Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: dkaufman on September 20, 2016, 12:32:51 pm
-
The news today that Phase One is discontinuing the 60 and 40 Megapixel sensor backs is very disconcerting. I shoot stitched images all the time on an Arca Swiss M-Line Two, formerly with a Phase One 40+ and now with an IQ160, and make full use of the image circles of both Rodenstock and some Schneider lenses. The 60 and 40 megapixel backs make full use of the image circles possible, the 50 and 80 and 100 Megapixel backs much less so, even with LCCs, according to all tests. These latest announcements make me think I should buy at least one spare 40 or 60 used back, as it appears there will no upgrade path for users who push the lens circle boundaries to the fullest. Do others share this view and think there is still a market for backs with larger photosites and more useable sensor architecture?
David Kaufman
www.davidkaufmanphotography.com
-
The news today that Phase One is discontinuing the 60 and 40 Megapixel sensor backs is very disconcerting. I shoot stitched images all the time on an Arca Swiss M-Line Two, formerly with a Phase One 40+ and now with an IQ160, and make full use of the image circles of both Rodenstock and some Schneider lenses. The 60 and 40 megapixel backs make full use of the image circles possible, the 50 and 80 and 100 Megapixel backs much less so, even with LCCs, according to all tests. These latest announcements make me think I should buy at least one spare 40 or 60 used back, as it appears there will no upgrade path for users who push the lens circle boundaries to the fullest. Do others share this view and think there is still a market for backs with larger photosites and more useable sensor architecture?
As we found in the tests that created our massive IQ3 Raw File Download Collection (https://digitaltransitions.com/iq3-100mp-100-megapixel-raw-file-download/) the IQ3 100mp is a great back to use on a tech camera with Rodenstock lenses. You lose a couple mm of the usable image circle of wider Rodenstock lenses - that is true, and is a shame. But you gain so much compared to the CCD backs: resolution, live view, long exposure, HDMI. We have many users now using it with tech cameras.
Most of the loss you speak of comes from the loss of compatibility with legacy Schneider wide-angle tech camera lenses. But Schneider doesn't make tech camera lenses anymore, so it's hard to imagine planning the future of Phase One's digital back line around them. A loss, for sure; I for one wish that Schneider had kept going with them.
-
The news today that Phase One is discontinuing the 60 and 40 Megapixel sensor backs is very disconcerting. I shoot stitched images all the time on an Arca Swiss M-Line Two, formerly with a Phase One 40+ and now with an IQ160, and make full use of the image circles of both Rodenstock and some Schneider lenses. The 60 and 40 megapixel backs make full use of the image circles possible, the 50 and 80 and 100 Megapixel backs much less so, even with LCCs, according to all tests. These latest announcements make me think I should buy at least one spare 40 or 60 used back, as it appears there will no upgrade path for users who push the lens circle boundaries to the fullest. Do others share this view and think there is still a market for backs with larger photosites and more useable sensor architecture?
Hi David,
I wonder if anybody ever tested a 40/60 MP back with shifts, versus a normal shot with a 100MP back, with software correction for perspective instead of lens shifts. Capture One since Version 9 has much improved resampling algorithms that maintain most resolution without introducing varying resampling blur. The added resolution from a 100 MP back also helps, and shooting with the center of the image circle also helps. There is also less need for drastic LCC corrections when shooting with a centered lens.
For extreme angles of view, stitching may be a better solution.
Maybe a new way of doing the same thing? Just a thought.
Cheers,
Bart
-
I use a P45 still for architecture. I have also used the IQ 60mp backs. I like both. My gear delivers. Yes, the latest would be nice, and certainly live view, etc... would be quite welcome. I feel the same way though...There are less and less great optionsfor architectural work that don't compromise in a pretty big way. On a cambo at least, the Schneider 35 gave 20mm shift. Yes, I do use that. I have often use it for stitching, but many times just to make a shot in a situation that can't be made the same otherwise. The rodenstock 35 can only shift 6mm on the same setup according to cambo's website.
Digital correction works in many situations...and can be just fine...but not always. It also means I have to guess what my final composition needs to be or actively engage corrections while shooting tethered. There are many situations digital correction can't fully fix perspective. You also can't hide your self from mirrors like you can with shifting. This comes in quite handy to a working architectural photographer.
I would be curious to see tests of the Rodenstock 40mm on a 100mm back. at full shift.
