Future MF Digital Decisions ~ 2018

[Frankly]

I'm an armchair follower... I'm a long established photographer with an extensive Nikon system, 35 years of large format film experience, blah blah blah. With the improving market I'm actually at the point where a medium format system is within grasp so I've been speculating and researching. This forum has been very helpful, I've read most of the manufacturer's literature, Anders Torger and Ming Thein, logged my YouTube hours.

I wouldn't want to replace my Nikon system, I value the long telephoto and professional auto-focusing too much. To be honest the newer lenses like the 105/1.4e and the Zeiss manual focus options are highly satisfactory and good enough. And besides I only print large for sales a few times a year, I no longer exhibit or make a pretense about galleries, all of my work is on my website and if it looks good on a 4k monitor that is all that really matters. I look forward to the day when I can hang digital monitors on the wall instead of prints.

Still I have the urge to increase bit depth and resolution for my landscapes in the range of "normal" focal lengths, 35mm to 85mm equivalents. I know it's overkill and simply shooting the Nikon is more than sufficient but I do miss the large format pace, the deliberate fussing and contemplation. And to be honest, it is mostly psychological, knowing you're doing something the nth degree, following the best practice and chasing the highest quality possible.

I'm not interested in using an older MFDB system, I'm thinking that a 100mp CMOS sensor is the way forward in order to achieve a significant "jump" in quality over a modern Nikon/Sony FX sensor. I understand that there is also a quality and look to the older fat pixel CCD sensors but I'm not at the point where I want to indulge in "retro-digital" quite yet. Maybe in 2030 it will be the cool thing?

My understanding is that these newer CMOS sensors do not play well with movements especially with wider angle lenses. And both Hasselblad and Phase One will be developing mirrorless systems (like the X1D) with the newest sensors and lenses specifically designed to take advantage of the lack of depth required for an SLR mirror box (less optical compromises).

So does this mean that waiting another season or two for them to create a 100mp mirrorless with a 35mm equivalent prime non-retrofocus design lens is going to be pretty much the ultimate?

Will there even be a purpose to technical cameras after this? With resolution overkill and automated focus stacking why would I want the complexity and chance of misalignment from a tech cam?

To me and probably a few thousand others if it means we can plunk down $30k and get the best of the best in Phase One's version of the X1D concept and be "done" isn't this the Holy Grail?

The only thing sticking here is the cost, and that old professional habit of always having a duplicate back up system at such a cost!

But please, speculate with me and tell me where my thinking is off?

 

[thompsonkirk]

Your thinking comes across primarily as a passion for ultimate image quality, though this seems to contradict your current plans for presenting your work. Among MF systems the margins of difference in IQ that will be visible on monitors and in occasional prints are probably too slight to merit putting things off for further research or technical developments. And stitching is so easy that quite large and detailed files are already at hand.

My background is similar (except Leica and now Sony instead of Nikon), and I've accomplished with considerable satisfaction what you seem to have in mind using the GFX. I don't see the point in waiting for 100MP because I'm not sure that more pixels, if they're smaller, will afford an interesting IQ gain. So my suggestion is get started now, not at $30K expense, but at a third of that. Just dive in at a modest price and see if you need to go anywhere beyond that?

Since you mentioned years of large format film experience, one caution: MF digital does not differ from DSLR by reproducing LF film IQ; it has its own aesthetic. The curves are different, without the extended toe and shoulder that your film experience accustomed you to. Dynamic range is really the most important technical development for you to look for (not more megapixels). With MF you can retain considerably more shadow detail. The challenge is to preserve minor gradations of highlight detail in prints. But this is an issue mainly if you're carefully crafting your own fine prints. On monitors – by transmitted light – it's not hard to make everything glow.

Another caution, if you have large-format IQ expectations, is to be cautious about the latest in MF lenses, almost all of which err in the direction that folks call 'clinical,' with emphasis on resolution and micro contrast rather than on '3D' rendering. I quickly gave up on modern primes, stepped back to older 645 Contax lenses, and then found (to somewhat embarrassing surprise) that the basic GFX zoom lens is 'gentler' than the new primes and not too different from the older ones – and a lot easier to pack around than an extended lens set.

Summary: My two cents is stop thinkin' and dive in experimentally. If you're like most of us, the 'holy grail' of what's best or right is bound to shift and evolve.

