handheld typically around 30-40l/mm?


BTW, some compact digis are probably pushing 300 lines/mm

~12mm/~3000 sites/12mm sensor = 250l/mm and I recall that some sensors are smaller and higher-res

They are NOT, however, large lenses -- holding a level of sharpness at sizes appropriate for 35mm or larger would be a far more difficult lensmaking task.

If rumours about the new high-end Canon SLR are correct, it will be a bit less than 6K*4K pixels or for 24x36 framing, ~175l/mm

This is WAY high for an SLR and I expect that few SLR lenses can approach that. Seems kind of pointless beyond that point to talk about greater precision give the limitations of the 35mm-style format.

 

[jd callow]

Kevin,
I believe that l/mm requires at least 2 samples to create the line so a (6k/36mm)/2 or 166.7/2 = ~83 lines per mm. If I'm wrong I'm sure someone will correct me.

I do agree wholeheartedly that the practical use of the information is pretty minimal.

 

[HerrBremerhaven]

The lines/mm came from video usage of lines/height, and is often used on newer D-SLRs. What is missed is that actual resolution capability has nothing to do with photosites nor file sizes. There are dead areas on imaging chips, unlike film which does not have dead areas.

On the subject of handheld photography, it is rare to do much better than 40 lp/mm (line pairs/mm) regardless of technique or camera gear. While there are some things that can help that out a bit, a tripod is the only consistantly useful way towards higher resolution. There is nothing magical about image stabilization nor imaging chip that suddenly changes that; handheld photography means accepting lower resolution. Of course, the reason to use smaller cameras is that we like shooting handheld with them.

With Zeiss, and the high capability of their lenses, even they point out that the reason to even state this is to point out that their lenses will not be the greatest limiting factor in your photography when you want high resolution. Unfortunately, many on internet forums and in publications jumped all over part of the Zeiss statement, without understanding the context and implication (my opinion). It would not be surprising if other lenses showed a great capability too (from other companies), yet our practical limit should be near 40 lp/mm.

Ciao!

Gordon Moat
Dead Link Removed

 

[rfshootist]

T, yet our practical limit should be near 40 lp/mm.
Ciao!
Gordon Moat
Dead Link Removed

This was what an old master answered to me decades ago when I asked him, why in his opinion tripods should be welded to camera bottom plates

A truth not very popular tho and therefore seldom mentioned, spoils all the party for the many gear heads among the "serious " amateurs.

Regards
bertram

 

[Dan Fromm]

<snip>
With Zeiss, and the high capability of their lenses, even they point out that the reason to even state this is to point out that their lenses will not be the greatest limiting factor in your photography when you want high resolution. Unfortunately, many on internet forums and in publications jumped all over part of the Zeiss statement, without understanding the context and implication (my opinion). It would not be surprising if other lenses showed a great capability too (from other companies), yet our practical limit should be near 40 lp/mm.
Dead Link Removed

Y'know, Gordon, I usually shoot at effective apertures of f/11 and smaller. At effective f/11, its possible that the film, given that my technique is meticulous, limits the negative's enlargeability. Possible at effetive f/16 too. But from effective f/22 down, diffraction kills all lenses equally and sets the limits.

I mention effective aperture, not aperture set, because I do a fair amount of closeup photography.

All of this assumes good technique. The big idea that leaps out of the Modern Photography article I mentioned earlier in this thread is that small format shooters are usually pretty casual about focusing perfectly and are absolutely sloppy about motion control. MP had to bracket focus to get really good resolution and went to considerable effort to control motion. Camera motion at high frequency (mirror locked up) and low (on a very sturdy tripod) and subject motion.

Shooting handheld is a recipe for low resolution. Most of us trade off resolution in the plane of best focus, wherever it might be, for depth of field. And many of us shoot low resolution emulsions and so can't get the best our lenses can give. And, as LF shooters like you never fail to remind the world, what matters is resolution in the final print, not resolution on film.

All of which are reasons to pay more attention to technique and to the desired result than to lens tests.

Cheers,

Dan

 

[HerrBremerhaven]

Hello Dan,

Definitely agree on many points. As a large format shooter, I would hesitate to claim more than 60 lp/mm from any shots, though I feel that is a good working target. Based on the average viewer eyesight being able to resolve maybe 5 lp/mm (some better, some worse), that would give a 12x enlargement capability. It might be of interest that commercial printing has long stated a preference for 300 ppi image files, which oddly enough equate to near 5 lp/mm.

I do shoot quite a bit of 35mm images. It is rare that I would have any of those printed larger than 10" by 15", and I think smaller prints have a better chance of avoiding softness. I do have hanging two large 24" by 36" B/W (silver) prints done from 35mm Kodak TriX. Both of those look quite good at that size, and get lots of compliments, but I feel those are the exceptions. It is possible to achieve high resolution images, but not practical to consider it easy.

Ciao!

Gordon Moat
Dead Link Removed

 

[rfshootist]

The big idea that leaps out of the Modern Photography article I mentioned earlier in this thread is that small format shooters are usually pretty casual about focusing perfectly and are absolutely sloppy about motion control.
Shooting handheld is a recipe for low resolution. Most of us trade off resolution in the plane of best focus, wherever it might be, for depth of field. And many of us shoot low resolution emulsions and so can't get the best our lenses can give. And, as LF shooters like you never fail to remind the world, what matters is resolution in the final print, not resolution on film.

All of which are reasons to pay more attention to technique and to the desired result than to lens tests.

Cheers,

Dan

How true ! But facts which are mostly ignored anyway. These words should be tatooed on the forehead of all these test chart nuts , and so they would be reminded to the TRUTH each morning while brushing their teeths !!

Regards,
bertram