-
Hi David,
I wonder if anybody ever tested a 40/60 MP back with shifts, versus a normal shot with a 100MP back, with software correction for perspective instead of lens shifts. Capture One since Version 9 has much improved resampling algorithms that maintain most resolution without introducing varying resampling blur. The added resolution from a 100 MP back also helps, and shooting with the center of the image circle also helps. There is also less need for drastic LCC corrections when shooting with a centered lens.
For extreme angles of view, stitching may be a better solution.
Maybe a new way of doing the same thing? Just a thought.
Cheers,
Bart
This does not provide an exact replication of shifting. The shift performed in computer distorts the building to a degree more then shift in camera and can provide disappointing results.
Insofar as the high MP backs, I am not sold. I did look at images from my dealer and another photographer captured with a Rod 32 and 40 and found in both cases poor results with sift on the IQ3 100. Even with bright cloudy skies, where the amount of information recorded is significantly higher then darker objects, noise was prevalent.
This is the first I heard of this. I am now thinking more.
It also means I have to guess what my final composition needs to be or actively engage corrections while shooting tethered.
This is another annoying aspect of correcting in post. I would much rather compose my shot within my frame, not guess what that frame may be.
What drives me the most nuts about this is all of the bill shit P1 feeds us on how C1 will fix the lens cast without any issues. We, as architectural photographers, are probably the most technically astute group of shooters out there. This marketing mumbo-jumbo is such a insult to our intelligence.
I can only hope that the high acute angle response of mirrorless cameras will bring us sensors that once again work with tech camera wides.
-
Phase did not really abandon the tech camera, with the 50mp and 100mp backs, instead they did quite a bit of work on the LCC process to make both backs work with tech cameras.
Per crops below, all taken with the IQ100, 32mm Rodenstock, 12mm of shift, with CF on.
Before the LCC is applied, yes there is an apparent loss of both color fidelity and saturation, along with the standard magenta shift (note the magenta shift happens with CCD also). However also note that there is very little loss in overall details due to noise, and the shift will pull up quite well. Once the LCC is applied you have a very good match to the color from the center of the file. Is it exact, no, never will be, and for solid sky (blue) it's still a trick to get an even blend, but to my way of looking, no harder than working with the IQ260 or 160. As has already been mentioned, the gains IMO outweigh the losses.
Paul C
-
What drives me the most nuts about this is all of the bill shit P1 feeds us on how C1 will fix the lens cast without any issues. We, as architectural photographers, are probably the most technically astute group of shooters out there. This marketing mumbo-jumbo is such a insult to our intelligence.
Hi Joe, I routinely use RawTherapee's flat field correction, it works exceptionally well, give it a try. Here is the wiki page on various automation features and how it works: http://50.87.144.65/~rt/w/index.php?title=Flat_Field
-
I would be curious to see tests of the Rodenstock 40mm on a 100mm back. at full shift.
Full range of Phase One 100mp tech cam tests downloadable for free here: Phase 100mp Raw Download (https://digitaltransitions.com/iq3-100mp-100-megapixel-raw-file-download/)
-
In reply to Doug's implied point that a 100MP back can make up for loss of image circle, it's really a complex calculation. With Rodenstock's 40mm lens for example, if you have a stitched image that covers the full width, or nearly the full width, of the image circle with a 60MP back, using a 100 MP back would result in a loss of 10 mm of circle and would force you to go to a wider lens to cover the same subject, in this case a 32mm lens. When you calculate the results from the higher resolution back with the wider lens, the actual image coverage will actually be of slightly lower resolution (approx 5% lower linear number of pixels) than with the 60MP back. So that's not necessarily a good trade off, and a very expensive one. Admittedly with longer focal length lenses (70, 90, 120), this will be less of an issue and the extra pixels will compensate for the presumably lesser loss of image circle.
Of course, a 100 MP back is a very high end solution in dollar terms, offering technological advantages, but is not affordable for many photographers. At the present time, I think that
60 MP backs are definitely the sweet spot in backs, from a pixel resolution / cost perspective, for stitched images with technical or view cameras. 40MP backs, with a smaller sensor size, offer somewhat wider stitching possibilities but at somewhat lower image resolutions.
