1. The Question
“So what does it look like when a nature nut obsessed with large prints gets his hands on the 645 system?”
I think we all wondered about this question. Because, let’s not fool ourselves; the 645 system has always been for landscape photography. It is not there to be cuddled. It is meant to be covered with flurries in the high Sierra, beaten by hail in the stormy Grand Canyon, and blasted with sand showers in the scorching Death Valley.
It took five years to build my photography around these expensive chunks of metal and glass, and I think now I can say a thing or two about them.
2. (More Formal) Introduction
I always thought the 645 system is primarily designed for landscape photography. Having been an old-timer with the 645N II in a worn-out bag and a heavy user of the 645D and the 645Z since they were first introduced into the market, that perspective never left my mind nor was ever challenged. Since hauling a 40-pound pack on a regular basis for backpacking trips in the Sierra cannot be good for my beat up joints, I might have some suggestions for future improvements, but all of these cameras and lenses stood the tests of the most unforgiving terrains and elements so far. So I thought there must be a review from who knows a little bit about landscape photography.
It’s not only the camera but also the photographer that has seen changes over the years. My fixation with the ‘a photograph must be a single exposure’ ideology, stuck in the head for quite sometime especially during the film days, has seen noticeable compromises, as I’ve learned more about the differences between the way the human eye sees the world and the way cameras do. And when I was educated more about the physical limitations that any high-resolution photographic system is subjected to, I decided to take that painful path that embraces new digital techniques. So this is also a piece about exploring new ways to best utilize the photographic system’s potential.
Here I wanted to share some of the images that I photographed recently, as a way to examine the capabilities of the Pentax 645 system; especially the new 645Z and some of the key optical pieces that have found space in my bag. As the prime technical objective of my work is to produce large prints of the highest quality, I thought providing some insights into the pixel-level quality, also not in its raw form, but in the final phase of the images after all processing is complete and just before being fed into a Giclee printer, would be useful.
3. What’s in the Bag
- D-FA 25mm
- DA 28-45mm
- FA 45-85mm
- FA 80-160mm (occasionally)
- Memory cards (Sandisk Extreme Pro 95mb/s)
- Remotes (like chinese ones better than the Pentax ones)
- Lee 4X6 GND filters and home-made filter holder (only for seascapes)
- Polarizers and ND filters
- A modified inspection mirror (attached near the focus window of a lens)
4. Technical Details
Almost all images are produced from multiple shots with different exposure and focus settings. For those who are not that comfortable with that notion, here is a simple fact; the human eye doesn’t work like a camera at all. It constantly captures pieces of the real world and our brains stitch them up and turn into what we call a memory, an artificial reconstruction and also often incomplete rendition. Each little piece is of different brightness and placed at a different distance, but our eyes do a tremendous job covering almost over 20 stops of contrast ratio and infinite depth in a heartbeat. So in order to see the way our eyes do, something has to be done.
The extent of this multi-exposure / multi-focus process depends heavily on the scene. To give you a point of contact, ‘Echoes of the Silence’ utilizes almost 18 shots, as it represents the worse case scenario for landscape photography; a large depth of field and a high-contrast backlit lighting condition. On the other hand, ‘A Night with the Lights’ only(?) needed three mostly to extend the depth of field, as the light was tamable and the nearest objects were still good tens of feet away. ‘The Highland’ is a single exposure image, that’s what I like about photographing with a telephoto lens sometimes; the odds are you don’t have to worry too much about depth of field (well not always, sometimes it gets even worse).
The Sony 51.4MP sensor, just like any high-resolution medium, is prone to optical diffraction, and it becomes noticeable at any aperture smaller than f/10. If you stop down for increased depth of field, say to f/16, your effective resolution is not that different than using a 24MP camera (the same thing can be said about large format film cameras too. If you stop down too much, that negates the very purpose of going big). On the other hand, at a wider aperture, diffraction may be minimized, but the corners deteriorate quite rapidly due to optical aberrations. So for most images in this review, the aperture was kept at f/10.
