[rdenney]

EDIT: I have added links to the threads containing the tests for each lens, and a post at the end with some general conclusions.

But the teaser is that every lens I tested looked really good at f/11. The differences were wide open. Only one FA lens benefitted from a slight focus adjustment.

I tested the following:

30 Arsat Fisheye: Grand Lens Test 5: The Arsat 30mm Fisheye
35 FA: Grand Lens Test 3: 645 FA 35/3.5
45 Hartblei PCS: Grand Lens Test 7: PCS lenses Part A: The Hartblei PCS 45/3.5
55 Arsat PCS: Grand Lens Test 8: PCS lenses Part B: The Arsat PCS 55/4.5
55 DFA AW SDM etc.: Grand Lens Test 4: 645 D-FA 55/2.8 AW AL SDM
75 LS (had to tape the contacts on this one): Grand Lens Test 11: 645 A 75mm f/2.8 Leaf Shutter
120 A Macro: Grand Lens Test 9: Pentax 645 A 120mm f/4 Macro
200 FA (plus with 1.4 extender): Grand Lens Test 2: 645 FA 200/4
300/4 A* (as above): Grand Lens Test 10: The 645 A* 300mm f/4
400 FA: Grand Lens Test 1: 645 FA 400/5.6 ED(IF)

45-85 FA: Grand Lens Test 6: The 45-85 FA Zoom
80-160 A: Grand lens Test 12: 645 A 80-160 f/4.5
28-45 DA REJECTED (added November, 2018): REJECTED 645 DA 28-45mm ED AW SR—Lens damaged and returned.
28-45 DA APPROVED (added December 2018): Grand Lens Test 13A: APPROVED. The real DA645 28-45 ED SR AW is in the house

1.4x 645A Teleconverter: Grand Lens Test 4.5: 645 A 1.4x Teleconverter

And then I tested a 165/2.8 P67 lens, and this one was noticeably soft at wide apertures—the only one to really show that. I'll look at this one later.

I set up on my front porch. The day was cloudy with flat lighting but rich colors. I used my Gitzo 3532 tripod with an original Arca-Swiss Monoball. The 300 and 400 required mirror lockup, and the tripod ring I had put on the 300 wasn’t stiff enough, and I had to mount the camera body to the tripod.

Here's the scene, slightly recomposed and on a sunny day in the golden hour:


Every lens was able to make images sharp to the 1-2 pixel range, at least by f/8, with only a couple of exceptions that were slightly less good. Some could do it wide open. There just wasn’t a real dog in the bunch, if we consider the image as a whole, and not just a seven-foot print at 100 pixels/inch.

I conclude that the current notion that the 645z lacks a breadth of good lenses is relevant only to pixel peepers photographing test charts—at real print sizes, with real scenes, and with proper technique, the camera is sensor limited, not lens limited. The fleet of lenses old and new deserve the full respect earned by Asahi of old.

I’m as happy as I can be, but I still don’t know which lenses to take to Alaska

Rick “details to dribble out” Denney

---

Last edited by rdenney; 12-10-2018 at 12:30 AM.

[desertscape]

And then I tested a 165/2.8 P67 lens, and this one was noticeably soft at wide apertures—the only one to really show that.

I've always thought the 165 f/2.8 was a bit too long for a Double Gauss design. On their newer 165, the LS, they went with an Ernostar design, which is more of an accepted type for that focal length. A sharp lens although I don't shoot it at f/4 much. The newer 200mm for the 67 is also an Ernostar. Oddly, this design type is not used for any 645 lens.

[rdenney]

I've always thought the 165 f/2.8 was a bit too long for a Double Gauss design. On their newer 165, the LS, they went with an Ernostar design, which is more of an accepted type for that focal length. A sharp lens although I don't shoot it at f/4 much. The newer 200mm for the 67 is also an Ernostar. Oddly, this design type is not used for any 645 lens.

The 645 200 is in the Ernostar/Sonnar camp, it seems to me. Probably the 150, too. It’s a pretty broad category, considering all the variations.

Rick “noting the wide variation even within Zeiss variants” Denney

[texandrews]

But the teaser is that every lens I tested looked really good at f/11. The differences were wide open. Only one FA lens benefitted from a slight focus adjustment.

