Digital medium format in 2024 – it's not dead!

[Paul Spinnler]

Yes, CI published info on this a few days ago.

Next stop: 90 HR, 150mm, XT XL.

Gonna be X-Mas for tech cam peeps in P1 land!

 

[buildbot]

And I have to repeat it apparently ad Infinitum too that it is not realistic to have an IQ5 with global shutter anytime soon.

The massive circuitry behind the current sensor is laid out to maximise dynamic range (the amount is pixel can absorb is called full well capacity) which is absolutely crucial for photography. Combined with an x shutter you have solved both the flash sync, motion skew problems while being able to benefit from the high dynamic range setup.

If you now want a global shutter you need to replace part of the circuitry to enable read out simultaneously of every single pixel. You can still have remaining circuitry for dynamic range, but lesss so.

It is my understanding that the new Sony sensor starts at ISO 250 and the dynamic range is not the same as the best normal sensors. Since the IQ4 with its big case is already beyond anything in the market in terms of DR it will be very difficult and unlikely that we see an IQ5 with GS soon. The new sensor precisely demonstrates that there is a trade off and that they haven’t been able to add readout circuitry while keeping the full well capacity the same. That why ISO starts at 250 and not 50. It’s also a whole different matter to do this on a more than double as large sensor with a multiple of pixels. The A9 has 24 megapixels and already the design compromise couldn’t be resolved while keeping dynamic range intact. It means we are still far away from a 150 or 200 megapixel back with GS and equivalent or better DR.

People repeating this dream just have no idea how sensor design works; it’s a bit annoying because it seems like this typical forum myth that’s perpetuating in the fantasy of people and that’s alive because no one really understands how these things work. Just because there has been a camera now that came out it doesn’t mean it can be implemented in a useful way on the scale of an IQ5 or with the requirements of DR in mind.

It's just a dream first of all. I don't expect Phase to suddenly have a perfect 200MP GS sensor. In 2025 or anytime soon. People keep bringing it up to your apparent annoyance because they want it so badly, and I think most of us are aware we would trade some DR for that ability. It seems that the ability to IDK, take multiple frames for increased DR drops out of peoples heads. So maybe a GS back has worse dynamic range - frame average more. Dual Exposure? Why Quad exposure! Align in camera like an iPhone does for increased dynamic range.

I do know how sensor design works, my education is digital electrical engineering. In undergrad, I worked on building readout systems for prototype ccd and cmos sensors that were to be used for the ATLAS detector at CERN. I am well aware of the tradeoffs required in terms of area. So at least one person bringing this up knows something.

What's changed is the ability to stack silicon and build 3D structures into your chip in a much more vast way that was possible before. You can have your pixel well silicon on top, and your readout circuitry below. You can even do Pixel->RAM->Readout circuitry, to limit having a readout channel per pixel. From the IMX661 sensor info page (The one used in the iXM-GS120 Phase One Back !):


Yes the first mainstream photographic GS sensors is limited compared to generations of improvement, first in rolling shutter CCD, then in CMOS sensor design. With the physical limitations being reduced with stacking, I expect we will see big improvements in the GS sensor space. This Sony generation has 80db if dynamic range, current gen and not stacked GMAX sensors top out around 70db. Will the next generation be at par with rolling shutters? Probably not, but the gap will close again. I'm also pretty sure that 80db of dynamic range is probably more than an old "fat pixel" 22MP CCD back puts out...which people seem to find fit for photographic purposes still.

If I am wrong I'd love to be corrected and learn though!

 

[Paul Spinnler]

No, no, that's right – its just that it isn't around the corner at all and given the strong push from them to sell their x shutter lenses, P1 just re-released all XT lenses with MK II shutter at eye watering prices, it seems to me that for the next gen the focus will be more resolution and / or DR + new electronics (I/O, battery life, on-board processing). The readout speed might improve further to the point where the effect of GS will be even less important and maybe, just maybe, the combo of higher end ob-board processing and readout speed combined with P1's image processing know how might lead to new DE+ modes, who knows.

It's just that the notion that there's this A9 III out there and that P1 will now implement this in an IQ5 is just not realistic. The compromise is still a compromise, despite advances in stacked sensor design.

It may still take a few years and in-between we just might have an IQ5 with 200 megapixels, slightly improved DR and better SoC and if that's the case I am still very happy even if there's no GS sensor. In the meantime P1 will have a fully developed XT tilt lens lineup which will be a great set of lenses useable without cables.

