How sharp does a lens have to be ?

gliderbee

Well-known
Local time
8:25 PM
Joined
Mar 2, 2008
Messages
750
What I mean is:

I don't know much about the resolving power of the human eye, a computer screen and photography lenses.

Is it possible that, at a certain level, maybe the image is sharper then we can distinguish ?

Just as an example (since I don't know the real figures):

If at a distance of e.g. 10 inch, the human eye can see the difference betwee two lines that are 1/25th inch apart, does it make sense to have lenses that can "see" the difference at the same distance between two lines that are 1/50th inch apart ?

Or IOW, how sharp does "sharp" has to be ?

I came to this question because of the last messages in the R-D1 "mediocre" thread.

Stefan.
 
Lenses can show -sometimes- more detail than we use to see...

Most lenses by most brands are sharp enough for professional work, and just a minimal amount of photographs show all a lens can give technically.

Sharpness is highly overrated: most of the best photographs ever in all fields colud have been a bit sharper if... lots of technical things. But who cares? In fact, they could have been a bit less sharp too, and they'd be just as good.

Everything, really, is a lot more important than sharpness, or the lens, camera, or brand you use.

Too much sharpness is avoided sometimes... Sometimes details can disturb if the overall look and image is more important, as in classic portraiture.

Cheers,

Juan
 
I have one of those Nikkor 50mm lenses and they are indeed very sharp. The broader questions should be, how sharp does lens need to be or how sharp does an image have to be.

Many of HCB's images were fuzzy, so where Cappa's and many other famous photographers. Well maybe not AA's and Weston's but that is a different beast altogether.

If you have a compelling photograph that is well composed etc a bit of fuzziness would be overlooked in most cases, however a tack sharp boring photo is still a boring photo.

IMHO people get to hung up on specs and don't spend enough time learning about composition, lighting etc.
 
What I mean is:

I don't know much about the resolving power of the human eye, a computer screen and photography lenses.

Is it possible that, at a certain level, maybe the image is sharper then we can distinguish ?

Just as an example (since I don't know the real figures):

If at a distance of e.g. 10 inch, the human eye can see the difference betwee two lines that are 1/25th inch apart, does it make sense to have lenses that can "see" the difference at the same distance between two lines that are 1/50th inch apart ?

Or IOW, how sharp does "sharp" has to be ?

I came to this question because of the last messages in the R-D1 "mediocre" thread.

Stefan.

Well I can see a blond human hair on a black leather sofa about 12ft away in fairly poor light; so while you may well be correct about the human eye's limitation I suspect you under estimate the brain's ability to fill in the gaps
 
Depends on the capture too. No point in demanding the very highest resolution when shooting 35mm TriX as its resolving power is pants. It is, however, lovely :)

Overally, I am moving more and more away from sharpness as something that matters to me for 35mm work. All my lenses are sharp enough, although sometimes you would ask for more at certain apertures.
 
If at a distance of e.g. 10 inch, the human eye can see the difference betwee two lines that are 1/25th inch apart, does it make sense to have lenses that can "see" the difference at the same distance between two lines that are 1/50th inch apart ?

Say the lens can only resolve 1/25 of an inch. Now you take a photo. And then you want a print from that photo and you blow it up twenty times, ening up with a print that is roughly 10x14. Now the image will have no detail beyond 0.4 inches, which is fairly large. Grab a photo that's about that big and measure out 0.4 inches on it in various spots, imagine that those things are no longer clearly visible. That's what you'd get.

Of course, if you shot large format, you can contact print your 8x10's and then if the lens only resolved down to 1/25 of an inch, you'd see down to 1/25" on the print.

The lens and the film need to resolve greater than we can see to allow for enlarging after the fact.

There's significant difference in resolution, resolving power, detail, grain, contrast, etc. It's not just one number you can label an image with.
 
This is as sharp as you can get in 35mm film: Zeiss Makro Planar 100/2 and Ilford Delta 100:

1794817250_87eeea846d_b.jpg


And this is about as sharp as it gets on 35mm Tri X - Zeiss Planar 85/1.4

2838455241_6ce70f7674_b.jpg


Personally, I prefer Tri X and I think it is sharp enough for the type of enlargements I do.
 
Well I can see a blond human hair on a black leather sofa about 12ft away in fairly poor light; so while you may well be correct about the human eye's limitation I suspect you under estimate the brain's ability to fill in the gaps

As I wrote: the figures are just an example, as I don't know the actual values; the question remains the same: what's the use of having lenses with a higher resolving power then the human eye ?

Stefan.
 
Say the lens can only resolve 1/25 of an inch. Now you take a photo. And then you want a print from that photo and you blow it up twenty times, ening up with a print that is roughly 10x14. Now the image will have no detail beyond 0.4 inches, which is fairly large. Grab a photo that's about that big and measure out 0.4 inches on it in various spots, imagine that those things are no longer clearly visible. That's what you'd get.