David Kaufman
www.davidkaufmanphotography.com
-
In reply to Doug's implied point that a 100MP back can make up for loss of image circle, it's really a complex calculation. With Rodenstock's 40mm lens for example, if you have a stitched image that covers the full width, or nearly the full width, of the image circle with a 60MP back, using a 100 MP back would result in a loss of 10 mm of circle and would force you to go to a wider lens to cover the same subject, in this case a 32mm lens. When you calculate the results from the higher resolution back with the wider lens, the actual image coverage will actually be of slightly lower resolution (approx 5% lower linear number of pixels) than with the 60MP back. So that's not necessarily a good trade off, and a very expensive one. Admittedly with longer focal length lenses (70, 90, 120), this will be less of an issue and the extra pixels will compensate for the presumably lesser loss of image circle.
Of course, a 100 MP back is a very high end solution in dollar terms, offering technological advantages, but is not affordable for many photographers. At the present time, I think that
60 MP backs are definitely the sweet spot in backs, from a pixel resolution / cost perspective, for stitched images with technical or view cameras. 40MP backs, with a smaller sensor size, offer somewhat wider stitching possibilities but at somewhat lower image resolutions.
David Kaufman
www.davidkaufmanphotography.com
David, I agree with some of your points on this. At some point, though, manufacturers move on from a product, and the 40/60 Dalsa sensors harken back to 2009. I can't think of a digital equivalent where the same sensor was available in the (almost) original chassis as a new product that one could order so many years after the original launch date. At any rate, there will be Phase One pre-owned IQ160 models available for quite a few more years. We still commonly have P40+/P65+ models in our pre-owned inventory, years after they were discontinued. It's not all bad, David.
Steve Hendrix/CI
-
In reply to Doug's implied point that a 100MP back can make up for loss of image circle, it's really a complex calculation. With Rodenstock's 40mm lens for example, if you have a stitched image that covers the full width, or nearly the full width, of the image circle with a 60MP back, using a 100 MP back would result in a loss of 10 mm of circle and would force you to go to a wider lens to cover the same subject, in this case a 32mm lens. When you calculate the results from the higher resolution back with the wider lens, the actual image coverage will actually be of slightly lower resolution (approx 5% lower linear number of pixels) than with the 60MP back. So that's not necessarily a good trade off, and a very expensive one. Admittedly with longer focal length lenses (70, 90, 120), this will be less of an issue and the extra pixels will compensate for the presumably lesser loss of image circle.
Of course, a 100 MP back is a very high end solution in dollar terms, offering technological advantages, but is not affordable for many photographers. At the present time, I think that
60 MP backs are definitely the sweet spot in backs, from a pixel resolution / cost perspective, for stitched images with technical or view cameras. 40MP backs, with a smaller sensor size, offer somewhat wider stitching possibilities but at somewhat lower image resolutions.
David Kaufman
www.davidkaufmanphotography.com
I just don't see the loss as that much.
The 40mm HR-W max shift on a full frame IQ sensor is 16mm and 16 mm is still pushing the IC indicator of the Rodenstock which creates a hard vignette and ruins the edge.
With the 260/160 I was able to get 16mm with the IQ100 15mm.
Each back has trouble with extreme shifts and from my work the CCD backs can be harder at times to correct. Each back is relying heavily on the LCC to recover the color and saturation on the shifted images.
16mm vs 15mm is not that much.
Plus on the 50MP back due to the crop you can shift a bit further out as the crop keeps you from hitting the IC indicator and image damage until around 18mm.
Rise fall perform better than shift also. But I rarely use them.
Paul C
-
Recent image from a landscaping client with the IQ3100 and 40hr. Two-shot stitch each shifted 18mm L/R. That's a 98.4mm IC. As you can see from the corrected LCC, it is indeed beyond the claimed 90mm IC for the 40hr. Second image is the uncorrected right-shifted image (back shifted right 18mm). Third image is the combined pano, cropped on the L/R to ~ the lens IC (you can still see the dark vignette in the upper right). This image will undoubtedly be cropped further on the right, but for y'all I thought I would show the as-delivered image.
Sure, if this was one of Joe's white kitchens (ha - see what I did there?) there might be more post required for color correction. But frankly the lens gets soft out at the edge of its IC anyway.
Has P1 abandoned the technical/architectural shooter? I don't know, but I agree with Paul; the end results I'm getting are better than my IQ180 was, and darn close to the P65+ before that. I'll just keep using the tools while the debate goes on.
