Which lenses will be up to 150Mp backs?

[narikin]

Seeing how many of my (dedicated digital MF) lenses struggle with even 100mp back, I'm wondering who/what will be up to the upcoming 150Mp sensors?

Your technique needs to be spot on to extract the maximum from these sensors already, and as resolution is about to jump 50% (!) I'm wondering whats up to the task, or if any lens maker will step up the plate on this one. My 90mm HRSW Roddy is probably there or close to, but that's a big tech lens (with big Image Circle) Not for backpacking. The Rodenstock HR Digaron-S (zero movements /IC) were really great, but they don't make them anymore.

An updated Digaron S line would be great? Or an MF Otus from Zeiss? Dream on... I know.

Forget about legacy lenses. I tried a Hassy 60mm CFi recently, on IQ100, and it was hopeless. Supposed to be one of their best resolving lenses, the late model with redesigned optics, but it was very soft with lots of aberrations. No comparison to 60mm Roddy.

 

[Shashin]

Going from a 100MP to 150MP sensor is only a 23% increase in resolution. Resolution is a linear measure, not an area one. You are now dealing with a 3.8um pixel. That pixel size would be 60MP on a 35mm sensor.

Still, the increase in resolution will not make any of your images any softer in appearance. So, what your optics gives on 100MP is not diminished at 150MP. You may simply move from sensor-limited resolution to optics-limited. But with more sophisticated computational photography, it might not be a big deal.

Still, you have a point--what is the point of just continuing to add pixels to silicon wafers?

 

[tcdeveau]

Going from a 100MP to 150MP sensor is only a 23% increase in resolution. Resolution is a linear measure, not an area one. You are now dealing with a 3.8um pixel. That pixel size would be 60MP on a 35mm sensor.

Still, the increase in resolution will not make any of your images any softer in appearance. So, what your optics gives on 100MP is not diminished at 150MP. You may simply move from sensor-limited resolution to optics-limited. But with more sophisticated computational photography, it might not be a big deal.

Still, you have a point--what is the point of just continuing to add pixels to silicon wafers?

“Because they can” is good enough reason for me. Optics and computational algorithms will also continue to improve, and I also imagine the engineers at camera manufacturers will continue to squeeze info out of those pixels to make it worth our while.

 

[Shashin]

“Because they can” is good enough reason for me. Optics and computational algorithms will also continue to improve, and I also imagine the engineers at camera manufacturers will continue to squeeze info out of those pixels to make it worth our while.

But without improvements to the human visual system, there is little to gain from resolution after a point. Pixel peeping is fun, but images are to be shown. Even a 44" printer is not going to be able to print the detail in a 150MP file. I really like Supercars, but where exactly are you going to be able to drive it to its limit?

 

[tcdeveau]

But without improvements to the human visual system, there is little to gain from resolution after a point. Pixel peeping is fun, but images are to be shown. Even a 44" printer is not going to be able to print the detail in a 150MP file. I really like Supercars, but where exactly are you going to be able to drive it to its limit?

It’s not all about pixel peeping. Imagine the crop latitude - large panos without stitching, a prime becomes a zoom, etc.

You make good points regarding practicality, but that doesn’t mean higher resolution lacks utility IMHO. Higher resolution sensors will also mean current 100mp backs will get cheaper, which is what I’m really looking forward to
-Todd

 

[jng]

Forget about legacy lenses. I tried a Hassy 60mm CFi recently, on IQ100, and it was hopeless. Supposed to be one of their best resolving lenses, the late model with redesigned optics, but it was very soft with lots of aberrations. No comparison to 60mm Roddy.

Chosen carefully some of the old V system Zeiss lenses hold up remarkably well. But in general I agree that they are no match for the dedicated tech cam lenses, which are really in a league of their own, provided, of course, they are available in the focal length you want. This is why the 250 and 350 Superachromats are coveted by some on this forum - outstanding optics in the longer tele range not covered by high-end tech cam glass.

I compared my 40mm IF CFE (actually a more modern design, touted to be the sharpest of the V system lenses) against the 45mm XCD on my X1D. Although this wasn't done rigorously enough to call it a "test," sharpness seemed comparable but the 45 XCD showed no detectable chromatic aberration, whereas the 40 IF showed just a touch of longitudinal fringing toward the edges of the smaller sensor. And the 40 IF is a beast. I haven't compared directly the 40 IF with the Rodie 40 HR - the rendering of the latter is so drop-dead gorgeous that running such a comparison seems rather pointless.

John

 

[Shashin]

...Imagine the crop latitude - large panos without stitching...