I tested the following:

30mm Arsat Fisheye on Hartblei adapter.
35 FA
45 Hartblei PCS
55 Arsat PCS
55 DFA AW SDM etc.
75 LS (had to tape the contacts on this one)
120 A Macro
200 FA (plus with 1.4 extender)
300/4 A* (as above)
400 FA

45-85 FA
80-160 A

And then I tested a 165/2.8 P67 lens, and this one was noticeably soft at wide apertures—the only one to really show that.

I set up on my front porch. The day was cloudy with flat lighting but rich colors. I used my Gitzo 3532 tripod with Arca Monoball. The 300 and 400 required mirror lockup, and the tripod ring I had put on the 300 wasn’t stiff enough, and I had to mount the camera body to the tripod.

Every lens was able to make images sharp to the 1-2 pixel range, at least by f/8. Some could do it wide open. There just wasn’t a dog in the bunch.

I conclude that the current notion that the 645z lacks a breadth of good lenses is relevant only to pixel peepers photographing test charts—at real print sizes, with real scenes, and with proper technique, the camera is sensor limited, not lens limited. The fleet of lenses old and new deserve the full respect earned by Asahi of old.

I’m as happy as I can be, but I still don’t know which lenses to take to Alaska

Rick “details to dribble out” Denney

Thank you Mr. Denney! That's a great teaser, too----I very rarely shoot anything but stopped down, although maybe this summer I'll go a little more gauzy. But I haven't seen much of anything that I thought was bad with any of my lenses, and the 55 is the only DFA I've got. In fact, I let go of my 45, but then got a cracking deal on an FA, NIB-old stock one out of Canada, so I re-bought it. Very handy FL on the Z.


I completely agree with you about the complaints of no or not enough new lenses.



My last quick trip was a drive, so I took a suitcase full of gear, and that's what I'll do again this time---gone for 6 weeks (although only shooting for several days, the rest is all other work...). Maybe you should just bear the pain for Alaska? BTW, I got one of these

Rock solid, fine quality. I can walk under it. $570.00

[desertscape]

The 645 200 is in the Ernostar/Sonnar camp, it seems to me. Probably the 150, too. It’s a pretty broad category, considering all the variations.

Yeah, the 645 150 and 200 are Ernostar-Sonnar but the 165LS and 200 Pentax for the 67 fit the Ernostar cross section. The original 4 element 200mm for the 67 used an Ernostar-Sonnar but was replaced by a 5 element Ernostar. But yes, similar designs.

[rdenney]

I want to add something that I will emphasize again when I wrap it all up.

I frequently read the detailed lens testing done by Roger Cicala at Lens Rentals, and that is NOT what I’m trying to do. He tests at a very high level (higher than DXO Mark) using an optical bench, and anyone with a quantitative capability ought to read his stuff.

He’s trying to determine when lenses need realignment, and how well the lenses are holding up to renter abuse. Most of his testing reveals analytical subtleties that would hardly influence my decision to use a lens.

I’m trying to determine whether lenses will meet my needs and fulfill my requirements, on a qualitative level, not how many line-pairs/picture height at MTF 50 they will resolve at maximum aperture.

My requirements include weather the lens, when carefully used, will sustain a sense (illusion) of endless detail on a 16x20 print, and, for each lens, the definition of “carefully used”. I want to invite close inspection. I also want to know whether a lens will or won’t be useful when I need to take the edge off a portrait, or provide a smooth and distraction-free blurred background. MTF testing does not reveal that.

I have done MTF testing before, without a bench and without Imatest, using film viewed under a microscope. Never again! I’ll leave that to the better equipped.

For example, I have loved my 45-85 zoom for years, but now I know that it needs to be at f/8 to resolve maximum detail, but wider open takes the edge off of portraits. This gives me a practical strategy for use as a wedding lens, and for use as a landscape lens. I would not have learned that from MTF testing.

Finally, I want to present my results with enough detail to allow others to compare against their own needs and requirements.

I love what Roger Cicala does—it really cleanses the Internet of subjective and anecdotal evaluation given with a cloak of quantitative fact. My tests are unabashedly anecdotal and qualitative, but they are not subjective.

Rick “putting things in context” Denney

[kaseki]

I want to add something that I will emphasize again when I wrap it all up.

I frequently read the detailed lens testing done by Roger Cicala at Lens Rentals, and that is NOT what I’m trying to do. He tests at a very high level (higher than DXO Mark) using an optical bench, and anyone with a quantitative capability ought to read his stuff.