So essentially it looks like the next year is another one of those:

INVEST IN LENSES YEARS BECAUSE YOU DONT NEED EXTRA CASH FOR A BACK. OR A LENS + CAMERA BODY YEAR.

This year had also quite a few SK Magic Four lenses come to market (B3K even had a factory new SK 120 Asph (which sold for 8k methinks)), which to me was a sign of the economy and I would think in that regard it was an exceptional year of opportunity to get some SK gems ... not sure we'll see that many next year!

I also hope the rumour around the SK 150mm is true ... in X shutter with TILT!

 

[ThdeDude]

Not having read all the previous postings, even a new DB (IQ5) using the current sensor could offer major advances as to usability over the IQ4-150.

 

[Paul Spinnler]

Not having read all the previous postings, even a new DB (IQ5) using the current sensor could offer major advances as to usability over the IQ4-150.

There's many angles they could tackle.

Even a slot on top of the back to add an EVF like the one in the M11 would be amazing. The M11 EVF connects to a hotshue connector port and maybe that could be the solution for a compact camera. Imagine an XC or TC with a back which a detachable compact 2-4k resolution EVF.

Better SoC allows for many options in firmware ... they could also work on colour science like Arri with Reveal.

 

[jduncan]

Ok, so we are closing in towards the end of the year.

Time to summarize what's coming in 2024 and analyze what this means for the high-end medium format market:

Phase One:
+ 70 XT tilt was just announced with MK II shutter assembly and 5 degrees of tilt (vs. 3 degrees on the 32, 40, 50 tilt XT variants)
+ 90 XT would be the next natural announcement, I would think sometime beg. of next year
+ XT XL fully integrated large tech shift cam (my hope is 22.5mm left and right, my hunch is 18mm left and right as to be differentiated from the Cambo product line – would be a shame)
+ XC – standalone availability as soon as China sales taper off
+ Digital back – probably 2025 due to lack of new sensors; XT XL, 70 and 90 should give P1 a good cadence for the year
+ Plus variant a possbility, but I would assume they'll aim for an IQ5, ideally with higher DR, better I/O, battery, on-board processing
+ XF: Let's just hope it is not discontinued

Hasselblad:
+ New 100 megapixel V mount back at an affordable price point – this should be a boon for Alpa and other open tech cam manufacturers

Arca:
+ There's this new compact mini-bellows camera – unsure about the value proposition
+ Mostly sees itself as an accessories manufacturer with main focus on tripod heads

Alpa:
+ Continued focus on expanding media presence and focused on new Hassy back (their IG is alive and they send out demo kits to youtubers, see the latest Peter McKinnon feature for example)
+ Focus on luxury segment - highest end tools and workmanship for a select clientele which appreciates beautiful cameras; you pay for a camera and a beautiful object

Linhof:
+ Is owned by one of the wealthiest Munich based real estate owners, so they'll continue to be here, but no innovations expected; still wouldn't wait forever to order a MT3000, there are no guarantees ...

Cambo:
+ No new developments beside JV with P1; they have the broadest manufacturing capabilities for tech cam stuff, including a whole range of helicals!

Sinar:
+ Still alive – selling new and refurbished P3s!

xxxxx Crop segment xxxxx:

Leica:
+ S4 in development, expected in 2025. First new medium format system since HX ...
+ "Ultimate Leica" – compatible with all lens families, M, SL, S vintage, and S next gen all in one body

Fuji:
+ Cementing the lead in crop MF and MF if you consider absolute sales volumes. Workflow, workflow, workflow. Best way to move up the ladder from Sony and CaNikon, price-wise outright affordable compared to the traditional tech cam approach

xxxxx Lenses xxxxx

Rodenstock:
+ No new lenses given lack of market viability; there's a 25mm 90IC lens in the drawer and at one point ruminations about an 70-SW, but would not expect anything to materialize

*****
Lens prices

1. Rodie HR: lenses have increased in price and used market is dry; ie people are holding onto their lenses still = stable prices
2. SK: Magic four expected to increase as economy improves given rarity; non-magic-four slightly lower in price, e.g. 47 digitar or rarely used teles

Nothing beats Rodie HR and SK late stage lenses in terms of high resolution and large image circle; even Fuji T/S has IC of 80 only. Interestingly enough, Rodie HR lenses are very rare on eBay still unlike the prior gen sironar digitals, for example.

There is no alternative to these lenses if you want to achieve the highest resolution imagery today incl. ability to shift and stitch large images for fine art applications.