In this case the client is happy because they got exactly what they wanted. The hardest part of this job was finding a clear sunny day in Northeast Ohio!
Dave
-
Not enough attention has been given to just how much work is being done by the LCC and C1, and Phase One does deserve credit for this. Dave's example shows the hard vignette due to the IC (image circle indicator) put inside the lens by Rodenstock which is an unfortunate design element IMO.
Due to slow shipment rate, and early problems with the first LCC corrections, the whole issue of color cast correction got a bit blown out of proportion.
I realize if you dig into the files and break them out pixel by pixel with software tools, faults will be found, however:
1. A lot of the same faults show up with CCD backs, or there are other problems.
2. Will the issues show up in print? no at least not for me at 38 x 96 several times.
3. Will the issues show up in image sized for web output No.
There exceptions, I guess a pure white wall is not a good candidate, but from my experience a pure white wall did not work well with a 160 or 260 with shifts over 10mm with same lenses.
A pure blue sky has always been problematic, but I had just as many issues with the 160/260, possibly more, blending in a 3 part stitch and pure blue sky.
Paul C
-
Paul,
Good point about recent improvements in the LCC process. I am sure my comparisons to the P65+ and to some extent the IQ180 are influenced by exactly that.
Dave
-
I can also attest to the remarkable improvements in C1 regarding the LCC process. I shoot with the Schneider 35XL which is long known to have serious cast issues. With the IQ3100 I am able to shift a full 10mm - I know its not 15mm but 10mm is a miracle and more than I will ever need. Its just remarkable how well the color casting is completely neutralized.
Victor
-
Really quite remarkable. On another Forum, there is a shot of a 23mm shifted 12mm, which is also just lovely. Makes one rethink. They've been busy in Denmark.
-
Yes Geoff it is. I want to respond to David's point in his posts and also to Tony's post: At least for the 40hr, I don't see the IQ3100 limiting the use of the image circle. I can't imagine anyone using more of the 40's image circle since the corners go black beyond what I've done in the example above. As Paul points out, this is because of the mechanical stop in the lens and has nothing to do with the back.
I know there are some tests floating around. As usual, photographers should do their own tests before drawing conclusions and making decisions. I tested all my lenses using both my IQ180 and a demo IQ3100 before I made the swap (40hr, 60xl, 90hr-sw, sk150). Did the same in 2011 when going from the P65+ to the IQ180. Obviously other lenses could have different results, as might specific subject matter.
Ciao,
Dave
The 60 and 40 megapixel backs make full use of the image circles possible, the 50 and 80 and 100 Megapixel backs much less so, even with LCCs, according to all tests.
-
Dave, I believe there is an error in your post about the 40HR. Two shifts of 18mm plus the sensor size of 54 mm equal 90mm, not close to a hundred as you write. Furthermore, tests by the major US sellers of these backs and lenses conclude there is reduced image circle usability with the 80 and 100 MP backs. I'm not trying to get into an argument over this. I'm just saying that for stitching applications , the new backs are less flexible with wide angle lenses, which is a concern to me, and that extra pixels don't always make up for the loss of image circle when you are forced to go to a wider lens. I understand there are trade-offs with new technology, but for me, I don't see an upgrade path at this time from the 60MP sensors. It may not matter to many other people.
David Kaufman
-
Dave, I believe there is an error in your post about the 40HR. Two shifts of 18mm plus the sensor size of 54 mm equal 90mm, not close to a hundred as you write. Furthermore, tests by the major US sellers of these backs and lenses conclude there is reduced image circle usability with the 80 and 100 MP backs. I'm not trying to get into an argument over this. I'm just saying that for stitching applications , the new backs are less flexible with wide angle lenses, which is a concern to me, and that extra pixels don't always make up for the loss of image circle when you are forced to go to a wider lens. I understand there are trade-offs with new technology, but for me, I don't see an upgrade path at this time from the 60MP sensors. It may not matter to many other people.
Have you personally tested or otherwise meaningfully shot with the 100mp?
If not, I might suggest, based on reading your concerns, and working with a large number of tech cameras users on exactly this transition, that you do. I don't know where you're located – if in the US I'd be glad to help arrange it out of our NYC or LA offices.
-
I've been meaning to do a more in depth post about the state of camera platforms in relevance to architectural photography. While I don't believe Phase One or any other manufacturers have intentionally abandoned the genre, I do feel like the natural evolution of the industry has left us arch shooters somewhat handicapped.