Ok, but you will not look at a picture that is 10x14 from a distance of 10 inch: you'll look at it from a bigger, more comfortable distance, so if the viewing distance doubles, the resolving power can be halved for the same effect.
Of course, if you print on a format that would be viewed from a distance of 10 inch, and you make a crop that would result in a complete print of 10x14, your point is valid.

Let's say that the resolving power of the lens is larger then the resolving power of the film .. Would that be useful in some circumstances ?

Stefan.
 
As I wrote: the figures are just an example, as I don't know the actual values; the question remains the same: what's the use of having lenses with a higher resolving power then the human eye ?

Stefan.

Go back up and read post #7. If your 35 film lens gives you the same sharpness as your eyes, and you are satisified with 35 format contact prints hanging on your wall, you are good to go.

If you want to enlarge however, you need more resolving power because most photographers, much less photo viewers, know or respect viewing distance. And nobody has talked about cropping yet either.
 
Human eyes is said to have a resolution of 1 arc-minute, if rated 20/20...about 0.3mm per metre of distance. The further away a subject, the less detailing you can see. [You cannot see monitor dot pitches finer than 0.25mm if your viewing distance is 32" away.]

Sharpness is a perception...the best description I have read is "...a pleasing combination of resolution and contrast." What pleases me may not please you.

A lens' resolving power is often expressed as line-pairs per millimeter [and evidenced at film plane]...80 lp/mm is considered very good. Each white/black line pair is therefore about 12u. Black/white is maximum contrast, so such a lens might resolve objects >6u?

Film is believed to have a grain size about 3~5u, depends what film and processing, of course. Your grain size may vary...

Printing paper has a resolution of ~20 lp/mm [source: Ilford, long ago.]

A popular resolution target photographers' like to use is human hair or cat's whishers...human head hair is ~60u in diameter, depends on what shampoo/conditioner were used...
 
As I wrote: the figures are just an example, as I don't know the actual values; the question remains the same: what's the use of having lenses with a higher resolving power then the human eye ?

Stefan.

The problem is that the eye and brain are integrated, and a lens and film are not. The eye is a pretty poor lens but the brain that controls its movements and interprets the eye's output is a class bit of kit, think perception rather than resolution
 
I think the sharpness is often confused with micro-contrast. The first one tells you how small detail can the lens record. However it is the contrast on the level of small detail that gives that bitting feeling of sharpness (sometimes film grain boost the visual sharpness if the grain is tight). The Sample from the Zeiss Macro Planar is a great example. You do not see sharpness in a photo of less than 1 Megapixel (well, up to some extend you do), but the micro-contrast that was preserved (and possibly boosted) in the post processing (resizing at least, right) is what you see.

Indeed a sharp AND contrasty lens is a SUPER sharp lens ;)

well, that is how I understand the topic ...
 
This is as sharp as you can get in 35mm film: Zeiss Makro Planar 100/2 and Ilford Delta 100:

1794817250_87eeea846d_b.jpg


And this is about as sharp as it gets on 35mm Tri X - Zeiss Planar 85/1.4

2838455241_6ce70f7674_b.jpg


Personally, I prefer Tri X and I think it is sharp enough for the type of enlargements I do.


that first picture is going to give me nightmares.
 
As with depth of field and perspective, there is a lot of the psychology of vision involved, but distressingly many people don't want to hear that. All they care about is Gradgrind-style 'facts'. Well, here's a fact: the human eye can normally distinguish a smaller discontinuity in a line -- so-called 'vernier' resolution -- than would be suggested by line pairs/mm or visual acuity in seconds of arc. As usual, the answer to all questions about resolution in the final print (lens resolution is a complete red herring unless you are making contact prints) is a firm, unequivocal IT DEPENDS.

One reason why 35mm was so successful was that enlargements were generally just about sharp enough. As expectations about enlargement size increased, lenses and film got better, keeping pace. Or did better lenses and film lead people to expect bigger enlargements?

At around 18-20 high quality (low noise) megapixels, digi can roughly equal 24x36mm film at its best, and out-resolves most 35mm in the real world. Is that adequate for A4, A3, A2? Depends on viewing distance and subject matter -- and how important fine texture is in the pic. With digi, too, there are no losses via an enlargng/projection lens, which is an important point. Enlarge 80 lp/mm x10 with an enlarger and you DON'T get 8 lp/mm.

Cheers,

R.
 


Nikkor 300 2.8 IF-ED AIS on F2AS on Tech Pan 24 ASA.

Nice shot with the Zeiss. However, I think the Nikkor and Tech Pan give it a run for its money :).
 
Last edited:
What surprises me is how sharp (and by sharp, I mean our impression of sharpness as defined by Barry Thornton) some lenses are that are relatively humble. My favorite is the Olympus Stylus Infinity F3.5:

2093631969_076706498c.jpg


Blow up:

2100920857_6f13fd948e.jpg
 
Back
Top Bottom