*****

So this leaves us with P1 holding the torch, followed by Hassy's new back which is expected to drive sales of non P1 tech cams. Lens prices stable given lack of alternatives and continually increasing production costs.

It is a nice glimmer of hope that P1 is still investing in photography.

Fingers crossed for an IQ5!

Hi happy holidays.

My question will we see new sensors? If the market is stuck with 100 and 150 MP sensors forever it will become more difficult for each generation for MF to compete.
Hasselblad did a very good job with the New firmware. Since we are not taking action pictures with the X cameras the X2D feels modern. I will say it's the most important development of the year. We now have two companies producing "modern" MF systems. Like you I expect Phase One to continue adjusting to the market next year.

Some challenges from my perspective:
1. Continue to improve software, as Hasselblad did this year.
2. Low volume, microelectronics is capital intensive, and it's difficult to keep up when one sales in a year what Nikon sales in one week (or less), even if Nikon is tinny.
3. MF no longer has an advantage in terms of flash sync speed.
6. Next year or the following if nothing pretty bad happens, Canon is likely to intrude a 100mp camera. needed to respond the the Sony A9 III could delay the introduction.
7. Global economy at peril from war and social and economic unrest.

Best regards,

 

[buildbot]

Not having read all the previous postings, even a new DB (IQ5) using the current sensor could offer major advances as to usability over the IQ4-150.

Better SoC allows for many options in firmware ... they could also work on colour science like Arri with Reveal.

Ah my personal annoyance (with Phase One).

They created this “infinity” platform using Linux and an AMD (née Xilinx) SoC, then preceded to minimally update it. Frame averaging and dual exposure are awesome features and the continuing firmware updates are nice.

But it could be so much more!!! Open up the platform, high more software devs for updating and improving the existing firmware, etc. Maybe even live up to the bespoke name and get a little bit of the recurring revenue pie by offering very expensive subscriptions that come with the ability to request features, with a 1-2 month turn around.

They don’t need new hardware. The hardware is excellent and newer versions of the sensor and SoC don’t exist even. The only thing that needs upgrading is the wifi, which shouldn’t be a new IQ5 IMO. If they used the same setup as the IQ-3 series, the wifi adapter is literally a USB module which could be easily replaced/upgraded.

 

[Pieter 12]

Being a cynic, the MP sensor race seems pretty ridiculous to me. For what? Most imagery today is viewed on a screen, and there are limitations to screen sizes and resolution. Plus, the larger the screen (or print for that matter), the longer the viewing distance, so at a certain point resolution is moot. I would venture to say the primary market for high-resolution images is retail, in-store displays where large images can be presented and fine detail appreciated. Sure, fine-art photgraphers might benefit, but that market is pretty limited. A lot just comes down to GAS and bragging rights. To be truthful, the overwhelming majority of images I see done with the proliferation of high-end cameras is mediocre hobbyist work, at best.

 

[Paul Spinnler]

Being a cynic, the MP sensor race seems pretty ridiculous to me. For what? Most imagery today is viewed on a screen, and there are limitations to screen sizes and resolution. Plus, the larger the screen (or print for that matter), the longer the viewing distance, so at a certain point resolution is moot. I would venture to say the primary market for high-resolution images is retail, in-store displays where large images can be presented and fine detail appreciated. Sure, fine-art photgraphers might benefit, but that market is pretty limited. A lot just comes down to GAS and bragging rights. To be truthful, the overwhelming majority of images I see done with the proliferation of high-end cameras is mediocre hobbyist work, at best.

Also most gallery sold "fine art" is mediocre copy paste, especially landscape stuff. So many boring "pier", "iceland" and "treescape" fotos done with frame averaging. I think at this stage the majority is just enthusiasts having a nice hobby and spending cash on photo gear rather than watches or cars.

Regarding resolution and end format – yes, you probably could have a nice photography portfolio site made with midjourney 6 imagery.

What I am saying: its clear that most P1 systems are not used to the fullest potential, worst case to photograph flowers in the backyard.

 

[Mexecutioner]

worst case to photograph flowers in the backyard.

Or cats

 

[akaru]

Hey, it ain’t easy shooting a cat indoors with a tech cam at f8. Not that I would know…gotta sell Matt on this for some 40” x 60” Soup prints.

I’d be surprised for even many gallery shots to be done with an iq4. I’m more bothered by the Sony/Canikon 8x10 prints at landscape tourist spots that are shot straight into a sunset too early, full of noise and flare, and *gasp* even pixelated at times.

