[troyz]

This is the nice thing: A reverse-coupled lens is used exactly as it was designed for: at infinite focus. So please, what are the darn good reported MTF measurements (I need > 500 lp/mm)?

I see what you're getting at: instead of taking a subject at infinity and forming an image at the film plane, the reversed camera lens takes a subject at what was previously the film plane and forms an image at infinity (which becomes a subject at infinity to the camera-mounted lens). It's a plausible argument if the lens attached to the camera is focused precisely at infinity, the subject is precisely at the focal plane of the reversed lens, and the two lenses are perfectly aligned.

Unfortunately, my statement about "darn good MTF measurements" could be misinterpreted as "precise MTF measurements"; what I really meant is "favorable MTF measurements". The center MTF measurements I'm thinking of (see for example, the Pentax DA21 from photozone.de) aren't even close to 500 lp/mm you specify, and have all the shortcomings the previous posters have mentioned.

[falconeye]

The center MTF measurements I'm thinking of (see for example, the Pentax DA21 from photozone.de) aren't even close to 500 lp/mm you specify.

Well, the DA21 has excellent center resolution. But not a stellar one. What you can see by the fact that it has its best resolution at f/5.6, i.e., not even f/4.

A 500 lp/mm lens on a K10D as used by photozone.de, however, would measure at only about 1/(1/82.5 + 1/500)=70.6 lp/mm (or 2225 LW/PH). According to an old Kodak formula which I don't trust. But you get the idea. Any photozone figure very close to the sensor resolution (2592 LW/PH for the K10D) must indicate a lens with stellar performance (e.g., the 30mm Sigma is 95% close!). Another indicator is a lens with best resolution below f/4. Because it means that already below f/4, all abberations except diffraction have almost disappeared.

Both indicators (LW/PH close to sensor resolution, best resolution at low f) can be deduced from photozone. But their choice is limited and it looks like the best lenses from this choice are Zeiss and Canon 24/1.4L.

So, I am looking for more resolution databases to extend my search. Esp. to rangefinder or old glass.

[Wheatfield]

Have you looked into enlarging and process lenses?
Many of the Schneider and Rodenstock lenses are excellent, as are many Nikkors and Fujinons. I believe this class of lens might be better optimized to your needs.

[falconeye]

Have you looked into enlarging and process lenses?

Good idea! Thanks.

A first result is this: I found MTF figures for Rodagon here:
Paul Butzi Photography

All Rodagon, Grandagon, Sironar etc. don't have their best resolution below f/4.

At least, Rodenstock has interesting stuff, like LINOS Photonics: High Aperture Laser Objectives (HALOs). So, something to study.

[Lance B]

The pentax 50 macro isn't in the top league. Thanks for the other pointers. I'll look them up.

I think he meant the FA50mm f2.8 not the DFA50mm f2.8. The FA50mm f2.8 was tested by Photodo.com and judged to be one of the sharpest they have ever measured.

[falconeye]

I think he meant the FA50mm f2.8 not the DFA50mm f2.8. The FA50mm f2.8 was tested by Photodo.com and judged to be one of the sharpest they have ever measured.

I wasn't aware of this lens. Thank You for the correction.

The photodo.com 40 lpmm center MTF figure for the F 50mm f/2.8 macro (called "Claimed to be one of the sharpest Pentax Macro lenses", no FA50 MTF figures) suggests that its center sweet spot is well above f/2.8 (f8 figure better than f2.8 figure). The weighted figures do increase with higher f, too. If the data hold true, the lens may be excellent, but not stellar

Only looking at weighted photodo.com figures, the following lenses seem to be noteworthy (f4 better than f8 at least -- as no lens there has f2.8 better than f4 -- and only the Canon has f2.8 better than f8):

Canon EF 200mm f/1.8 L
Contax-Zeiss G Planar 45mm f/2
Pentax SMC-A 85mm f/1.4
Contax-Zeiss Planar T* 85mm f/1.4

Those are the four best-resolving photodo.com lenses. And Pentax is part of the pack Unfortunately, not good-resolving enough

[loudbay]

I haven't posted on the board for quite some time. You, falconeye, have brought me out of hiding. Congrats. Or rather, thank you...

Eagerly waiting for more 'developments'.

Matt

[lesmore49]

What about Leitz lenses that are/were designed to be used with the famous and expensive Leica Rangefinder ??