The backs over 60mp do indeed have less usable movement when shifting for composition. I found the quality from my SK 35 so unusable when shifting over 10mm that I frequently dropped down to the DSLR and 24mm TS-e. Also, 35mm is just not that wide when you're shooting architecture.
If you really consider the degree of shift available on any 35mm lens with a MFDB, and I'll be generous here... Let's say you can get 12mm on an IQ 260 (the last back I owned). That represents about a 22% shift for a horizontal composition (54mm wide sensor). On a 35mm sensor you can get 13mm shift from the 25mm TS-e (and I can get 15mm when mounting it on the Arca) that represents a 36% image shift.
So, by using the smaller system, I can compose images that would be impossible to obtain on MF. Add to that the fact that the 24TS-e on a little sensor performs better at the edges than my 35mm SK did on the DB.
On top of that, there is the 17mm TS-e which will do things you could never dream of on MF.
All in all, I think the state of things has developed almost randomly. MFDB as a capture platform is much more attractive now than it was for the 10 years I used it... just not for the way I compose architecture. The little camera, though it certainly produces lesser quality files, is just a more capable platform for me.
I've shot the IQ3 100 and I love the image quality. I really wanted to try to find a way to make it work. I just don't see it happening.
#frustratedarchshooter
-
Dave, I believe there is an error in your post about the 40HR. Two shifts of 18mm plus the sensor size of 54 mm equal 90mm, not close to a hundred as you write. Furthermore, tests by the major US sellers of these backs and lenses conclude there is reduced image circle usability with the 80 and 100 MP backs. I'm not trying to get into an argument over this. I'm just saying that for stitching applications , the new backs are less flexible with wide angle lenses, which is a concern to me, and that extra pixels don't always make up for the loss of image circle when you are forced to go to a wider lens. I understand there are trade-offs with new technology, but for me, I don't see an upgrade path at this time from the 60MP sensors. It may not matter to many other people.
David Kaufman
It mattered to me, for sure, but everyone has their standards. The press on the IQ100 has been very limited to early adopters, many of who were using a incorrect version of C1. The CCD backs you reference really are not better, they may be equal, but not better. As I have tried to point out, each has their own issues. But the CMOS backs can work well and very well on tech cameras. Strongly recommend working with a dealer, renting, etc and trying for yourself.
Phase is partly to blame for this also, as they have not been very good on getting the positive message out, which surprises me especially with all the work they have done on the CMOS LCC.
But, one thing is for sure, a CCD back IQ260/160 whatever, will have the exact same limits on the HR-W 40mm, due to the internal IC indicator. That is in the lens, Dave's image shows this very clearly. Depending your image, this may have adverse effects/non recoverable back to even 14mm, which is why I always stop at 15mm with the 40 HR-W.
Paul C
-
David,
I was quoting the image circle, i.e. the diagonal, not the horizontal image size. SQRT ((53.7+36)^2 + (40.4)^2) = 98.38mm.
No problem with you disagreeing. If we all felt the same way about everything and saw everything the same way it would be a boring world. I'm not trying to say going to an 80 or 100 mp back has zero downsides. I'm sure there are lenses and situations where the 60mp back would be the better choice. I get that. I'm just quite convinced the 40hr is not one of those, and that is from my own personal testing. My whole point in posting was because the 40hr was specifically called out in your post and in Tony's. I don't think these limitations exist when referring to to the 40hr. In my opinion, this is one of those cases like when we were kids and the message gets whispered around the circle. By the time it comes back the story is quite different, and usually much worse than it really is.
Actually, the site where this photo was taken drops down about 8 feet right beyond where I had the camera set up. Believe it or not, I was glad I had the 3100 and not my 180 because I would not have trusted the LCC's ability to clean up an IQ180 image as well as the IQ3100. If I still had the 180 I would have wanted the camera back another foot or so. I would have needed some complicated ladder / scaffolding to reach the camera.
Dave
Dave, I believe there is an error in your post about the 40HR. Two shifts of 18mm plus the sensor size of 54 mm equal 90mm, not close to a hundred as you write. Furthermore, tests by the major US sellers of these backs and lenses conclude there is reduced image circle usability with the 80 and 100 MP backs. I'm not trying to get into an argument over this. I'm just saying that for stitching applications , the new backs are less flexible with wide angle lenses, which is a concern to me, and that extra pixels don't always make up for the loss of image circle when you are forced to go to a wider lens. I understand there are trade-offs with new technology, but for me, I don't see an upgrade path at this time from the 60MP sensors. It may not matter to many other people.