Kirk

 

[vjbelle]

I don't see the point in waiting for 100MP because I'm not sure that more pixels, if they're smaller, will afford an interesting IQ gain. So my suggestion is get started now, not at $30K expense, but at a third of that. Just dive in at a modest price and see if you need to go anywhere beyond that?

So, so true. Not many realize that the next 100MP smaller sensors and 150MP larger sensors will be highly diffraction limited regardless of the 'so called' diffraction compensation software. I have lots of reservations regarding the next generation of sensors and hope that enhancements will be forthcoming for the current size sensors.

Victor

 

[jng]

What Kirk and Victor said... More megapixels can be tempting but at <4-5 micron pixel pitch the point of diminishing returns may have come and gone. Right now I think the main differentiating factors are sensor size, optics and of course how the files are handled in the raw conversion process. My "old" 60 Mp CCD back gives me files I much prefer over the ones from my 36 Mp Nikon, but I can't very well shoot anything that's moving with my tech cam (talk about contemplative shooting...). My brief experience with the X1D tells me that I'll see a difference here as well compared to the Nikon, in a nice compact package. But I'm skeptical that the step up from 50 to 100 Mp on the crop sensors will make a noticeable difference if not printing large. In the meantime there are a lot of great images to make!

My 2 ¢.

John

 

[Jamgolf]

I think there are two separate things you are asking about ...

1) merits of medium format digital backs over 35mm DSLR or smaller formats
2) merits of movements or technical cameras in light of technological advancements in resolution and auto image stacking etc.


Since you wish to hold on to your Nikon system for its autofocus and telephoto prowess, and are really only interested in medium format in the "35mm to 85mm equivalents" i.e. 50mm to 132mm MF lenses. It seems you are not really interested in wide angle Rodenstocks such as 23HR, 32HR, 40HR or 28SK, 35SK etc. that will limit your technical lens options to 50HR, 70HR & 90HRSW (120SK ASPH or 150SK if you can find a preowned one which is not easy). So, the technical camera argument becomes far less compelling for your stated usage.

And getting a medium format SLR camera such as XF or H6 or GX etc. perhaps is not of interest to you anyway and would be a weird setup to own only for "35mm to 85mm equivalents".

You also said "I no longer exhibit or make a pretense about galleries, all of my work is on my website and if it looks good on a 4k monitor that is all that really matters" - which is strong argument against a higher resolution camera. Thats like buying a Ferrari but driving it at only 60 mph. The biggest point of using a 100+ MP 54x40 sensor is to be able to print big. If you know you are not going to do that, then I think it is similar to the Ferrari analogy.

Now, having said that, there are people (like myself) who actually dislike the automation of everything and the possiblility of auto image stacking and automatic perspective correction and automatic this and that. For those people, technical cameras are tools that let us enjoy the photographic experience. But thats a very subjective reason to own such a camera - not objective at all.

 

[JohnBrew]

Good thinking here. I'm pretty happy with my RM3di and Rodie lenses for now. I think I would enjoy an X1D Series II but not necessarily at 100 mp. A refined 50 mp X1D might be all I need. What I don't need is to replace my computer so I can edit 100 mp files!
Should be an interesting year!
And while on the subject - a Happy New Year to everyone here on the forum and all the best.

 

[ErikKaffehr]

Hi,

I am pretty sure we will see 100 MP on 33x44 mm sensors. It is sort of obvious that they are coming as Sony published a roadmap with those coming in 2018. But sensor development takes time.

Both Hasselblad and Fuji designed their 33x44 mm mirrorless systems with 100 MP taken into account. Jim Kasson is a retired scientist and has run much testing on the GFX and those lenses are about the best ones he had in his hands. No, he has not tested the X1D.

I don't think we are going to see a lot of symmetric lenses on 33x43 mm, it seems the new designs are quiet complex. Look at the Fujinon 63WR. Essentially, it is a symmetric design, a double Gauss in front, with an anomal dispersion element added to the front group. The rear group is quiet complex.


Increasing MP is essentially always beneficial. All other factors being equal, a smaller pixel camera will always yield a sharper image than a camera with larger pixels. Think this way:

The optical system has a single task, to reproduce a point of the subject as a point in the correct place on the sensor. All parts of the optical path will cause some "smearing" of that point. The lens has residual aberrations, that smear the resulting point, to that comes diffraction, OLP filter if present and pixel aperture. All these effects reduce sharpness. Smaller pixels normally mean smaller pixel aperture.

