Very high resolution lens?

[Q.G.]

?
There are interesting things to discuss we may not know about?
Yet we should be told that you two discuss them in private?

 

[Henning Serger]

Hello Ralph,

probably the difference is because of the following reasons.

1. ....

2. .....

3......

Best regards,
Henning

Let me just add another relevant points:

4. The resolution values I have described refer to our tests with the new Zeiss Makro-Planar 2/50 ZF. This lens surpassed all of my other very good prime lenses so far. Probably it is one of the best 35mm lenses on the market.

Have a look at these data:

http://www.zeiss.de/C12567A8003B8B6F/EmbedTitelIntern/MP_2_50ZF_DE/$File/MP_2_50ZF_DE.pdf

http://www.photozone.de/nikon--nikk...zf-makro-planar-t-50mm-f2-review--test-report

5. With BW film, developers have visible influence on resolution. Not huge, but visible. Due to our tests the best developers concerning resolution deliver about 5-10 % higher resolution values.

6. To achieve better focus accuracy, a test chart with integrated Siemensstern is very helpful. The Siemsstern gives much better indication of focus accuracy than the line patterns.
A viewfinder magnifier (2x loupe) is also very helpful to improve focus accuracy.

Best regards,
Henning

 

[Henning Serger]

??? Are we talking about LENS resolution, or FILM resolution?

Hello Ed,

to be precise: We are talking about system resolution. Aron tested his 55m Fuji lens in combination with Ilford Delta 100.
I have tested the Zeiss 2/50 (and other lenses) in combination with Delta 100 (and lots of other films). And with other films, the resolution values differ, of course. In each case a more or less different system resolution dependant on the film used.

Best regards,
Henning

 

[Ed Sukach]

to be precise: We are talking about system resolution. Aron tested his 55m Fuji lens in combination with Ilford Delta 100.
I have tested the Zeiss 2/50 (and other lenses) in combination with Delta 100 (and lots of other films). And with other films, the resolution values differ, of course. In each case a more or less different system resolution dependant on the film used.

The original title was "Very High Resolution LENS" , not "Camera" or "System".

Of course, far more often than not, we will use our lenses in cameras, but I think it is a grave error to judge the value of a LENS from a test with so many potential biases possible from an entire SYSTEM ... and this is done all too frequently in photography. I am NOT trying to be anally "pecise", just clinging, hopefully, to reality.

 

[Henning Serger]

The original title was "Very High Resolution LENS" , not "Camera" or "System".

Of course, far more often than not, we will use our lenses in cameras, but I think it is a grave error to judge the value of a LENS from a test with so many potential biases possible from an entire SYSTEM ... and this is done all too frequently in photography. I am NOT trying to be anally "pecise", just clinging, hopefully, to reality.

Hello Ed,

reality for us photographers in most cases is that we are using our lenses in combination with film. A lens designer will be interested in the fact that an excellent prime lens at f4 can have an aerial resolution of about 400 Lp/mm (near its diffraction limit).
A photograper probably is not interested in this single value.
He wants to know what will be on the film in the first step, and on the print in the second step.

And that is what Aron is interested in: He has a lens he likes, and with Delta 100 a film he likes, and he wants to know what resolution at which aperture he can achieve with his preferred combination.
He started a test, and he has asked us whether his results may be reasonable.
For his photographic reality that is a question he is interested in. And there is nothing wrong with that I think.

Best regards,
Henning

 

[RalphLambrecht]

??? Are we talking about LENS resolution, or FILM resolution?

In LENS testing (at least in the old, Luddite, Optical Bench way), no film or ground glass is used: the microscope is focused on the aerial image - much the same as "Grain Focusing" in the enlarger.
Introducing film into the testing provides the opportunity for a number of elements that will invariably degrade the results: flm flatness, emulsion/ developing charactersitics, mechanical "shake" of the support system ...

Is it possible to see exteremly high resolution in any area of any given lens? Absolutely, but maximum resolution, by itself, is an incomplete test of the performance of a lens.

