Pentax 645Z

[tsjanik]

Most of these "FA 645" lenses were originally design for 645 format film cameras, before the 645D. I only know of one new "FA 645" lens that covers the 70mm image circle of 645 format even though covering only the 55mm image circle of 44x33mm would have simplified the design: the 55/2.8 mm.  Some plausible explanations for that include:
a) covering the possibility of returning to 645 format, and
b) making that lens attractive to the many people still using Pentax 645 film cameras (who greatly outnumber owners of the 645D), and so increasing its sales.


P. S. The 25mm is an "DA 645", meaning I think that it only overs the smaller 44x33 format.  What other new "645" lenses has Pentax introduced since it discontinued the 645 film cameras?

The 90mm macro is D FA and the 25mm in Asia is also D FA; for some unknown reason, Pentax sells the reduced image circle DA 25mm in North America.

[leeonmaui]

Aloha,

That interesting and good to know, so there is a difference in the D FA and what is sold here, in the USA.
Do you think there would be any noticeable difference in the lenses on a 645D or 645Z
As I'm most likely to pair the 25mm with the new body.
(I miss ultra wide with the 645D 35mm combination.)

As a side note, I told my friend I was going to get it and it might be one of the last primary cameras i bought, as the Z denoted the camera was at the end of the alphabetical line...
He quipped back; well there's always the 645Z1
lol
 

[ErikKaffehr]

Hi,

The sensor size is the same, but the 645Z has more pixels so it will be a bit more demanding on lenses. The difference in linear sensor resolution is just 13%, that is not a lot.

So a lens that is very good on the 645D will also be very good on the 645Z. On the other hand a lens that does not work well on the 645D will be a small bit worse on the 645Z.

On technical cameras it would be a bit different, as the sensor can be tilted and shifted, but with the Pentax 645 the sensor is fixed and the wide angle lenses are retrofocus designs, to give room for the mirror box. So beam angle will be no problem.

Best regards
Erik

Aloha,

Do you think there would be any noticeable difference in the lenses on a 645D or 645Z

lol
 

[Bart_van_der_Wolf]

The sensor size is the same, but the 645Z has more pixels so it will be a bit more demanding on lenses. The difference in linear sensor resolution is just 13%, that is not a lot.

So a lens that is very good on the 645D will also be very good on the 645Z. On the other hand a lens that does not work well on the 645D will be a small bit worse on the 645Z.

Hi Erik,

Only worse per pixel. The higher sampling density (5.3 micron pitch instead of 6 micron pitch) will improve the combined resolution in cycles/mm of any lens. Output to the same size will improve. It's all about system MTF response, and sampling density makes a lot of difference.

Cheers,
Bart

[peterv]

The difference in linear sensor resolution is just 13%, that is not a lot.

Hi Erik,

I'm always struggling with this concept. Could you please explain how you came to this 13% figure? Thanks!

[Bart_van_der_Wolf]

I'm always struggling with this concept. Could you please explain how you came to this 13% figure? Thanks!

Hi Peter,

The resolution increase mentioned is derived from the higher sampling density (5.3 micron pitch versus 6 micron pitch, ). There are 13% more sensels that sample the same sized projected lens detail. That does not translate to a 13% system resolution increase (one would need to combine the MTF curves for both sensor scenarios to determine the actual system improvement), but a significant (anything more than 1% is visible, and thus significant) boost in micro-detail contrast and smaller features will be achieved.

Cheers,
Bart

[ErikKaffehr]

Hi Bart,

You are obviously right, I was a bit sloppy in formulating my response.

To elaborate a bit. A really good lens will shine on any sensor. A lens that is barely adequate will not be worse with a sensor having better resolution but the softness/aberration will be more obvious with the better resolving sensor.

I don't think a 12-13% difference is a lot, if we can print 20x26" with 39 MP we could 22.5x30" with 50MP, just as an example.

Best regards
Erik

Hi Erik,

Only worse per pixel. The higher sampling density (5.3 micron pitch instead of 6 micron pitch) will improve the combined resolution in cycles/mm of any lens. Output to the same size will improve. It's all about system MTF response, and sampling density makes a lot of difference.

Cheers,
Bart

[peterv]

Thanks Bart & Erik,

BTW, on the topic of MTF curves combined by sensor and lens I've found this lengthy tutorial by Panavision's John Galt and Canon's Larry Thorpe to be very interesting:

http://www.youtube.com/watch?v=gqq8QKMmtYg

It's a few years old and about resolution in motion/video mostly, but quite a lot of it applies to stills as well. Although I must admit I didn't understand all of it



 

[Theodoros]

Most of these "FA 645" lenses were originally design for 645 format film cameras, before the 645D. I only know of one new "FA 645" lens that covers the 70mm image circle of 645 format even though covering only the 55mm image circle of 44x33mm would have simplified the design: the 55/2.8 mm.  Some plausible explanations for that include:
a) covering the possibility of returning to 645 format, and
b) making that lens attractive to the many people still using Pentax 645 film cameras (who greatly outnumber owners of the 645D), and so increasing its sales.