There certainly are some exceptions. When the image includes elements that cover some depth and change rapidly (waves in seascapes), I resort to smaller apertures. When a sun-star plays a role in the pre-visualized image, I occasionally stop down to include the desired effect. When I am dealing with stars and the moon, I open up to let more light in.
I know the newest trend in landscape photography is oriented heavily towards overuse of shadows and Orton effects to make the images appear more dramatic, but that style has not really spoken to me personally nor is well suited for print-oriented photography like mine (printers do not handle shadows very well).
Total eight images are presented as examples, and nine different 800px X 800px 100% crops are provided for each image. All of these images are photographed with the 645Z. Please note that these images are optimized for canvas prints and therefore sharpened quite aggressively. There are some artifacts because of the level of sharpening and also the Bayer construction of the sensor, but based on trial and error, we figured it works better that way for prints.
(1) Echoes of the Silence (2015)
D-FA 25mm / Eastern Sierra, California
(2) Winter Thirst (2015)
D-FA 25mm / Lake Tahoe, Nevada
(3) A Daydream (2015)
A 35mm / Grand Canyon, Arizona
(4) A Window to the Desert Winter (2014)
DA 28-45mm / Canyonlands, Utah
(5) And a River Runs Through (2014)
DA 28-45mm / Yellowstone, Wyoming
(6) The Highland (2014)
FA 80-160mm / Mt. Rainer, Washington
(7) The Grand Stage (2015)
FA 45-85mm / Grand Canyon, Arizona
(8) A Night with the Lights (2015)*
D-FA 25mm / Banff, Alberta
All images are compiled in the following link, to see the 100% crops, just click on any image.
To me, the final verdict is quite simple and as predicted. The 645Z is a great camera with amazing DR and low light capability. The DA 28-45mm is fantastic and outperforms the rest. The D-FA 25mm follows very closely but it’s got some issues with field curvature and chromatic aberrations (CAs). The FA 45-85mm, FA 80-160mm, and A 35mm are all great performers, but their days are numbered for any future high MP sensors.
These are things that I learned over the years.
A. High-resolution Photography
(a) The current 51.4MP sensor, when used properly, can make prints with amazing details easily up to 45 X 60 size (about 137.43ppi).
(b) Fully utilizing a 80-100MP sensor will be a real challenge, but it doesn’t hurt to have more (at the expense of more computing power and storage, but personally I’d rather take the resolution).
B. The 645Z compared with 35mm full frame
(a) The 645 system still holds an edge over any 35mm full frame system in resolution.
(b) With the 51.4MP Sony sensor, the 645Z offers better low light performance than most of the high-resolution 35mm full frame cameras in the market.
(c) The 4:3 ratio works much better for vertically composed images.
(d) The 645 system needs to offer a better ultra-wide angle solution. Even the 25mm feels limited sometimes.
(e) Much bigger and heavier than any 35mm full frame systems.
C. The 645Z compared with the 645D
(a) Much faster operation overall. With the 645D, you need to be patient, but the 645Z feels just right and smooth.
Buffer and writing speed are also greatly improved, enabling responsive interface for the photographer (reviewing images, checking exposure, histograms, etc.) even in a continuous, high-speed shooting situation (for example, bracketing five exposures with constantly changing focus under fast-changing light condition)
(b) Vastly improved dynamic range and low light / noise performances.
(c) Shutter / mirror vibrations are reduced significantly, crucial in high-resolution photography.
(d) Articulated screen is a very useful addition for low or high angle shooting.
(e) Live-view is incredibly useful. Especially for accurate focus and exposure.
(a) D-FA (DA) 25mm
- It is a valuable addition, even when you have the DA 28-45mm. The extra wideness is very useful. Often I find myself using the 25mm more than the 28-45mm.
- The lens exhibits noticeable field curvature and CAs. Multi-focus shooting can mitigate the field curvature, and the CAs can be fixed relatively easily in post-processing.
- The lens offers great resolution, but I doubt the corner performance will hold up very well when the sensor resolution is increased beyond 60-80MP.
- It performs very well with the 645Z, offering fantastic corner-to-corner sharpness through out the entire zoom range.