He’s trying to determine when lenses need realignment, and how well the lenses are holding up to renter abuse. Most of his testing reveals analytical subtleties that would hardly influence my decision to use a lens.

I’m trying to determine whether lenses will meet my needs and fulfill my requirements, on a qualitative level, not how many line-pairs/picture height at MTF 50 they will resolve at maximum aperture.

My requirements include weather the lens, when carefully used, will sustain a sense (illusion) of endless detail on a 16x20 print, and, for each lens, the definition of “carefully used”. I want to invite close inspection. I also want to know whether a lens will or won’t be useful when I need to take the edge off a portrait, or provide a smooth and distraction-free blurred background. MTF testing does not reveal that.

I have done MTF testing before, without a bench and without Imatest, using film viewed under a microscope. Never again! I’ll leave that to the better equipped.

For example, I have loved my 45-85 zoom for years, but now I know that it needs to be at f/8 to resolve maximum detail, but wider open takes the edge off of portraits. This gives me a practical strategy for use as a wedding lens, and for use as a landscape lens. I would not have learned that from MTF testing.

Finally, I want to present my results with enough detail to allow others to compare against their own needs and requirements.

I love what Roger Cicala does—it really cleanses the Internet of subjective and anecdotal evaluation given with a cloak of quantitative fact. My tests are unabashedly anecdotal and qualitative, but they are not subjective.

Rick “putting things in context” Denney

Care should be taken to discriminate modulation transfer function (MTF) from contrast transfer function (CTF). MTF is based on sinusoidal density patterns (e.g., cycles/mm), whereas CTF is based on black/white density patterns (e.g., line pairs/mm) such as in the USAF 1951 resolution chart. The relationship between them (transform equation) is given in section 8.3 of the (now ancient) RCA Electro-Optics Handbook, and likely other optical texts. CTF is the easier to measure; MTF is the more useful in calculating resolution through a system as the MTF's multiply. Portra technical data provides MTF.

The RCA Handbook was later published and put on-line by Burle, which is now part of Photonis. I am unsure whether the free pdf still exists. Printed copies of this useful text are available from Amazon and others ranging from modest cost to stratospheric cost.

[rdenney]

Care should be taken to discriminate modulation transfer function (MTF) from contrast transfer function (CTF). MTF is based on sinusoidal density patterns (e.g., cycles/mm), whereas CTF is based on black/white density patterns (e.g., line pairs/mm) such as in the USAF 1951 resolution chart. The relationship between them (transform equation) is given in section 8.3 of the (now ancient) RCA Electro-Optics Handbook, and likely other optical texts. CTF is the easier to measure; MTF is the more useful in calculating resolution through a system as the MTF's multiply. Portra technical data provides MTF.

The RCA Handbook was later published and put on-line by Burle, which is now part of Photonis. I am unsure whether the free pdf still exists. Printed copies of this useful text are available from Amazon and others ranging from modest cost to stratospheric cost.

Good reminder. I'm quite familiar with that distinction. I performed a range of MTF testing using charts that provided sinusoidal patterns. It was extremely difficult to do using film without specialized equipment, and required a good bit of judgement (and associated error functions). To be sure, my warning is that I am NOT attempting anything like either MTF or CTF in my testing. I simply want to know which of my lenses provide performance acceptable to me, and under what general guidelines (such as which apertures to use), in making a photo of a typical scene.

Rick "thanks" Denney

[kaseki]

Good reminder. I'm quite familiar with that distinction. I performed a range of MTF testing using charts that provided sinusoidal patterns. It was extremely difficult to do using film without specialized equipment, and required a good bit of judgement (and associated error functions). To be sure, my warning is that I am NOT attempting anything like either MTF or CTF in my testing. I simply want to know which of my lenses provide performance acceptable to me, and under what general guidelines (such as which apertures to use), in making a photo of a typical scene.

Rick "thanks" Denney

And thank you for this performance tour de force.

[sinar.de]

The RCA Handbook was later published and put on-line by Burle, which is now part of Photonis. I am unsure whether the free pdf still exists. Printed copies of this useful text are available from Amazon and others ranging from modest cost to stratospheric cost.

Here is the link: http://www.ok1rr.com/tubes/burle/Electro-Optics-Handbook.pdf

[kaseki]

Here is the link: http://www.ok1rr.com/tubes/burle/Electro-Optics-Handbook.pdf

Good work locating that. I've saved a copy, even though I have the original (2nd Edition).