Diffraction is a property of light. It will be little affected by optical design. Good lenses are normally diffraction limited at f/8 or often larger apertures. Adding more pixels will not affect diffraction.

If we pixel peep an image at actual pixels, we are looking at a small part of a very large image at decently close distance. If we say double the number of pixels, the magnification will increase 41%. So, if we look at 1:1 a 50 MP image may be 90" wide (on a 24", full HD monitor), the 100 MP image would be 127" wide. So the image may look soft on screen, but it depends on the increased magnification at viewing.

The real downside of having more pixels is that we also need more storage and faster computers.

So, I would say that 100 MP is coming. I would expect a modest increase of price on the coming models of the present 33x44 cameras as sensor size is same. So, I guess that the 100 MP Fuji model will still be well below 10k$US.

I would also guess that the 100 MP 33x44 systems will be able to compete head on with the 100MP 54x40 mm systems. At that time there will be 150 MP at 54x40 mm (again, based on Sony sensor road map).

Pricing will be affected, obviously.

I would guess that Hasselblad's main business will be 33x44 mm. I would be a bit surprised if Phase One went into 33x44 mm mirrorless, that business is lower price and higher volume and needs a new lens system.

Best regards
Erik

 

[Frankly]

I see this akin to the older arguments when the Nikon D800 was introduced and many people thought the 36mp pixel density would be detrimental or not as useful as expected... diffraction limited, high ISO noise, moire, giant file size being overkill for workflow, etc.

Few (none) of these problems materialized in real life.

But even for my sports and action work, seen on screen or modest print size, the D800/D810 series has worked out wonderfully as a general purpose workhorse that really had no competition until the Sony A7 matured. And even then....

And yes, even though I am an infrequent printer I do value the editability of my files and have a longer range is always better.

And even if I decide to stick with a lower rez 50mp version they'll be that much further along and a better value when the 100mp versions are out.

Right now my ideal would be a mirrorless MFD with a classic 35-40e prime and a short fast tele for portraits. Seems like we're getting close. @jamgoff thank you, it's clear the technical camera and SLR platforms wouldn't be ideal for me and I was looking for confirmation that I wasn't overlooking anything in my reasoning.

In the end it will probably come down to pricing, weighing the HB versus the Fuji while paying attention to whatever Phase decides to do.

 

[thompsonkirk]

I agree with what Erik has said above, except that "Increasing MP is essentially always beneficial. All other factors being equal, a smaller pixel camera will always yield a sharper image than a camera with larger pixels." Perhaps so, if in theory all other variables are equal and one imagines oneself producing mural-sized prints. But I don't think the 50 to 100 MP difference will matter for the OP's or most of our real-life practice.

For the OP's and many of our needs, the advantage of MF isn't just that it yields a sharper image. It's the smoothness and subtlety of tonal transitions that led me to digital MF. (Also what led to alternating Leica, Rollei, and sometimes LF in film days.)

I've had no problem making files with sufficient resolution for prints on 17" and 24" paper at normal rather than pixel-peeping viewing distances. (That's with A7R2 and Zeiss lenses, or A7r2/Kolari with Leica lenses.) But when I've rephotographed the same subjects with GFX and a number of lenses, the most notable difference for me isn't just sharpness. It's the subtlety of edge contrasts and the smoothness of tonal shifts, especially within highlights. It's very hard to predict that in practice, this will increase much – obviously not proportionally! – when MF sensors rise from 50 to 100 MP. So I wonder in whose work this will make a visual as opposed to a theoretical difference.

Particularly for the OP, who's concerned with monitor display, let's expect zero visible difference. I enjoyed his website – a strong one for sure – but I didn't see images that call for heightened preoccupation with resolution – for example detailed product photographs, or XL prints used for commercial displays.

Speaking for myself, my standard is that while I don't want digital printmaking just to mimic analog work, I do want my older analog images and newer digital work to maintain some stylistic continuity. That means scrupulously avoiding artifacts and using no sharpening on files made without AA filters – not even the 'default' raw conversion sharpening. Then for post processing I'm starting with what a nicely-rendering lens has put onto a MF sensor. Most of my prints go on 17x22" paper. I've printed larger on request but not because I thought it did more justice to my vision. It's at this level of practice that I don't expect important gains from a 100MP sensor.