- Copy-to-copy variation is quite significant. Please test your lens before purchase. Tested about 3 samples, all perform differently. The most inconsistency comes from the corners at the wide end (28mm).
- Awfully heavy and big, feels almost over-engineered. SR is almost completely useless for landscape photography. I can see how Pentax wanted to appeal to the studio crowds with this, but my knees and back are paying the price.
(c) A 35mm
- Early observations (when tested with 645D) indicated that this performs better than the FA 35mm, which showed more field curvature.
- It still performs very decent with the 645Z, but compared with the DA 28-35mm, high frequency details are subpar.
- Copy-to-copy variation seemed quite minimal, based on 3 different samples I personally tested.
- I don’t see this will make a good lens for the future sensors with higher resolution.
- Still a valuable piece for backpacking trips due to its weight and size.
(d) FA 45-85mm
- Still a fantastic performer. The performance at 45mm is almost comparable to the DA 28-45mm.
- Copy-to-copy variation seemed quite minimal, based on 4 different samples I personally tested.
- Will need to see some updates for 80+MP sensors.
- The best performance-weight ratio of the bunch.
(e) FA 80-160mm
- A great lens at a reasonable price. Overall performance with the 645Z is decent.
- At this focal range, very prone to vibrations. Needs sturdy support.
- Needs to be updated for future 80+MP sensors.
7. Suggestions for Future Development
Here are some things Pentax can do to make it better.
A. 80-100MP sensor
Practically, this will be ultimate resolution for the 645 system. All kinds of logistic challenges to go beyond this; lenses need to be sharp corner to corner at f/4-5.6 to avoid diffraction and near-to-far focus will require tens of different layers and extensive post-processing, which means it’s almost practically impossible to utilize this level of resolution. One way to address this issue is to go with non-bayer, Foveon-type sensor. This will increase the resolution by a factor of about 1.5 at the same level of diffraction.
B. Ultra-wide angle
Be a prime or a zoom, please make one. Personally, anything close to 18mm (14mm on a 35mm full frame) can be really useful.
I know it will look funky with the current flange-back distance, but I am up for whatever can shave off a pound.
D. Downsized lenses
Please keep the lenses light. Offer non-SR and not-so-fast variants designed specifically for outdoor photography. The current system is almost unacceptable for any serious hiker. You will need to add about 20 lbs to the base weight to carry 645Z + 25mm + 28-45mm + 45-85mm + Gitzo 1 series carbon tripod.
E. Live View
Zooming is fine, but moving to a particular part of the frame is not so smooth and too slow. This considerably slows down the focusing process in live-view shooting.
F. Automated multi-focus shooting
Multi-focus shooting and focus stacking in the post-process are becoming an essential technique in many applications. Implement an algorithm to do this. Please allow an option to be used in conjunction with exposure bracketing; also let the photographer customize what focus range should be bracketed and what not.
G. Focus windows
Additional windows on the sides (or bottom) of lens barrels will be a great addition, so that the photographer has complete information and control of focus at all times (you can’t check the focus when your camera is higher than your eye-level). More precise and finer ticks will be very helpful as well. If there is a way to do this electronically (show a number that precisely displays the current focus distance on the live view feed), it may be better.
H. More custom modes
Three custom modes (U1-U3) are amazingly useful, but one can use one or two more.
Personally, I am a consumer of practically zero brand loyalty. I take whatever is best for the job. My take is, a preconceived notion of any company does not really improve my work in any useful way.
Given that, I want to thank Pentax and it’s engineers that have made these great products possible. The 645 system held up my photography for the past five years, and it saved a number of images that might otherwise have turned out quite differently. I am an engineer by training and I know how difficult it is to develop these low-volume, high quality products and keep them afloat. And they have done an excellent job over the years.
Though I still do not call myself a fan (because that kind of rhetoric or mindset often hinders objective assessment), I sincerely want to thank Pentax for the unconventional devotion to image quality, because it resonates deeply with the principles I hold as a photographer. And I think I really need to say these things. Because I see something more than simple business plans in these beautifully strange products.