My tests include the lens, film, developer, agitation...

I should call it negative system resolution.

 

[Keith Tapscott.]

I've made two tests using one of my recently acquired cameras (Fujica ST801 with EBC Fujinon 1.8/55) on the resolving power of the lens. For the first test I used a moderate resolution film (Fomapan 200) and the results showed actually the resolution of the film (110 lp/mm) and not lens.

Today I made new tests using Delta 100 developed in Rodinal 1:25. For the illumination of the test chart I used a 1000W halogen lamp, so contrast was high.

These are my results (lp/mm)

Aperture: centre, very close to edge, edge
f1.8: 85, 57, under 50
f2.8: 156, 69, under 50
f4: 161, 89, under 50
f5.6: 147, 101, 55
f8: 115, 101, 69
f11: 69, 55, 69
f16: 69, 69, 69

Do you think I was doing something wrong during this test or it is actually possible to achieve such a high resolution with this setup? The lens is a 'Planar' type design.The prints from this camera do look very sharp (to me), even heavily cropped ones taken at f1.8. I'm somewhat surprised. :rolleyes:

It also is somewhat interesting for me (I used a Tessar lens for a long time), that diffraction already takes its toll at f5.6-f8, most notably at f11.

I think that the sales of second-hand Fujica ST801 Cameras with an EBC Fujinon 1.8/55 lens have risen dramatically since you have made this post.

 

[RalphLambrecht]

The figures look to be about twice what one would expect. How are you counting lp/mm?

I agree. After checking my tests again, these numbers look about twice as high as mine too. Henning has invited me to see his results, and I may do so, since he is not too far away and sounds like a credible guy, but until then, I'm with Nicholas on this one.

 

[dynachrome]

Years ago a car's horsepower was measured at the flywheel in bhp (brake horsepower). In the U.S. this was later changed to SAE (Society of Automotive Engineers) Net. The SAE Net figure is a more useful figure because it takes into account friction and other loss of power to the wheels. My father's 1968 Dodge Coronet 440 station wagon with the 383 V8 engine had a 2 barrel carburetor and was rated at 290 bhp. My 2008 Honda Odyssey has a 3.5 Liter V6 (210 ci) with fuel injection. It's rated at 244 hp SAE Net. I was 16 when I first drove the Dodge and I can tell you it was much faster than my Honda. If I have to compare the two, the Honda is a much more civilized vehicle.

Every time I see a lens test I wonder what conditions it was done under. Was the contrast 1000:1 or 50:1? Was a tripod used or was it a bean bag? Was the mirror locked up if an SLR was used? Was a film like TP or Imagelink used or was a less sharp and more grainy film used? Were the results interpreted by scanning or other digital means? When I think of a lens like the Zeiss 50/2 macro I'm sure it's quite good. The question for me is whether it is enough of an improvement over a lens like the 55/2.8 Micro Nikkor to make a difference for anything I might do. Very little close-up or macro work is done at f/2. The exception might be some work with micro lenses. These are actually optimized for use at or near wide open.

If I shoot hand held and use Tri-X or TMY2 or Plus-X, will I get these same test results? Will I ever be able to get results like the ones mentioned in practical use? If I just need to make a large print I'll use a 100/3.5Zenzanon PG on a Bronica GS-1 with Pan F+. You can use any other film you like with any Zeiss, Leica, Nikon or Canon lens on a 35mm camera. We'll make 11X14s, 20X30s or even larger prints. The Bronica shots will be better every time. How much was the 100/3.5 Zenzanon? Less than $75. I'm sure it cost a lot more new but some of these lens tests with 35mm equipment are not very useful. If you need to make a large print, use a larger format.

 

[Keith Tapscott.]

Every time I see a lens test I wonder what conditions it was done under. Was the contrast 1000:1 or 50:1? Was a tripod used or was it a bean bag? Was the mirror locked up if an SLR was used? Was a film like TP or Imagelink used or was a less sharp and more grainy film used?
Were the results interpreted by scanning or other digital means?