YMMV,

Kirk

 

[ErikKaffehr]

Hi,

My personal experience is with the A7rII and an older CCD back, the P45+. Regarding prints, I normally print at 16"x23", it is called A2 here in Europe. At that size, I didn't see the advantage of the 39 MP P45+ and my "than" camera the 24 MP Sony Alpha 99.

A while ago, someone posted a few questions on the DPReview MFD forums, essentially the following topics:

1. I print at 40"x50", is the GFX beneficial compared to Sony A7rIII/Nikon D850?
2. At what size will the difference show?

DPReview published a couple of sample photos with the A7rII and the GFX and indicated that the GFX only had a small advantage. But, what happen when you print at 40"x50"?

I don't print at that size... But, I have found that on a "full HD 24" monitor an image resized to 40"x50" at 180PPI would be a decent approximation of viewing a print.

I made the following processing on both images:

• Develop in Lightroom with no sharpening or noise reduction
• Open in Photoshop and resize to 40" height at 180PPI
• Sharpen using FocusMagic with radius 2 and at 75%

Actual pixels crops are shown below:


I would say that the difference is significant. It could be discussed if the A7rII image could be made better with a bit of noise reduction and extra sharpening, but the same would apply to the GFX image.

The aspect ratio of the GFX is closer 4x5 than the aspect ratio of the A7rII. The comparison here was for matching vertical dimensions and here the GFX has a 17% resolution advantage, 6192 vs 5304 vertical pixels. The GFX lens is very sharp and it is also helped by the larger sensor dimensions.

Best regards
Erik

I agree with everything Erik said above, including the main point that "Increasing MP is essentially always beneficial. All other factors being equal, a smaller pixel camera will always yield a sharper image than a camera with larger pixels." But I don't think it matters for the OP's or almost anyone else's uses.

For the OP's and my needs, which do not include large prints, the advantage of MF is not particularly that it yields a sharper image. It's the smoothness and subtlety of tonal transitions that led me to digital MF – and what led to alternating Leica, Rollei, and 4x5 in the good ol' film days.

I've been making files with all the resolution needed for prints on 24" paper and normal (rather than pixel-peeping) viewing distances – using A7R2 and Zeiss lenses, or A7r2/Kolari with Leica lenses. But when I've rephotographed the same subjects with GFX, the notable difference is not about sharpness. It's about the subtlety of edge contrasts and the smoothness of tonal shifts, especially within highlights.

If resolution isn't one's holy grail and other aspects of image quality matter at least equally, then I really don't expect to see an impressive or even interesting improvement when sensors 'rise' from 50 to 100 MP.

Particularly for the OP, who's concerned mainly with monitor display, let's predict zero visible difference. I've enjoyed his website – a strong one for sure– and I just don't see the kind of images that call for 'resolution preoccupation' – for example detailed product photographs, or XL prints for home or boardroom decoration.

Kirk

 

[thompsonkirk]

I’m happy to concede the point, noting only that neither you nor I nor Frank, the OP, make 40x50 prints. I see these mainly at Fraenkel and Koch Galleries in SF and wonder who actually collects prints that size. It’s a world beyond me.

Happy new year all,

Kirk

 

[f8orbust]

I'm not interested in using an older MFDB system, I'm thinking that a 100mp CMOS sensor is the way forward in order to achieve a significant "jump" in quality over a modern Nikon/Sony FX sensor. I understand that there is also a quality and look to the older fat pixel CCD sensors but I'm not at the point where I want to indulge in "retro-digital" quite yet. Maybe in 2030 it will be the cool thing?

People are going to jump on me, but I believe the IQ of MF has not translated from CCD -> CMOS, even though 'numbers are just numbers'. It's rather like high-end analogue audio to digital. If you really want the 'MF look', pair an IQ CCD back with Schneider glass. Nothing else quite matches this.

My understanding is that these newer CMOS sensors do not play well with movements especially with wider angle lenses. And both Hasselblad and Phase One will be developing mirrorless systems (like the X1D) with the newest sensors and lenses specifically designed to take advantage of the lack of depth required for an SLR mirror box (less optical compromises).

That's correct, and I doubt they ever will - 99% of these backs are going to be used on an SLR body, so the manufacturers have no reason to go out of their way to make these backs behave well when shifted/tilted.