P. S. The 25mm is an "DA 645", meaning I think that it only overs the smaller 44x33 format.  What other new "645" lenses has Pentax introduced since it discontinued the 645 film cameras?

It says here clearly http://www.pentaxforums.com/lensreviews/SMC-Pentax-D-FA-645-25mm-F4-Lens.html that the 25mm covers the whole image circle of 645 film image area with 109 deg of AOV...

[Bart_van_der_Wolf]

To elaborate a bit. A really good lens will shine on any sensor. A lens that is barely adequate will not be worse with a sensor having better resolution but the softness/aberration will be more obvious with the better resolving sensor.

Agree.

I don't think a 12-13% difference is a lot, if we can print 20x26" with 39 MP we could 22.5x30" with 50MP, just as an example.

While not earth shattering, another way of looking at it is represented in the attached MTF simulation, which compares both cameras with a very good lens, and a more mediocre one, at f/5.6. If we assume that MTF=10% (according to the ISO standards on resolution) is approximately comparable to the visual resolution limit we get the following:

Good lens, 0.7 sigma Gaussian blur radius, at f/5.6:
MTF10 is approx. 92 cy/mm versus 82 cy/mm, 12.2% higher resolution limit, with 7% higher modulation for the 645Z at the limit of the 645D. Medium size spatial frequencies get something like a 9.5% boost in modulation, when output at the same size. That's almost an order of magnitude, I'd call that significant.

Mediocre lens, 0.9 sigma Gaussian blur radius, at f/5.6:
MTF10 is approx. 72 cy/mm versus 63 cy/mm, 13.3% higher resolution limit, with 6% higher modulation for the 645Z at the limit of the 645D. Medium size spatial frequencies get something like an 8% boost in modulation, when output at the same size. That's slightly less than for a good lens, but still approaching an order of magnitude, I'd still call that significant.

The additional modulation also allows much better detail quality as a basis for post-processing, e.g. with Topaz Labs Detail.

Cheers,
Bart

[Gel]

The lens quality does concern me slightly because I love my 100mm Hassy 2.2

But as I concentrate a lot of my paid work on people and weddings I could well achieve a more unique look with a less than perfect lens.

[JV]

The 90mm macro is D FA and the 25mm in Asia is also D FA; for some unknown reason, Pentax sells the reduced image circle DA 25mm in North America.

Slightly disturbing.  Why would Pentax sell the reduced image circle DA in the US and not the D FA?  

The DA 25mm is being sold for $4,996 which already is not exactly cheap.  You are basically talking Hasselblad and Phase One prices here.

[tsjanik]

Slightly disturbing.  Why would Pentax sell the reduced image circle DA in the US and not the D FA?  

The DA 25mm is being sold for $4,996 which already is not exactly cheap.  You are basically talking Hasselblad and Phase One prices here.

The first offering in the US was D FA version; I have never seen an official explanation for the switch to DA, just internet speculation (better flare control, less manufacture cost, etc.).  The DA version has a longer built in hood yet is 30 g lighter, so  the change is more than just a longer hood.  I've read that many Japanese still use the 645 film cameras, so the D FA version continues to be offered there.
As to the shocking (for Pentax) price: as much as I like the 645D and would love to have this lens, I find the price too much of a deterrent.  I keep telling myself I could get a D800 and Zeiss 15mm for the about the same price, or an A7r and a Samyang 14mm for a lot less, or just continue to stitch with the excellent 35mm for no cost at all.

Tom

[Theodoros]

The lens quality does concern me slightly because I love my 100mm Hassy 2.2

But as I concentrate a lot of my paid work on people and weddings I could well achieve a more unique look with a less than perfect lens.

You may use Hasselblad V glass on it and even the Schneiders (60 Curtagn, 80 xenotar, 150&250 tele-xenar) for the Exakta 66/P6... Even some CZJ for the Exakta 66/P6 can be worth considering, but be careful with the later ones because there can be huge sample variations...

By the way... I wonder why nobody does a Rollei SL66 adapter for all 645 cameras that bear a focal plane shutter... there is surely enough space for a simple/cheap focusing mechanism to work properly... Most of the Rollei SL66 glass is simply superb! 

[ErikKaffehr]

Have you checked the flange distances involved?

Best regards
Erik

You may use Hasselblad V glass on it and even the Schneiders (60 Curtagn, 80 xenotar, 150&250 tele-xenar) for the Exakta 66/P6... Even some CZJ for the Exakta 66/P6 can be worth considering, but be careful with the later ones because there can be huge sample variations...

By the way... I wonder why nobody does a Rollei SL66 adapter for all 645 cameras that bear a focal plane shutter... there is surely enough space for a simple/cheap focusing mechanism to work properly... Most of the Rollei SL66 glass is simply superb! 

[BJL]

The 90mm macro is D FA and the 25mm in Asia is also D FA; for some unknown reason, Pentax sells the reduced image circle DA 25mm in North America.