When I think of a lens like the Zeiss 50/2 macro I'm sure it's quite good. The question for me is whether it is enough of an improvement over a lens like the 55/2.8 Micro Nikkor to make a difference for anything I might do. Very little close-up or macro work is done at f/2. The exception might be some work with micro lenses. These are actually optimized for use at or near wide open.

If I shoot hand held and use Tri-X or TMY2 or Plus-X, will I get these same test results? Will I ever be able to get results like the ones mentioned in practical use? If I just need to make a large print I'll use a 100/3.5Zenzanon PG on a Bronica GS-1 with Pan F+. You can use any other film you like with any Zeiss, Leica, Nikon or Canon lens on a 35mm camera. We'll make 11X14s, 20X30s or even larger prints. The Bronica shots will be better every time. How much was the 100/3.5 Zenzanon? Less than $75. I'm sure it cost a lot more new but some of these lens tests with 35mm equipment are not very useful. If you need to make a large print, use a larger format.

I can`t disagree with that. 35mm is still a good format though.

 

[Andrew Sanderson]

Taken any pictures lately?
OK, that was perhaps a little flippant, but photography is about more than resolution and LPmm.
It's easy to get too obsessed with testing, rather than looking.
Andrew.

 

[RalphLambrecht]

Here is another sobering thought:

The equation for system resolution is 1/R^2=1/r1^2+1/r2^2+...1/rn^2

What that means is that the resolution of the system is not equal to the lowest resolution of any component in the imaging chain, but lower than the lowest component, following above equation.

Here is an interesting example:

1. 35mm camera lens = 90 lp/mm
2. film = 200 lp/mm
3. enlarger lens = 120 lp/mm

makes a system resolution of 67.7 lp/mm so far, but we must divide that by 8.5 to cover the magnification to make an 8x10 print, and end up with another system

4. projected image resolution = 8.0 lp/mm
5. paper resolution = 100 lp/mm

to get the total system resolution from camera lens to print of 7.9 lp/mm, which is OK and about what healthy eyes can resolve.

Now let's change one thing, we are going to use a 150 lp/mm camera lens. Our total system resolution will change to 9.9 lp/mm. A 25% improvement. Not bad.

But how about this? Leave the camera lens as is at 90 lp/mm but make it medium format, now only needing a 4.5x enlargement to make an 8x10 print. The total system resolution goes up to 14.9 lp/mm. A 89% improvement. That's more likel it.

Why are we discussing this?

Forget unrealistic lens tests, which assume ideal conditions, high contrast targets and developers which are optimized for sharpness and not for tonality. Switch to medium format instead.

 

[Keith Tapscott.]

Taken any pictures lately?
OK, that was perhaps a little flippant, but photography is about more than resolution and LPmm.
It's easy to get too obsessed with testing, rather than looking.
Andrew.

Taking photographs of good subjects is much more fun than shooting Rez-Charts.

 

[Film-Niko]

Every time I see a lens test I wonder what conditions it was done under. Was the contrast 1000:1 or 50:1?

No one needs to wonder about tests with contrast of 1000:1 made by photographers using printed test charts.

Because there are no printed test charts with such a high contrast. It is physically impossible. With printing dyes on paper it is impossible to get contrasts of 1000:1.
And no lens can transfer contrasts of 1000:1 completely. Look at the scientific report of the chief optic designer of Zeiss, Dr. Nasse, published in camera lens news 30. He explaines there exactly that.

Film manufacturers are doing tests with contrast of 1:1000 (and 1:1,6). But they don't do that with printed test charts and lenses. The expose the film in contact copy with resolution patterns, using intense direct light. Thats a complete different thing.

On the market so far I have only seen test charts for photographers with object contrast of 1:10, 1:20 or 1:30. All low or medium contrast.

 

[RalphLambrecht]

No one needs to wonder about tests with contrast of 1000:1 made by photographers using printed test charts.