So does this mean that waiting another season or two for them to create a 100mp mirrorless with a 35mm equivalent prime non-retrofocus design lens is going to be pretty much the ultimate?

I wouldn't hold your breath.

Will there even be a purpose to technical cameras after this? With resolution overkill and automated focus stacking why would I want the complexity and chance of misalignment from a tech cam?

Yes, there are situations where focus stacking won't work, or distorting the image in post to fix 'technical' issues isn't an option. Also, the experience of using a technical camera - lust like a LF camera - can be a joy in itself. Hire an Alpa for a couple of days.

To me and probably a few thousand others if it means we can plunk down $30k and get the best of the best in Phase One's version of the X1D concept and be "done" isn't this the Holy Grail?

Slim-to-no chance that P1 will develop a mirrorless body and the lens ecosystem to go with it at a price point to compete with HB and Fuji. It's just not how they operate. At least, up to this point. If they do, it will show you that mirrorless has been hurting them, and that they figure this is where the market is headed. Regardless though, they'll never let you 'be done' ! Look at the trichromatic for instance. Because MP are out of their hands (now that Sony is R&Ding everything), and is going to max out at 150MP (DB) and 100MP (mirrorless) for the next few years, manufacturer's will come up with other ways to tempt you. Already people are dumping $$$$ to go from the IQ3 100 to the trichromatic version, which to me is just crazy.

The only thing sticking here is the cost, and that old professional habit of always having a duplicate back up system at such a cost!

Advantage here of the tech cam route. Buy a used P25+ and stick it in your bag.

Jim

 

[ErikKaffehr]

Hi,

I have the honor of having decorated a very long corridor at my workplace. The largest prints were 0.9x4 m that would be 35"x157", but those were stitched panos. Normal image were up to 31"x47", but those were on not very high quality canvas. One of the reason they asked me to provide pictures was that they were impressed by the technical quality.

It makes sense to have large images as wall decorations.

Best regards
Erik

I’m happy to concede the point, noting only that neither you nor I nor Frank, the OP, make 40x50 prints. I see these mainly at Fraenkel and Koch Galleries in SF and wonder who actually collects prints that size. It’s a world beyond me.

Happy new year all,

Kirk

 

[vjbelle]

Hi,

Increasing MP is essentially always beneficial. All other factors being equal, a smaller pixel camera will always yield a sharper image than a camera with larger pixels. Think this way:

The optical system has a single task, to reproduce a point of the subject as a point in the correct place on the sensor. All parts of the optical path will cause some "smearing" of that point. The lens has residual aberrations, that smear the resulting point, to that comes diffraction, OLP filter if present and pixel aperture. All these effects reduce sharpness. Smaller pixels normally mean smaller pixel aperture.

Diffraction is a property of light. It will be little affected by optical design. Good lenses are normally diffraction limited at f/8 or often larger apertures. Adding more pixels will not affect diffraction.

If we pixel peep an image at actual pixels, we are looking at a small part of a very large image at decently close distance. If we say double the number of pixels, the magnification will increase 41%. So, if we look at 1:1 a 50 MP image may be 90" wide (on a 24", full HD monitor), the 100 MP image would be 127" wide. So the image may look soft on screen, but it depends on the increased magnification at viewing.

Erik

I don't understand or agree with your disregard to the effects of diffraction. Any reputable lens test always gives reference to the aperture that will be diffraction limited and the imatest charts always show how the sharpness of the lens is reduced when diffraction limits are reached - and its rather dramatic.

I would agree that more pixel are always better but onlh if they are the same size pixels. If smaller pixels are used to obtain a higher MP sensor then diffraction will appear sooner (at a larger aperture). I shoot my Schneider Digitar lenses at F11 which is diffraction limited even with my 100MP back. However, I'm willing to forgive that small amount of diffraction (I'm just at the beginning of diffraction) for the added pixels. My Digital back has pixel dimensions of 4.6 microns. The upcoming 150MP/100s back will have pixel dimensions of around 3.7 microns. That, for sure, will cost me an F stop for the same sharpness - an F stop that I can't afford not only for optimal lens sharpness but also for DOF.

You can find an in depth discussion regarding diffraction here (https://www.cambridgeincolour.com/tutorials/diffraction-photography.htm).

Victor

 

[MGrayson]

That, for sure, will cost me an F stop for the same sharpness

Victor

Only if you demand the same pixel-scale acuity. At the same print size, it costs you nothing and you still gain something.

