The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Question for the DMF landscape masters

Geoff

Well-known member
Is it fair to say there are two "plateaus" for MFDB - one is the top level of 80 mp backs, which gives the highest quality, but is less forgiving of defraction and color cast, requiring the newer lenses for shifting? The other sweet spot is around 35-40 mp, and allows for f16 and f22 without defraction loss, and also use of the next tier of lenses. For the more relaxed (?) shooter of LF or MF digital, the lower resolution back and its larger pixels has its advantages, if one can avoid the siren's call.

Here's a quick shot grabbed at f16, Apo-Sironar 55 mm, iso 50 all natural light. The crops show the workable DOF.
 
Last edited:

Shashin

Well-known member
If you stop down to f/22 on there will not be much of a difference between a p45+ and a iq180 in a print in terms of resolution, up to you if you want to waste the peak resolution in order to get a more equally sharp rendering or if you prefer a style more similar to the 8x10" way to shoot. When you've paid for all the pixels it can be nice to actually get to use them ;).
You are sure about that? Have you shot an IQ180? The difference in pixel pitch is 5.2um vs. 6um (that is 86% of the pixel pitch and 40% more resolution). I have shot 6um sensors at f/22 and I would say that statement is incorrect--you are not "turning" an IQ180 into P45+. And lets put pixel peeping of an IQ180 in context, on a 24" monitor, you are looking at about an 8" section of a 60" print.

Apart from that, resolution is never wasted. A higher resolving system will always resolve more. You will always "use" the pixels.
 

torger

Active member
You are sure about that? Have you shot an IQ180? The difference in pixel pitch is 5.2um vs. 6um (that is 86% of the pixel pitch and 40% more resolution). I have shot 6um sensors at f/22 and I would say that statement is incorrect--you are not "turning" an IQ180 into P45+. And lets put pixel peeping of an IQ180 in context, on a 24" monitor, you are looking at about an 8" section of a 60" print.

Apart from that, resolution is never wasted. A higher resolving system will always resolve more. You will always "use" the pixels.
It's correct that resolution is never fully wasted, it does get sligthly better, especially if we value the reduced aliasing issues (less color aliasing/moire issues if we have some diffraction onset, less jaggies issues when printing ultra-large etc). But the returns are diminishing and I'd say that it may not be worth playing the upgrade game for the resolution alone if you often shoot at f/16-f/22, and if you are new to MF, getting in at 33-39 megapixels is a lot cheaper than getting in at 60 or 80, especially if pre-owned options are considered.

I have not personally shot an IQ180 but I've made prints with other systems, sharpened at different diffraction levels and compared, seen others make tests, know the theory and have seen files. But sure, to be 100% certain I'd prefer to make a true A/B test of my own :).
 

alajuela

Active member
Is it fair to say there are two "plateaus" for MFDB - one is the top level of 80 mp backs, which gives the highest quality, but is less forgiving of defraction and color cast, requiring the newer lenses for shifting? The other sweet spot is around 35-40 mp, and allows for f16 and f22 without defraction loss, and also use of the next tier of lenses. For the more relaxed (?) shooter of LF or MF digital, the lower resolution back and its larger pixels has its advantages, if one can avoid the siren's call.

Although might not be directly to the point being made, there are several advantages overall to the 180, apart from being able to have really large strong files and great color transitions.
1. On the sensor plus, you are now shooting 20 meg files instead of 15 on the 160. this is a big deal.

2. The ability to crop if necessary, this is the same principle as large negatives.

The things being listed as negatives, such as requiring flawless technique and limited selection of lenses which are expensive, are things which I feel are enviable as the technology progresses.

There are horses for courses, but the 80 DMF is more than just big prints.

Thanks

Phil
 

torger

Active member
the 80 DMF is more than just big prints.
Yep, if we compare with the P45+ specifically the color is much more accurate too in the 80.