Because there are no printed test charts with such a high contrast. It is physically impossible. With printing dyes on paper it is impossible to get contrasts of 1000:1.
And no lens can transfer contrasts of 1000:1 completely. Look at the scientific report of the chief optic designer of Zeiss, Dr. Nasse, published in camera lens news 30. He explaines there exactly that.

Film manufacturers are doing tests with contrast of 1:1000 (and 1:1,6). But they don't do that with printed test charts and lenses. The expose the film in contact copy with resolution patterns, using intense direct light. Thats a complete different thing.

On the market so far I have only seen test charts for photographers with object contrast of 1:10, 1:20 or 1:30. All low or medium contrast.

Very good point.

This note got me to measure my test chart contrasts. I have a few, but the one I use the most is the large Resolving Power Chart No. 83.001 from the Edmund Scientific Company. It surprised me with a black relative density reading of 2.07, which means its contrast is about 120:1. I didn't think one can get that from ink, but then again, inkjet printers deliver up to 2.4 density (250:1). All my other tests charts max out at 30:1, just as Film-Niko said. The ones I have from RIT measure as advertised at 32:1, 6.3:1 and 1.6:1.

 

[Film-Niko]

Here is another sobering thought:

The equation for system resolution is 1/R^2=1/r1^2+1/r2^2+...1/rn^2

What that means is that the resolution of the system is not equal to the lowest resolution of any component in the imaging chain, but lower than the lowest component, following above equation.

Here is an interesting example:

1. 35mm camera lens = 90 lp/mm
2. film = 200 lp/mm
3. enlarger lens = 120 lp/mm

There is a big mistake in your calculation: The formula you use is designed for calculation with aerial resolution of lenses. To get the system resolution of your total system as the result in this formula you have to use the aerial resolution of the lenses involved in the imaging chain as components in the formula.
But your values of 90 lp/mm and 120 lp/mm are already values of system resolution, so you get wrong results.
Lenses with only 90 lp/mm aerial resolution are crap. No 35mm lens of the main manufacturers has such low res values.

Example: At f4 a good prime lens has an aerial resolution of 200-300 lp/mm, a very good lens has about 350 lp/mm, and an excellent lens has nearly 400 lp/mm (it is diffraction limited; there are quite a few lenses from Nikon, Canon, Leica, Zeiss, Pentax, Schneider...which are excellent designed and only diffraction limited).
The better enlarging lenses, Apo-Componon and Apo-Rodagon, are such lenses which have outstanding resolution and are diffraction limited.

Why are we discussing this?

Forget unrealistic lens tests, which assume ideal conditions, high contrast targets and developers which are optimized for sharpness and not for tonality. Switch to medium format instead.

I am using test charts with contrast of 1:20. That is low to medium contrast, like the things I normally photograph, contrast like in reality.
All other photographers I know which are doing lens and film or sensor tests are using charts in the 1.10 - 1:30 contrast range. No one makes high contrast tests.

Developers: Why either resolution or tonality? I am using developers which give me both excellent tonality and high resolution / fine grain. For example Microdol 1:3, Spur HRX-3, CG 512, Moersch Finol.

These statements like "don't use 35mm, but medium format instead" are not helpful for Aron, the original poster.
He surely has very good reasons using his 35mm gear.
Horses for courses. A good photographer know that for some purposes 35mm is the way to go, for others MF or LF.
I can't imagine using MF for my wildlife photography, I will fail. 35mm is best for that, no qestion.

But in the situations where 35mm is best, of course we want the best lens and film quality for our pictures (the same we want in the other formats).
I want best technical picture quality in my wildlife, street or portrait photography.

And to get best quality, of course it makes sense to test lenses and films.
After the test we know what is good and what is working for us. And with this lens(es) and films we can make excellent pictures for many years to come.

Therefore Aron, go on. There is nothing wrong with your decision to test your lens and films.

Back to your question: I don' have your Fuji lens. But I have made tests with Ilfords Delta 100 as well and my 50mm lenses. With object contrast of 1:20.
At f4 and f5,6 I got resolution values in the 110 - 140 Lp/mm range with my primes.
My results are not so much different from yours.