Why not accept that there are tradeoffs throughout a system, and the optimal result is almost never achieved by sacrificing every other parameter to maximize a single one? To say nothing of the fact that we're arguing about tuning the piano and ignoring the performance.

I recommend Peter Serkin's recordings of the late Beethoven Piano Sonatas on a piano designed just after Beethoven's death. Once you get over the fact that it sounds like a bad toy instrument, the music comes through more beautifully than 99% of performances by modern athletes on modern instruments.

Having said that, I'll buy the next higher-pixel Leica S if it ever appears

--Matt

 

[ErikKaffehr]

Hi,

It is not the way it works.

See it this way: The lens is a part of the imaging chain. Each part of that chain has a measurable MTF. So the lens has an MTF which of course includes diffraction. But we can split the MTF of the lens into MTF_aberrations and MTF_diffraction. MTF_aberrations is increasing while stopping down, while MTF_diffraction goes down when stopping down. Lens MTF is MTF_aberrations * MTF_diffraction, that curve typically has a maximum. The best lenses today may reach maximum between f/2 and f/5.6.

Note that lens MTF is a propert of the lens, it has nothing to do with the sensor.

An imaging chain may include an OLP filter, that filter also has an MTF.

Finally, the sensor has also an MTF. That MTF depends solely on the pixel aperture. It is quite obvious, the size of the opening of the pixel decides the smallest detail the sensor can resolve. Smaller opening means higher MTF. In general smaller pixels have smaller apertures, although it is possible to have smaller aperture than the pixel size. That would increase MTF.

The MTF of the system is given my multiplying the MTF values of the chain. The only part of that depending on pixel size is the MTF of the pixel aperture and that increases with reduced pixel size.

When I got my P45+, I did an aperture range testing with the Sonnar 150/4 mounted:

You can see that LW/PH peaks at f/5.6. Stopping down to f/8 starts beginning to take it's toll even on 6.8 micron pixels.

Of course, the more you have, the more you loose. But, what remains will always be more. Those are real world facts.

Brandon Dube is a young optical engineer who works with OLAF lens testing and also does optics simulation for NASA. He has a nice write up on those things, here:
https://www.lensrentals.com/blog/2017/10/the-8k-conundrum-when-bad-lenses-mount-good-sensors/

I take some images from Brandon's write up:

Point image (PSF) of a perfect lens f/1.4 lens:

The same, with an OLP-filter:

Real world PSF on the optical bench without and with OLP:

PSF of a 5 micron pixel:

MTF curves, lens, lens + OLPF , lens + OLPF + Pixel

Clearly, the Pixel aperture is the dominant factor.

Now, look at an excellent lens with different resolutions:

And a good lens with different resolutions:

A lens that is "OK":

Here the lens is the limiting factor and sensor resolution still improves MTF, but not very much.

I don't think Brandon's write up discusses diffrcation. But, we have some older data from LensRentals. That data is on 24x36 mm.

Check the curves for the Nikon D700, Nikon's large pixel camera, from this posting by Roger Ciala: https://www.lensrentals.com/blog/2013/03/overcoming-my-fentekaphobia/

Let's for simplicity take the average curve. It peaks at f/8 and reaches say 695 lp/PH. At f/16 we have around 590 lp/mm.

Now, look at the D800 data:

The average curve peaks at f/5.6 with 950 lp/mm (or so). At f/16 it still has around 750 lp/mm.

So optimum performance increases 100 * (950/695 -1) -> 46%. Diffracted performance at f/16 increases 100 * (750/590 -1) -> 27% and the D800 will resolve higher at f/16 than the D700 at f/8.

Best regards
Erik

I don't understand or agree with your disregard to the effects of diffraction. Any reputable lens test always gives reference to the aperture that will be diffraction limited and the imatest charts always show how the sharpness of the lens is reduced when diffraction limits are reached - and its rather dramatic.


I would agree that more pixel are always better but onlh if they are the same size pixels. If smaller pixels are used to obtain a higher MP sensor then diffraction will appear sooner (at a larger aperture). I shoot my Schneider Digitar lenses at F11 which is diffraction limited even with my 100MP back. However, I'm willing to forgive that small amount of diffraction (I'm just at the beginning of diffraction) for the added pixels. My Digital back has pixel dimensions of 4.6 microns. The upcoming 150MP/100s back will have pixel dimensions of around 3.7 microns. That, for sure, will cost me an F stop for the same sharpness - an F stop that I can't afford not only for optimal lens sharpness but also for DOF.