Ideally we'd want as much resolution as possible, and an antialias-filter on top (aliasing is really no good!) so we would "outresolve" diffraction and lens with a wide margin.

Problem is that we get to pay with more color cast, rendering some fine lenses near useless (like the Schneider 35mm), and lots of more dollars. Sharpening workflows in today's software are also adapted to a certain level of outresolving, a file that is too fuzzy on the pixel level is hard to work with.

That makes me prefer balanced systems, a system that supports the personal way to work without costly excessive overkill in one or more parts. A user that intends to shoot at f/16 and f/22 often and want a balanced system and appreciate good economy I think should look into an older generation back.

I would guess that there are not many users that have followed the upgrade path from 22 to 80 megapixels and still use the same shooting apertures as they did on their 22 megapixel backs. When you upgrade to higher resolution, a desire to shoot at larger aperture and get higher resolution lenses appears. So maybe one started with a "all pixels equally sharp" style and shifts towards a more 8x10"-like style, i e one sharpest plane of focus at some good place and let other areas go quite much out of focus on the peeping level, and maybe even get a strong desire to focus stack scenes that you never would do with the 22 meg back. Not necessarily something bad, but may be worthwhile to think about when investing new or upgrading.
 

Shashin

Well-known member
Unfortunately, pixel peeping is the fashion and we have this distorted myth about diffraction which is resulting in turning photography into an object lesson in maximizing MTF. Resolution is the biggest myth as it has little to do with sharpness and a viewer is more attracted to sharpness than resolving power. Ultimately, it is whether the image is good, if if it takes an aperture smaller or larger than the MTF ideal, then that is what you need.

There is more to an image than how many pixels it is divided into. To say I should not have bought a 645D instead of a P25+ because I use small aperture is silly. Not only do I use other apertures, but I get a much better sensor with better DR and noise. But even at small aperture, I am getting the benefit of a higher-resolution sensor. And the same goes with high-resolution lens on "low" resolution backs, you are still benefiting.

I see a bleak future where photography is just a bunch of landscapes taken at the golden hour at f/11. All a result of pixel peeping and the distorted view of photography it presents. Fortunately, the lens manufacturers know how to make a lens an have given them more than one aperture.
 

Geoff

Well-known member
Its interesting how what I would have considered a gentle discussion has become hard edged. That was certainly not my intent - rather to suggest that there are different places in the hardware landscape and that different folks might gravitate to their own spot. Maybe it's my own rationale for not upgrading!

The extra resolution would be lovely - to be able to shoot one lens and crop more fervently, etc. But the present solution is pretty nice too - and as Torger points out, has a nice balance to it. I feel its like being a 4x5 shooter vs. an 8x10 shooter. Nothing really wrong with either one.
 

Don Libby

Well-known member
You can pixel peek all you want however in the end it's boils down to how does the image look at "normal viewing distance".

I'll admit that the first think I do to any file is view it at 100%, looking for dust/dirt and any anomalies that might interfere. The higher the resolution the more these anomalies show up - specks in the sky that when viewed normally look like dirt turn out to be birds when looked at 100% or higher.

To stack or not to stack. For me there's no hard fast rule. I don't view my subject before hand with a predetermined thought. What I'll normally do is let the subject decide. If I'm standing in an area where it appears that I want everything in focus from my toenail to horizon I'll check the f/stop and which lens I happen to be using (still too new with the HR40 t/s as I've only had it a couple months). If what I see is how I want to capture (hope that makes sense) then I'll do a couple shots and leave. If on the other hand I do want that sharp focus then I'll stay long and stack.

Writing and reading the above remained me of one thing. Remember shooting film? You (or at least me) never took just one shot. I'd shoot the scene at different f/stops for insurance; I'd do the same with digital.

I guess my answer (if it counts) is that I have no hard fast rule other than to let the subject matter/lens/conditions dictate stacking or not.