A 14,278 X 4760 pixel sensor would certainly get my interest...

 

[stevenfr]

I agree the ability to crop a single image to make a large pano with the next chip is a huge benefit for sibjects that don’t stitch easily. However, I will still stitch panos for the larger file size. A larger sensor just gives more flexibility with the files.

Currently, I am getting 3 to 5gb on some of my stitched panos with the IQ3 100.

Steven

A 14,278 X 4760 pixel sensor would certainly get my interest...

 

[Oren Grad]

Even a 44" printer is not going to be able to print the detail in a 150MP file.

Assuming a 4:3 aspect ratio in the sensor, at an Epson's maximum setting of 720 ppi, 150MP is enough for a print area just shy of 15x20 inches. Even at 360 ppi it's just short of 30x40, which is within 44" on the long dimension and well within on the short dimension.

And that leaves aside the image quality benefit in color photography of "spending" nominal resolution by downsampling to eliminate the interpolation forced by the Bayer CFA. Take that into consideration as well, and the print size you can achieve with 150MP, if you are pursuing the maximum achievable technical quality, is actually pretty small.

Sure, the lenses may be a practical constraint, as will the increasing need for focus stacking if you want any semblance of apparent depth of field as the sampling rate goes up and the tolerable circle of confusion goes down. But at least on the sensor side, we are very, very far short of running out of the ability to make a visible difference in a print with more pixels, even with inkjet printing, which at the current state of the art is a pretty crude output medium.

That's not to say that every picture, every photographer or every viewer does or should need, want or care about that much descriptive power. But just as contact printing of sheet film negatives offers a distinctive and valuable "flavor" of expression within the medium of film photography, so too will ultra-high-resolution digital capture beyond what we have today offer distinctive possibilities for those who are interested in a certain kind of description of the world.

Ctein wrote about this almost ten years ago; though his calculation was framed in a slightly different way, the implications are the same:

The Online Photographer: Why 80 Megapixels Just Won't Be Enough...

 

[ErikKaffehr]

Hi,

If a lens is OK with 100 MP it will also be OK with 150 MP, Going from 100 to 150 is just a small improvement.

That is also saying that there will be little need to upgrade from 100 MP to 150 MP, especially as the 100 MP sensor is a pretty new design.

The 150 MP sensor will of course have new features.

Best regards
Erik

Seeing how many of my (dedicated digital MF) lenses struggle with even 100mp back, I'm wondering who/what will be up to the upcoming 150Mp sensors?

Your technique needs to be spot on to extract the maximum from these sensors already, and as resolution is about to jump 50% (!) I'm wondering whats up to the task, or if any lens maker will step up the plate on this one. My 90mm HRSW Roddy is probably there or close to, but that's a big tech lens (with big Image Circle) Not for backpacking. The Rodenstock HR Digaron-S (zero movements /IC) were really great, but they don't make them anymore.

An updated Digaron S line would be great? Or an MF Otus from Zeiss? Dream on... I know.

Forget about legacy lenses. I tried a Hassy 60mm CFi recently, on IQ100, and it was hopeless. Supposed to be one of their best resolving lenses, the late model with redesigned optics, but it was very soft with lots of aberrations. No comparison to 60mm Roddy.

 

[Shashin]

Assuming a 4:3 aspect ratio in the sensor, at an Epson's maximum setting of 720 ppi, 150MP is enough for a print area just shy of 15x20 inches. Even at 360 ppi it's just short of 30x40, which is within 44" on the long dimension and well within on the short dimension.

The PPI of a printer is not comparable to the DPI of an image--printers use a small point size to mix and lay down color, that does not mean it results in a 720 dpi image. Even if it did, 720 dpi would be beyond the resolution of the human visual system, which is my point. The 300 dpi has specific context: an 8 X 10 print viewed at about 12". A 16x 20 image at 150 dpi viewed at 24" would appear the same. Even if you half viewing distance, the print will still be resolved as photo quality (if you learn about viewing distance, it is not limited to standard viewing distance, but a host of other distances). A 30 x 40 print will not show all the details in a 40MP image--I know as that is my most common print size. So, I stand with what I wrote, unless you are making significant crops, you are just not going to see the benefits of a 150MP sensor with only a 44" printer.

 

[Oren Grad]

The PPI of a printer is not comparable to the DPI of an image--printers use a small point size to mix and lay down color, that does not mean it results in a 720 dpi image.