 

[RalphLambrecht]

Film-Niko

There is nothing wrong with that equation. You are right about aerial resolutions, but the numbers I enter were just for comparison But here is the calculation again, this time with much higher values.

resolution
[lp/mm]

camera lens 300.0
film 200.0
enlarger lens 300.0
aerial image 145.5

print magnification 8.5
paper 200.0
total system 17.1

The point was and remains true, a jump from 35mm to medium format gives a gigantic increase in image resolution, due to the smaller print magnification required. This by the way is anybody's experience who ever moved up from 35mm to medium format. The same is unfortunately not true for the next jump to large format, because we can't appreciate the quality increase at regular print formats.

 

[RalphLambrecht]

... your values of 90 lp/mm and 120 lp/mm are already values of system resolution ...

At f4 a good prime lens has an aerial resolution of 200-300 lp/mm, a very good lens has about 350 lp/mm, and an excellent lens has nearly 400 lp/mm. ...

Well, the equation can be solved for any of the components as well, given that we have a lens/film system resolution of let's say 90 lp/mm. The trouble is, if I enter your lowest recommended value for the lens (200 lp/mm), the film would be at a mere 100 lp/mm to get to a system resolution of 90 lp/mm.

Do you think that is realistic for Tmax-100 and a 120:1 contrast ratio target?

 

[RalphLambrecht]

...
2. If we photograph a test chart in a distance of 2 or 3 meters, absolutely exact focussing is necessary to achieve more than 100 lp/mm resolution. The focus must be exactly down on the millimeter.

The problem with focusing is, that we don't have really exact focusing systems. Neither our manuel focusing systems (split image, matte screen, microprism) nor the AF systems deliver the precision we need to exploit the capabilities of our lenses and films.

Therefore we have to do a simple, but effective trick: "Focus bracketing". Shooting several frames, each frame newly focused with a sligtly different focus. We make a series and choose the pictures were the focus is dead on.

When I do my lens and film tests, e.g. with TMX, then always some negatives have resolution values of only 80 or 90 lp/mm. Because the focus is a bit in front of or behind the flat test chart.
But some shots of the test series are absolutely precise in focus, because of the "focus bracketing"...
Henning

Henning

I followed your suggestion and did a test with something similar to your focus bracketing, but I just staggered five mini USAF charts behind each other, 20 mm apart and focused on the one in the center. This way I have a variety of focus distances to look at. The test setup was for an 85mm lens at 25x reduction (2298 mm from film to subject).

I was shocked to see that at f/2.8 my manual focus was totally off. Autofocus was much better but not fully reliable. I think AF was confused by these targets being so close together orthographically but so far apart in depth. So, yes, you are absolutely right, focusing makes a big difference in lenst testing!

The same test at f/8 was different. Depth of field masked almost all focus errors. There is hardly any resolution difference between the test targets. This is not a surprise to me. Even, when assuming a very conservative CoC of 0.007 mm for 35mm (typically 0.022 mm), you still get a depth of field of about 75 mm around the subject plane (almost exactly the max distance between my targets). Of course, for the f/2.8 test above, depth of field was reduced to 25 mm. No surprise that this is harder to get into focus.

You said in your post, 'down to the mm', but I think 'down to 0.1% of focus distance' is more like it. Anyway, focusing is still a big issue if we don't want to do our lens testing at the threshold of sharpness. Thanks for the insight.

 

[Film-Niko]

The point was and remains true, a jump from 35mm to medium format gives a gigantic increase in image resolution, due to the smaller print magnification required.

Ralph, in general you are right, of course.
But, there is no rule without an exception.

During the last months I have worked a lot with high resolution films in 35mm:
Adox Ortho 25, Rollei Ortho 25, Kodak Imagelink, Adox CMS 20, Rollei ATP and Spur DSX 64.