As always just my 2¢

Don
 

Shashin

Well-known member
I feel its like being a 4x5 shooter vs. an 8x10 shooter. Nothing really wrong with either one.
Look at the range of apertures given on a 4x5 lens vs. an 8x10 lens. The 8x10 stops down further. You see, diffraction limit is format specific, not resolving power specific. The advantage of the 8x10 was not simply resolving power and actually resolving power may not have been the defining factor. And this is part of the problem with a resolving power/MTF only approach--it does not work in context of pictorial images. That is kind of like saying the only thing that matters in a sports car is top speed; turning circle, handling, and acceleration are irrelevant.

Unfortunately, this MTF approach has taken over photography. People are constantly repeating this myth because they keep reading it. Where the problem is that photographers are versed in half the technical process, very few know about how an image is perceived. And the perception of an image counts. Connected to this diffraction thing is the idea that DoF decreases with resolution--how can image have less DoF simply because it is divided into smaller pixels? Pixel peeping at work again.

I guess what I object to is taking an open system of related variables and forcing it into a closed system of rules. Especially, when the rules are not really true. I am not impressed with painting by numbers either. Photography is a creative act and the science behind it let you make choices and evaluate results, it does not dictate how you photograph. The science is no problem for me--I have a science degree in imaging and do scientific imaging professionally--but I also understand that photography is an art. "Thou shalt" does not work well in art.
 

torger

Active member
If one chooses to get 80 megapixels because of the slightly better color or slightly improved DR (which according to me was really good already in the 22 meg backs), that's fine with me :).

Some get it to resolve significantly more details though, and then you cannot kill it with diffraction, which only in parts is recoverable with deconvolution. If I'm going to give advice I won't say that paying for 80 megapixels when having a shooting style that mandates f/16-f/22 in a large amount of pictures is a great idea without explaining what that will mean, and how you can relate to it.

I see no problem in if you choose shooting at the same apertures regardless of resolution of the back, but frankly I don't think that's how most people value their $40K pixels. They want to maximize the resolution if possible.

I use a lower resolution back and do not think more pixels will improve my prints much when my shooting style is as it is, as explained before. More pixels if you can have them for free is always better though of course.

(Looking at typical apertures of 8x10" and 4x5" landscape shots it seems to me that they used relative "large" apertures. 8x10" lenses are often limited to f/90, which is f/20 on an IQ180, but are rarely shot at those apertures. The 8x10" tradition uses the largest apertures in relation to format size to my knowledge, hence my examples)
 

jlm

Workshop Member
i use f8 or f11 because the sk's and rodies are at their best and that is the first link in the chain
 

Shashin

Well-known member
OK, lets talk real. Here is an image taken at f/22 on a 40MP sensor with a 6um pixel pitch.



Now, I make my tests challenging because I want to know what my systems do. I made this test under very flat lighting, because as the target contrast goes down, so does system resolution. (I also made a 3 foot print from this as well because I like to make real tests, not to flatter, but to know what I can achieve.)

Here is a 100% crop of a flat area in the image. I just quickly processed these in ACR. I certainly could do more. And I do not know if I am cropping at the plane of focus either. If this was a 44" x 33" print, this crop would be a 4.8" square section of it.



The problem with the optimized MTF argument is what if you lose resolution because you do not have enough DoF? Certainly, even at f/22, I am still getting my pixel's worth out of the 645D.
 

torger

Active member
OK, lets talk real.
Thanks for the interesting example. It mirrors well my own results. You make a good point showing that f/22 on a 6um back is not too bad. When you shoot low contrast structures that are imaged large it hurts less too, compared to say a brightly lit detailed city at a distance.

Side by side with a shot made at f/11 at the plane of focus would show a significant difference though, and if making a print it would not look that much worse if shot with 9 um pixels, hard to differ would be my guess.