It says here clearly http://www.pentaxforums.com/lensreviews/SMC-Pentax-D-FA-645-25mm-F4-Lens.html that the 25mm covers the whole image circle of 645 film image area with 109 deg of AOV...

Thanks tsjanik and Theodoros: so apparently there are two different 25mm lenses, the "D FA 645 25mm" offered in Asia (mostly Japan?) which covers the full 645 film format, and the "DA 645 25mm" offered in markets like the USA, covering only the smaller 44x33mm format.  (Maybe because the Japanese/Asian market has a far stronger contingent of Pentax 645 films camera users?)

Anyway, from what I know now, Pentax has introduced two new lenses in the "digital era" where it took an an extra design effort and costs in order to cover the 645 film format: the D FA 635 25mm f/4, and the D FA 55mm f/2.8.  I exclude the "digital era" D FA 645 90mm, because with lenses like that of FOV narrower than normal, optical designs naturally produce an image circle of diameter almost as large as the focal length, so any 90mm design will naturally cover the 70mm image circle of 645 film format. (The image is then cropped by baffles, the lens body, the lens mount and so on to the size needed.) It is only when a larger image circle involves a normal to wide angular coverage that it involves extra effort and expense.


P. S. Am I right that lenses designated "FA 645" without any "D" are designs from the film era? If so, I see just the four digital era Pentax 645 lenses mentioned above.

[BJL]

The sensor size is the same, but the 645Z has more pixels so it will be a bit more demanding on lenses.

As far as I can tell, the resolution of the new Sony 44x33mm sensor is limited to its Nyquist frequency of 94lp/mm, whereas there are monochrome films with high MTF to well beyond that. For example, Kodak T-MAX 100 has MTF of 50% at 125lp/mm according to Norm Koren at http://www.normankoren.com/Tutorials/MTF1A.html and IIRC, its MTF at 94lp/mm is very high, over 80%. 

So I expect that any lenses whose resolving power holds up well with fine-grained monochrome films will also resolve well enough for this sensor.

[bcroslin]

Regarding leaf shutter lenses with the 645Z. Does anyone know if there's a limitation with the 645D or Z to theoretically sync past 1/500th? I own the 75 f2.8 LS lens and once you get past having to manually cock the lens it's not a bad piece of glass and actually very sharp. The 645D fully supports the lens electronically. I wonder if the possibility exists for Pentax or some 3rd party to produce a line of LS lenses like the Phase/ Schneider lenses for the 645Z that would sync higher than 1/500th.

[tsjanik]

P. S. Am I right that lenses designated "FA 645" without any "D" are designs from the film era? If so, I see just the four digital era Pentax 645 lenses mentioned above.

Yes, and if you recall we made a five-year bet two years ago about MF sensor prices; I'm not sure who's winning. 

http://www.luminous-landscape.com/forum/index.php?topic=65392.msg520976#msg520976

[Lacunapratum]

Adapters to the Rolleiflex SL66 system with helicoid are available, but will set you back $800.  They work pretty well.  Alternatively, you can have a skilled mechanic rework those 645 helicoids.  And for the special lenses, you could also have a rigid adapter built.  I have all three versions.  

In terms of lens coverage, Pentax might have initially tried to maintain backward compatibility for 645 film users.  Remember - there was a huge Pentax 645 following in Japan.  Hence the original D FA design for the 55mm and the 25mm.  I seems that thereafter Pentax felt that especially the D FA 25mm did not perform to their expectation on the 645D and as the result limited the coverage to the smaller 44x33mm sensor format, choosing the DA designation for a revised DA 25mm, while keeping the D FA 25mm available in Japan for film users.  At that point they may have already decided that future models would rely on the smaller 44x33 medium format sensor, even though their third new lens (the D FA 90mm) once more appears to allow for full coverage.  

It seems that in their medium format program, Pentax largely depends on sensor manufacturers (e.g. Kodak, Sony) to guide their format choice.  Perhaps initially they were hoping to contract the production of large 54x40 sensors, but as it turned out, Sony's 44x33 CCD sensor gave them enough of a quantum leap to manufacture a new model.  Thus, they opted for the smaller sensor size, even though it doesn't take advantage of the full coverage of some of their wonderful lenses.  For Pentax, the bigger sensor would make sense, if it was offered, especially because of the nice adaptibility of matching P67 lenses.  But for right now, it seems, Ricoh/Pentax is betting on their new 44x33 645Z.  I don't believe that a larger sensor in the future is out of question, especially if the 645Z turns into a huge success and if there was a manufacturer to be had that would supply such a sensor at a reasonable cost to Pentax.  At present, the only limiting lens in terms of coverage is the DA 25mm, and I presume Pentax could re-tool for the manufacturing of the D FA version if necessary.  Some owners of the DA version might not be too excited about having to fork out another $4-5K for a D FA in addition to their DA, but they might have gotten sufficient usage out of the 645D and the 645Z to be willing to gain additional coverage on a larger sensor.