With the dedicated developers (especially with those from Spur for CMS 20, ATP and DSX) I got outstanding results:
Very good to excellent tonality, extremely fine to unvisible grain, outstanding resolution and sharpness.
I've made very large prints and showed them my photographer friends (all of them with experience in medium format, some doing LF as well):

All of them thougt the prints were made from 120 roll film! But it was 35mm.
With these films and very good lenses you can achieve a technical picture quality which is comparable to medium format.
But you have the flexibilty of the 35mm system.
A very nice combination .

 

[Film-Niko]

Do you think that is realistic for Tmax-100 and a 120:1 contrast ratio target?

Ralph, it is difficult to say, because I have never used a test target with a 120:1 contrast ratio. 1:30 was my maximum contrast so far.
With TMX I've got similar resolution numbers as with Delta 100. Not much resolution difference, only TMX being a bit finer grained.
And probably my primes are in the 300-400 lp/mm aerial resolution range.
And important: AFAIK this formula is not suitable for exact calculation, it is a rule of thumb.
If you look at the scientific research results of Zeiss (Mr. Nasse) you see that he achieved higher resolution values with TMX that you would expect based on the formula.
Probably this formula is a bit conservative, and resolution values are a bit higher in reality.

 

[RalphLambrecht]

Ralph, in general you are right, of course.
But, there is no rule without an exception.

During the last months I have worked a lot with high resolution films in 35mm:
Adox Ortho 25, Rollei Ortho 25, Kodak Imagelink, Adox CMS 20, Rollei ATP and Spur DSX 64.

With the dedicated developers (especially with those from Spur for CMS 20, ATP and DSX) I got outstanding results:
Very good to excellent tonality, extremely fine to unvisible grain, outstanding resolution and sharpness.
I've made very large prints and showed them my photographer friends (all of them with experience in medium format, some doing LF as well):

All of them thougt the prints were made from 120 roll film! But it was 35mm.
With these films and very good lenses you can achieve a technical picture quality which is comparable to medium format.
But you have the flexibilty of the 35mm system.
A very nice combination .

Niko

I have not used anything but 400 ASA film for years, except for testing purposes. I wouldn't know what to do with ASA 25. That film speed does not suit my type of photography.

 

[RalphLambrecht]

Ralph, it is difficult to say, because I have never used a test target with a 120:1 contrast ratio. 1:30 was my maximum contrast so far.
With TMX I've got similar resolution numbers as with Delta 100. Not much resolution difference, only TMX being a bit finer grained.
And probably my primes are in the 300-400 lp/mm aerial resolution range.
And important: AFAIK this formula is not suitable for exact calculation, it is a rule of thumb.
If you look at the scientific research results of Zeiss (Mr. Nasse) you see that he achieved higher resolution values with TMX that you would expect based on the formula.
Probably this formula is a bit conservative, and resolution values are a bit higher in reality.

The formula fits well with my experience and testing.
Why are you quoting aerial resolutions which are beyond the hypothetical diffraction limits? At f/8, diffraction limits the lens to 158 lp/mm for red light and 185 lp/mm for green light. This may fit into a research paper but has little to do with practical photography.

 

[Q.G.]

I have not used anything but 400 ASA film for years, except for testing purposes. I wouldn't know what to do with ASA 25. That film speed does not suit my type of photography.

Same here.

Two thoughts came to mind when reading "All of them thougt the prints were made from 120 roll film! But it was 35mm. [etc]"

One of them was exactly that: the flexibility of 35 mm format, but at ISO 25?
What flexibility would that be?
Being able to put a small camera with small lenses on a tripod, instead of a slightly larger camera with slightly larger lenses?

The other was: imagine you had indeed used the same film in an MF camera ...

 

[RalphLambrecht]

Same here.

Two thoughts came to mind when reading "All of them thougt the prints were made from 120 roll film! But it was 35mm. [etc]"

One of them was exactly that: the flexibility of 35 mm format, but at ISO 25?
What flexibility would that be?
Being able to put a small camera with small lenses on a tripod, instead of a slightly larger camera with slightly larger lenses?

The other was: imagine you had indeed used the same film in an MF camera ...

Good point! Most 35mm films come in 120 as well. Using them in a MF camera must be like cooking with gas.