I have used a number of Leica Rangefinder lenses and although I haven't any lens tests available...I would say that some of these lenses must be in the top group.

But then, I wonder how many photographers under 55-60 yrs. have had extensive use with Leica Rangefinders ? Not a lot I would think. If more did....I think we would see the Leica lens brought forward more often, for consideration.

As an aside....I have an old Leica rangefinder and I have to sat there is no other camera with a manual focus that clicks into optimum focus so quickly and easily as an Leica Rangefinder, at least that I have used.

[newarts]

Your work is an impressive example of what can be done with real equipment.

Are computational approaches to image formation appropriate for your work? In my view, de-blurring an image mathematically is a valid approach, especially when one is using data that is approaching perfection (like yours).

Even a perfect lens imposes a blur function to an image point and reversal of this blur can be accomplished mathematically. For an example see: Richardson-Lucy Deconvolution on Flickr - Photo Sharing!


Deconvolution can, in effect, negate diffraction overlap of image points. Easy to use software like FocusMagic and Raw Therapee are available for such work.

Dave

[Ben_Edict]

Good idea! Thanks.

A first result is this: I found MTF figures for Rodagon here:
Paul Butzi Photography

All Rodagon, Grandagon, Sironar etc. don't have their best resolution below f/4.

At least, Rodenstock has interesting stuff, like LINOS Photonics: High Aperture Laser Objectives (HALOs). So, something to study.

Process lenses, enlarging lenses and old(!) macro lenses are optimized for top-performance differently from normal lenses. They are optimized for peak performance at closer distances AND at smaller apertures, as for their intended applications, you would need the increased depth of field.

So, a macro lens etc. should NOT reach its peak performance at f/4 or even faster!

This changed slightly with current macro lenses (no new enlarging lens on the horizon, to my knowledge), as diffraction sets in earlier for smaller sensors, than for 35mm. So current macro lenses usually reach peak performance between f/5.6 and f/8 and already degrade at f/11, which in the past was near the best aperture.

Laser optics are usually corrected for short wavelengthes – and the shorter the wavelength, the higher the resolution gets. They usually have very small useable fields and are not corrected for visible light, which is a mixture of many wavelengthes (laser produce a single wavelength).

An interesting lens to test would be one of the old Ultra-Apochromatic Takumars, used with short wavelength light - but these are rare and excessively expensive.

Ben

[dantekgeek]

Sigma 30/1.4: 2453/2271

I knew the 30 was sharp at the center, stunningly so, but I had no idea it was (apparently) one of the very very sharpest!

This is an extract from this link NRFOTO Bjrn Rrslett

He is talking about the 28 mm f/1.8
Ultra-Micro-Nikkor

"Originating from a series of industrial lenses optimised for extreme sharpness (up to 1.200 lines/mm resolution!), this Ultra-Micro lens certainly delivers the goods. It is designed with an image circle of just 8 mm and optimised for 1:10 magnification, so the best way to employ it for macrophotography is reversing it on a medium- to large-format camera. At 10:1 (10X magnification) it then will give frame-filling shots for the 6x9 cm format with incredible and stunning image detail. I mostly use it on my 1955-vintage Linhof Technika field camera for extreme macro work. Optimum apertures are f/2-f/5.6."

Hope it helps

[Ron Boggs]

While you are researching, didn't Minolta make extreme macro lenses 25 years ago or more? A 28mm and/or 35mm as I recall, plus other rather wide macro super close-up systems...As I recall, these lenses weren't even of normal barrel diameter but were rather tiny. You'll have to dig a bit...

Not sure your premise of optimum performance at wide apertures can be substantiated. Yes, pro optics tend toward wider apertures, but that doesn't rule out optimum performance stopped down. This would seem to be particularly true of macro lenses as mentioned by Ben Edict.

The horrendous problems with diffraction with smallish digi sensors may make this issue moot, but make sure your premise is accurate before ruling out the killers at f.8 or f.11. As a guy who seldom shoots below f.8, I'd hate to see you rule out better resolution just based on the "wide aps" premise. Of course, I realize that a reversed lens will likely be used wide open, but can you stop down a reversed lens for these applications? Time to pull out a few of John Shaw's books just to play with the formulas he discusses...

Good luck and really nice images! I particularly like the extractions that are somewhat abstract and unrecognizable as human created items.

[Raybo]

falconeye,

Have you seen the movie "Fantastic Voyage"?

[Clicker]

One of the leica R like a Sumicron?