[johnha]

Hi Rick, thanks for doing all these tests, analysing the results and placing them in context. These are interesting threads, not just for the results but also a technical discussion on how best (or not) to measure lenses.

What I'm looking to get out of them is whether the D-FA & DA lenses are so much better than 'legacy' 645 lenses that they become the 'preferred' options (despite the price difference) for getting the best out of the D or Z - and whether my 6x7 lenses on the adapter are capable of 'good enough' results.

Given your assertion that in real world images, the Z is lens rather than sensor limited, could this be taken that the differences between the D and Z sensors (within the bounds of ISO capability) are perhaps less than the 'on paper' comparison?


John.

[rdenney]

Hi Rick, thanks for doing all these tests, analysing the results and placing them in context. These are interesting threads, not just for the results but also a technical discussion on how best (or not) to measure lenses.

What I'm looking to get out of them is whether the D-FA & DA lenses are so much better than 'legacy' 645 lenses that they become the 'preferred' options (despite the price difference) for getting the best out of the D or Z - and whether my 6x7 lenses on the adapter are capable of 'good enough' results.

Given your assertion that in real world images, the Z is lens rather than sensor limited, could this be taken that the differences between the D and Z sensors (within the bounds of ISO capability) are perhaps less than the 'on paper' comparison?


John.

The 645z is sensor-limited in a few narrow cases, but those cases occurred most broadly with my one D-FA lens. I only own one modern D-FA lens, the 55/2.8. The others I cannot afford at present (and "present" may be for quite a while--I have to absorb the expense of the camera and a versatile capability, which required lots of stuff like flashes and longer lenses, and then there was the Sinar P Expert Kit that happened to come my way, and the used Canon 5DII...).

Look at the test for the 55. What I see there that I don't see with the FA lenses is wide-open performance. At f/11, they are all pretty good, and some are excellent. But wide open, the 55 D-FA is more noticeably ahead. All the better Pentax lenses can approach the boundaries of the sensor in the center of the frame at f/11, but the 55 can do it over a wider range of apertures. Of course, it's a double-gauss normal lens--one of the easiest designs to correct to a very high degree. But I suspect the same holds true for the 28-45 and the 90. The 90 is better than the 120, according to a review by a commercial jewelry photographer (whose test I could not better for practical value), but not by much. I suspect the 28-45 is better than the 35 prime at any given aperture, but especially wide open.

I would suppose that I would prefer the D-FA lenses in all cases. But preference is one thing; choice is another. My choices have to answer to more constraints than do my preferences.

I only test one 67 lens--the 165/2.8. It's the only really fast long lens in the 67 line, and I don't have the 150/2.8 645 lens. The 165 was noticeably softer wide open than the 200/4 or the 120. Perhaps that's not a fair comparison, and I don't recall the images well enough to know how they compare at f/11. I'm thinking to redo that test with a different subject, and compare it to the Carl Zeiss Jena Sonnar 180/2.8, for more than just sharpness (neither are really sharp wide open, but they make superb portrait lenses, where surgical sharpness will reveal things the sitters often don't want to see).

Tonight, I'm adding the two perspective shift lenses I own to the series.

Rick "leaving the 120, the A* 300, the 80-160 and the 75LS still to come" Denney

[johnha]

I only test one 67 lens--the 165/2.8. It's the only really fast long lens in the 67 line, and I don't have the 150/2.8 645 lens. The 165 was noticeably softer wide open than the 200/4 or the 120. Perhaps that's not a fair comparison, and I don't recall the images well enough to know how they compare at f/11. I'm thinking to redo that test with a different subject, and compare it to the Carl Zeiss Jena Sonnar 180/2.8, for more than just sharpness (neither are really sharp wide open, but they make superb portrait lenses, where surgical sharpness will reveal things the sitters often don't want to see).

Thanks Rick, particularly the Interesting comparison between the 165/2.8 & 200/4 - a comparison with the Zeiss would be very interesting.

[Fenwoodian]

but I still don’t know which lenses to take to Alaska

I've been to Alaska a few times. My brother lived in Valdez for years.

When I first went there in the 1980's, I brought a giant telephoto that would get close ups of the sheep/goats/bear etc. But after I got home, I realized a photo with an animal that just about filled the frame was kind of boring - I could have just as easily taken it at the zoo.

These days I don't take the big glass any more. I prefer more of an "environmental portrait" of the wild creatures - photos that showcase their environment as much as the animal itself.