There's more with you though, this techy test shows that with proper sharpening techniques you can sharpen f/22 images to very closely resemble f/8 or even f/5.6 images, and that with a D800 which has even smaller pixels:
The Diffraction Limit. How Small is Too Small? - On Landscape
(I think they underestimate the low frequency effects though which affects the global sharpness impression). Digital sharpening is much more powerful than tools available when shooting film, so we can push diffraction further now.

Anyway, what people with high res backs often choose to do is to rather than shoot f/22 like they perhaps did with their old 22 meg back to get deep DoF, they shoot at say f/11 or even f/8 with the focus plane at something important and let it slip visibly out of focus (at nosing distance at least). While this style seems more relaxed in terms of DoF it requires precise focus placement to get the desired result (laser distance meter and high precision focusing ring of the pancake cameras). This way you get to appreciate the peak resolution of the system where the focus is set, at the cost of a less evenly rendered scene in terms of resolution.

I prefer to stop down more to get more equally rendered resolution, unless I want a short DoF effect I don't shoot larger than f/11. I'd guess something like 70% f/11, 20% f/16 and 10% f/22 - f/32, tilt used very often. It works well with a view camera focusing on ground glass (I use a Linhof Techno). I have 7.2um pixels today, I'll use the same style with 6um pixels.

It would be interesting to know what is taught in the MFD landscape shooting classes these days.

Although it might seem like we disagree in most things here, I think we are quite close in how to deal with DoF in practice, ie not be too afraid about stopping down. What I talk about here is that if one spends a very large amount of money on a very high resolution system one may want to choose a shooting style that really maximises that potential, and I see that many do. I also think that if one prefers to stop down more (like I do) it may be worthwhile to look into older generation backs as a more economical and balanced solution (I also like the distortion-free light f/11-optimized Schneider Digitar wides); while it's true that higher resolution system renders better results even when stopped down, the difference becomes rather small.
 
Last edited:
Interesting discussion.

On pixel peeping, as a landscape photographer who shoots f5.6/8 and routinely stacks for maximum clarity throughout, the delight is not in sharpness for its own sake, but because of the amazing beauty that is inherent and revealed in the detail of nature - sand patterns of a late-light dune, the texture and tones of snow on foliage, the form of rock...the wonderful macro world within the landscape is opened to us as never before. This new technology inspires ever greater appreciation of the levels of life and matter that make up a landscape.

It's also more than I can see by looking with 55 y.o. eyes!
 
Very interesting discussion. Yes, I do use focus stacking quite often, but rarely for anything shorter than 75mm. I will use it on a 45mm lens, but only when I absolutely have to have infinity at super sharp focus. Even then I will only take 2 images to stack, and not something like 3 or 4 like I might on my 75-150mm lens.
 

Guy Mancuso

Administrator, Instructor
I think we have throw in here the quality of glass as well. From what I have seen is diffraction on my 55 LS lens at F16 but with the 35mm phase lens not of the same quality I have even dropped to F22 with a 6 micron back and the 35 did not hit diffraction until than but I recall on one shot I guess depending on subject it popped up at F11 on the 55 LS .

Now with my tech cam lenses diffraction was even more of a issue getting into that zone of F16 and F22 so I used tilt at F11 or I would focus stack if needed. I think its a combination of micron size backs and glass. I remember my P25 plus 9 micron diffraction was less of a issue , maybe I had worse glass back than. Now admittedly if I need the DOF and can't focus stack ill shoot at f22 and live with it. Guess my point here outside of the tech stuff it seems to be a combination. Just a thought
 

alajuela

Active member
Very interesting discussion. Yes, I do use focus stacking quite often, but rarely for anything shorter than 75mm. I will use it on a 45mm lens, but only when I absolutely have to have infinity at super sharp focus. Even then I will only take 2 images to stack, and not something like 3 or 4 like I might on my 75-150mm lens.
Hi

May I ask what software you use to focus stack? I know there is Helicon and Photoacute, but I have no experience with either.

Thanks

Philip
 
Top