You can find an in depth discussion regarding diffraction here (https://www.cambridgeincolour.com/tutorials/diffraction-photography.htm).

Victor

 

[vjbelle]

Erik...... I always appreciate your contributions. I know just enough of this to be a little dangerous. The absolute bottom line for me is the print and how all of this will manifest itself at my print sizes. I'm always printing at 40 to 44 inches on the long side. With my current back (Phase 3100) this means almost (0) upsizing (native size of a file is 38 inches on the long side). I shoot at F11 almost exclusively and can see the effects of stopping down further on screen and print. If a 150MP back would keep me at the same level of sharpness at the same print sizes then I would consider that keeping me whole. However, and this has nothing to do with diffraction, it would never be worth $25,000.00 out of my pocket. If a GFX at 100MP can give me the same image I am getting now at the same aperture and size then that would be a very attractive system for me as I would be able to take advantage of all of my Schneider lenses and the movements of my Actus.

Thanks again for taking the time to respond.......

Victor

 

[Bernard]

Note that lens MTF is a propert of the lens, it has nothing to do with the sensor.
(...)
Finally, the sensor has also an MTF. That MTF depends solely on the pixel aperture.
(...)
The MTF of the system is given my multiplying the MTF values of the chain.

Erik,

You are over-simplifying. System MTF is strongly affected by ray angle. A very high resolution lens, like a Biogon, used with a high resolution sensor, will yield poor system MTF if they are not matched.

Your calculations are based on the assumption that light rays are perpendicular to the sensor plane, which they never are (outside of the optical centre of the image at small apertures).

We will all need to wait and see how the next generation of sensors works with current/future lenses. Those sensors may be more sensitive to lens types, or less so. They may even vary by application: for instance, Fuji may use a different protection glass than Hasselblad, effectively changing the optical formula of your lenses.

 

[jng]

If a GFX at 100MP can give me the same image I am getting now at the same aperture and size then that would be a very attractive system for me as I would be able to take advantage of all of my Schneider lenses and the movements of my Actus.

My intuition (often wrong, also falling within the area of a little knowledge being dangerous) tells me that you'll do better with 100 Mp on the larger sensor than on the crop sensor, at least when printing large, since the overall magnification will be less.

Thanks as well to Erik for his informative post re: MTF curves and overall system resolution. The comparison of the same lens on the 12 Mp D700 vs 36 Mp D800 is particularly interesting to me, given my own personal conundrum: I love the smooth tonality of the images from my old D700 (since sold, regretfully) but find the D800E to give a more clinical or "digital" rendering, which I never warmed up to (it's not all about resolution!). Thus started my journey down the MFDB rabbit hole...

John

 

[Frederic]

[...] Another caution, if you have large-format IQ expectations, is to be cautious about the latest in MF lenses, almost all of which err in the direction that folks call 'clinical,' with emphasis on resolution and micro contrast rather than on '3D' rendering. I quickly gave up on modern primes, stepped back to older 645 Contax lenses, and then found (to somewhat embarrassing surprise) that the basic GFX zoom lens is 'gentler' than the new primes and not too different from the older ones – and a lot easier to pack around than an extended lens set.
[...]

Kirk

Some very good point here, especially regarding the clinical rendering of modern MF lenses (which Capture One tends to exacerbate even further).

While my Rodenstock and Schneider/Phase lenses fulfill my architecture and repro needs, I just can't stand their high level of micro-contrast for portraits and even landscapes. Textures and small details always look harsh and messy/fuzzy to me unlike the ones seen in 8x10 scans (I'm using a 100MP back, hence the comparison). For some reasons this phenomenon was way less pronounced on Aptus backs.
My 'solution' for portraiture is now to use 6X6 and 6x7 lenses on Mamiya 645 bodies through adapters. The Pentax 105 2.4 has a gentler and more filmic look compared to the Phase 110 2.8 for instance, even when shot at the same aperture. And the Hassy Distagon 50 makes a great job for environmental portraits, unlike... my HR50 !

Back to the OP question, I'd advise he seriously considers and tries both the X1D and GFX. Since I rely a lot on flash and adapted lenses I'd probably favor the GFX, at least in a business context.