 

[Paul Spinnler]

Can we agree that owning a P1 system and a tech cam at this stage is not forcibly rational.

This forum’s mantra is and has been “Dante’s inferno”; to me, it is a combination of fascination for great optics, beautifully made cameras and of course the pursuit and passion for photography.

But it is a combination of all of these elements. No illusion that in today’s commodified world it is near impossible to produce art that is truly innovative.

We are walking on the shoulders of giants!

 

[jng]

Can we agree that owning a P1 system and a tech cam at this stage is not forcibly rational.

This forum’s mantra is and has been “Dante’s inferno”; to me, it is a combination of fascination for great optics, beautifully made cameras and of course the pursuit and passion for photography.

But it is a combination of all of these elements. No illusion that in today’s commodified world it is near impossible to produce art that is truly innovative.

We are walking on the shoulders of giants!

I agree with your first and last claim, but can't readily accept your penultimate premise about today's near-impossibility of producing art that is truly innovative. This view reminds me of the pioneering molecular biologist Gunther Stent who, after making many seminal discoveries in the late 1950's and early 1960's about the nature of genes and how they are expressed, somewhat infamously held in his book “The Coming of the Golden Age: A View of the End of Progress” published in 1969 that the field of molecular biology, like all human endeavors, had hit a dead end (he similarly made the same claim about music). Stent's prediction about molecular biology, at least, was dead wrong, as he would later admit.

Perhaps Stent's thesis reflected the fact that he had become bored with the field he played a major role in starting, an intellectuall restlessness not uncommon among brilliant minds. But I think it also points to the bar of novelty being raised with each discovery, insight or innovation, with the next tangible leap in progress not made in increments but because of some disruption.

Since Stent's time, the ongoing series of successive and generative revolutions in molecular biology (sequencing of full genomes, gene editing, structural biology, RNA vaccines, to name just a few) have been enabled through a virtuous cycle of disruptions in technology that allowed new and unexpected discoveries, which in turn inspired the next generation of technologies for discovery. And so on. I suppose the question for me is whether the incredible new tools that photographers have at their disposal today will inspire new art forms. While I cannot answer that question, at the moment I also cannot accede to the notion that creativity in photography is dead or impossible to attain. RIP Gunther Stent.

John

 

[rdeloe]

OK, so now that we've gone here....

To me, it does feel a bit like the end of the road for many genres of photography. Representational landscape photography is a good example. I have a subscription to On Landscape magazine, and I quite enjoy it because every once in a while I see something that I find a bit new and a bit more interesting. However, a lot of the time I will open a new issue and see very competent photography that I have seen many times before from different people (including myself). In recent issues, I have even seen the same basic photographs in the same issue, but from two or three different photographers. In one recent issue, I could have pulled out the photographs from two different people and assembled them in a single group and they would have looked like the work of the same person.

We can all speculate why this is happening. My diagnosis is that it's a combination of the following: (1) there are a lot of really good photographers out there producing good work; (2) if your goal is to show the world as it is (i.e., representational photography rather than 'digital art'), there's only so much variety in the world, so subjects get repeated; (3) there are certain compositional devices and other "tricks" or "tools" that work well in photography, and skilled photographers deploy them; and (4) even leaving aside trends and fashions that some people follow slavishly, certain kinds of images seem to catch the eyes of photographers, and are repeated endlessly.

For a forum like this, an awkward truth is that better gear does not solve the problem. When I'm looking at a photograph that looks like lots of other photographs I've just seen, the fact that it was made with the very best equipment one can use doesn't make one iota of difference if the photographer who made the image is trapped in my diagnosis. A case in point for me is the fact that the most interesting and authentic photography I've seen in On Landscape in a while was made by someone using a Panasonic Lumix G9 Pro.

Personally, I think John @jng is on the right track: the tools we have today, including generative AI, are going to inspire new art forms that don't look like what many people currently think of as "photography".

My crystal ball isn't better than anyone else's, but when I look into its depth, I see a future where the kind of photography that many of us on this forum enjoy will occupy the same kind of niche as large format film photography occupies today. In other words, it will be an anachronism because people will have moved on to other ways of expressing themselves. Those of us "left behind" will continue enjoying what we do. That's my plan.

 

[Mexecutioner]

Can we agree that owning a P1 system and a tech cam at this stage is not forcibly rational.