Indeed, PPI is not the same thing as DPI. The Epson printers deposit ink on the paper at a maximum of 2880x1440 DPI. However, the maximum PPI that the Epson driver can handle is 720. In fact, if you do send a file at 720 PPI with the driver set for "finest detail" and 2880/1440 DPI, the resulting print doesn't quite fully resolve the data sent, because of the fuzziness of the dot pattern deposited on the paper. However, the image is visibly cleaner and more detailed than it is at 360 PPI. I've done many identity-blinded comparison tests with my own files and printers (first a 3880, now a P800), and I can consistently tell the difference.

A further subtlety is also relevant here:

Even if it did, 720 dpi would be beyond the resolution of the human visual system...

Not quite. The extra information can still have visible effects even when it does not enable a higher number of distinct line pairs to be distinguished. More here:

The Online Photographer: Why 80 Megapixels Just Won't Be Enough...

...which is my point. The 300 dpi has specific context: an 8 X 10 print viewed at about 12". A 16x 20 image at 150 dpi viewed at 24" would appear the same. Even if you half viewing distance, the print will still be resolved as photo quality (if you learn about viewing distance, it is not limited to standard viewing distance, but a host of other distances). A 30 x 40 print will not show all the details in a 40MP image--I know as that is my most common print size. So, I stand with what I wrote, unless you are making significant crops, you are just not going to see the benefits of a 150MP sensor with only a 44" printer.

The term "photo quality" is ambiguous and therefore uninformative in the context of this discussion, as prints with widely varying levels of measurable detail and perceived acuity can be perceived as "photographic" and judged visually pleasing and suitable for purpose. A good contact print looks very different from a wall-sized enlargement from any possible capture, film or digital, though they may all be perceived as "photographic". But I disagree with your assertion about 30x40 prints. It's contradicted by my own tests and experience in making and viewing prints as well as by my understanding of the underlying science.

I won't belabor this further. At this point I see it as an argument not about what's technically feasible, but rather about how good is good enough, and there's no one right answer to that question. People will have different perceptions, different tastes, and different purposes for their work, all of which will imply different judgments of good enough, and that's entirely fine.

 

[voidshatter]

But without improvements to the human visual system, there is little to gain from resolution after a point. Pixel peeping is fun, but images are to be shown. Even a 44" printer is not going to be able to print the detail in a 150MP file. I really like Supercars, but where exactly are you going to be able to drive it to its limit?

My 5.5" iPhone supports a function called "pinch zoom". It can take as many pixels as it can hold in RAM. 150MP is nowhere close to making it choke.

8 years ago even 1920x1200 resolution over a 24" display is regarded good ppi. Nowadays not even 4k resolution over a 24" display is regarded high ppi.

Never underestimate future technology. Maybe your children will use what you shoot today as their digital wallpaper for decoration in their future houses, with dynamic zooming in and out.

 

[Shashin]

But I disagree with your assertion about 30x40 prints. It's contradicted by my own tests and experience in making and viewing prints as well as by my understanding of the underlying science.

The Online photographer seems to be taking about is detection, not resolution. Detection is indeed much higher than resolution. And it makes no difference in the discussion. Ultimately, a photograph is a perceptual target and most photographers place far too much significance to the technical attributes of an image. And maybe your experience is different, but the underlying science supports my position.

 

[Shashin]

My 5.5" iPhone supports a function called "pinch zoom". It can take as many pixels as it can hold in RAM. 150MP is nowhere close to making it choke.

8 years ago even 1920x1200 resolution over a 24" display is regarded good ppi. Nowadays not even 4k resolution over a 24" display is regarded high ppi.

Never underestimate future technology. Maybe your children will use what you shoot today as their digital wallpaper for decoration in their future houses, with dynamic zooming in and out.

But still you are left with the human visual system. Sure you can pixel peep, but then you can't see the image. Unless of course you are not making an aesthetic composition, but just want something like Google maps.

 

[cunim]

What we sometimes forget is that MTF is not limited by the worst component in the optical chain. Rather, it is a product of the chain MTFs. Therefore, there is always a benefit to increasing the resolving power of one component, even if the others remain the same. For example, we have a sensor/lens MTF chain of 0.8 x 0.8 = 0.64 for a given spatial frequency. If the lens remains the same and we improve the sensor to 0.9 the MTF now becomes 0.72. That could be visible, though whether or not it matters is subjective.

My example is grossly simplified because final MTF is actually a product of many components including sensor, lens, Bayer processing, postprocessing operations, monitor, printer, etc. The principle remains the same and improving resolution in any one component is always a good thing. 'Course, teeny-tiny pixels have disadvantages, and that is a balancing act that each of us must perform.