I agree 100%. Every year, when I renew my equipment insurance policy, and I am reminded of the equipment list, I shake my head in disbelief and wonder why I did it, but as long as it keeps bringing me joy I have no issues with it bending logic and making irrational decisions.

 

[Paul Spinnler]

I think two things are at play, namely democratization of information and accessability of tools:

In the 80-90s, if you were one of the few people wandering around with an 8x10 camera and shooting sth particular, let's say insides of concert halls, you were most likely one of very few. With the advent of large scale printers like the lightjet and in combination with cutting edge drum scanners and emerging image manipulation software you could create sth people would go into a museum and marvel at as they hadn't seen sth like this. Nowadays everyone can shoot a detailed concert hall shot with a P1 camera and print it large.

Inspiration, behind-the-scenes info and accessability have led to a proliferation of the production of fine art. Back then you needed to go to an art school, meet the right people to get inspired and be lucky to see and understand all the advanced options. If you put 1+1 together you created sth rare.

Now you watch a youtube video or read a mag, get inspired, buy the gear, travel there, make the shot, print it post it on insta. Boom.

The amount of effort and accessability are completely different.

Its the same with music production – it was an arcane art up until the end of 2010s and then youtube, tutorials and templates came and nowadays everyone can make great sounding generic music of any type. All info is there.

You can still be innovative, of course, but it requires more work. You can shoot film and document human condition by doing crazy staged shots with actors and not many will do that. You can also buy the XT, travel to Iceland, join a workshop and post your own frame averaged mountain shot. Many do that.

 

[Pieter 12]

Sometimes, it just comes down to having an eye and an idea. Talent is important, skills follow and gear is pretty much interchangeable.

 

[darr]

While it's true that technological advances have made creating art more accessible, this democratization doesn't diminish the value of creativity. It's not just about the tools or the ease of access; it's about what you do with them. Yes, anyone can take a high-quality photo or produce a piece of music, but it takes a unique vision and creative spirit to turn something ordinary into something extraordinary. In a world where everyone has access to the same tools, the true differentiator is individual creativity and the ability to see things in a new light.

 

[Thyl]

It is my impression that without a major breakthrough in the MANUFACTURING of sensors, we will not see much progress. Here is why.

-The abilility of silicon to absorb photons and produce electrons from this is limited. It is a surface effect, and hence, the larger the surface of a pixel, the more photons it can absorb. There are some tricks to increase that number, like dual gain, but not many. Most of the progress made is only for sensors with small areas for each pixel. But in the end, we need surface area. The only significant improvements can come from increasing the sensor size (well, like with film). Incidently, this is not only of importance for the dynamic range, but also for the physical gamut (norrow band dyes for the Bayer filters), if I am not mistaken.

-The costs for sensors is mostly governed by their area, and this is something that appears to increase exponentially with increasing area. There is no way to come up with a larger sensor at a price point that photographers would be able to pay (like north of 100k) - with present day manufacturing.

-besides, Sony will and can not help. They apparently have lost interest in medium format, reducing the number of photographic sensors advertised on their website from 4 to 2 (and a global shutter industrial sensor with lower dynamic range). But even if tehy hadn't lost interest, what should they do? Improvements go into smartphone sensors (where the money is), and this will be of limited benefit for sensors already having a larger area. You need to note that Arris Alev IV sensor was used for TEN years. And Arri is not price sensitive.

-Also a cooperation of the surviving medium format manufacturers to produce an improved, joint sensor cannot overcome the exponential price increase per area. So, this would not help improving the sensor tech.

in summary, imho, the only way forward in terms of dynamic range and gamut are sensors with larger area per pixel, but this is prohibited by the cost of manufacturing such sensors. It hence appears that we have to get away from lithography, and find a cheaper method of manufacturing such sensors. Printing and stamping come to mind, and actually, just now, Canon seems to have established such a novel manufacturing method. Alors, here we sit and have to wait for a miracle to happen.

 

[DNN]

Alors (Canon) isn't a manufacturing method. Its a method of breaking up the pixel surface area so different exposures are computed. A good review is at https://global.canon/en/news/2023/20230112.html .

The cost of a bigger die isn't exponential with die area increase - it is a function of defect density, and edges(partial die areas), and the formula is based upon die area, so is an x^2 function, Its also has a factor related to wafer volume as the fab costs, both physical plant and wafer making costs, have to be paid by the production of a wafer. So volume and defect density can be of importance for costing new sensors.