David Kaufman
-
Evidently I missed an announcement. The 40mp CCD had to die, has been dead, and really will have an amazing live in the used back market. Thousands were made, and folks have been upgrading from them to the 50/60/80/100mp backs. They won't die, and yes, until recently they were the entry back for both Phase and Hass. 50mp CMOS really has set itself up as the defacto starting point. Sony is making lots of money off them, and making lots of them since Phase/Hass/Pentax are all using it.
The X1D tells us there will be more, higher resolution 33x44 chips, sooner than later. With that info, to discontinue the 60mp CCD actually helps clear up pricing for folks - it makes the jump to the 100mp seem higher, but allows for a 75mp CMOS to slide in next year and be a middle ground with all the advantages of CMOS.
Hasselblad dropped their H5D-60 product, and my pre-coffee brain says that Dalsa most likely hasn't done a run of the 60mp chips in a while. It sucks, but Intel doesn't make Pentium III processors any more either (at least for retail sales). It's much better for Phase/Hass to be up front and say that yes, these cameras & backs work great, but we need to push forward.
I say this all with a touch of sadness as my first hit of MF was with an IQ260 on the World Tour.
-Joe
-
Hi David,
I wonder if anybody ever tested a 40/60 MP back with shifts, versus a normal shot with a 100MP back, with software correction for perspective instead of lens shifts. Capture One since Version 9 has much improved resampling algorithms that maintain most resolution without introducing varying resampling blur. The added resolution from a 100 MP back also helps, and shooting with the center of the image circle also helps. There is also less need for drastic LCC corrections when shooting with a centered lens.
For extreme angles of view, stitching may be a better solution.
Maybe a new way of doing the same thing? Just a thought.
Cheers,
Bart
There is a small comparison between tech cam and software correction in this Getdpi thread (http://www.getdpi.com/forum/medium-format-systems-and-digital-backs/59096-technical-camera-obsolete-keystone-correction-vs-shift.html) both with 100mp back. IQ is quite similar. I'm sure 40mp would be worse.
-
Sure, if this was one of Joe's white kitchens (ha - see what I did there?)
Dave
;D
If you only knew Dave, how many times I've heard that same joke. ::)
-
So, by my understanding, if they brought out at 50 or 60 CMOS that was full frame, due to it being 'less dense' it would work better on our tech cams?
Am I correct in this?
Given there were two full sized chips in the CCD range, is there any reason there couldn't be two full CMOS chips?
Other than us APs, would there be a market for a chip like this?
Do you think Phase has a Tech cam of any sort in their labs to test these things?
-
;D
If you only knew Dave, how many times I've heard that same joke. ::)
I can only imagine. Probably about as many times as I get, "Huh. I thought you were Chinese" when people meet me.
::)
Edit: Now that I think about it, maybe I should do food photography.
???
-
So, by my understanding, if they brought out at 50 or 60 CMOS that was full frame, due to it being 'less dense' it would work better on our tech cams?
Am I correct in this?
Given there were two full sized chips in the CCD range, is there any reason there couldn't be two full CMOS chips?
Other than us APs, would there be a market for a chip like this?
Do you think Phase has a Tech cam of any sort in their labs to test these things?
For the most part, that would be correct. A lower MP sensor will handle lens cast better, however there are other aspects that effect this as well.
For instance, the type of and curvature of the micro lenses used and how deep the pixels wells are, plus others. These aspects change from manufacturer to manufacturer; for instance Kodak's chips have a great reputation for handling lens cast. If you compare the 50 MP Hassy to the 60 MP Phases, both sets have the same pixel pitch, but the Hassy version (with a Kodak chip) still handles lens cast much better.
In the short run, I can't see any development specifically planned to help out us APs, and I doubt Sony has created a 60 MP chip to sub into the 60 MP backs.
However, with mirrorless cameras being the next wave, high acute angle response (essentially engineering talk for handling lens cast well) in newer chips will increase since lenses on mirrorless cameras can sit just as close to the sensor as in tech cameras.
This may wind up bringing sensors once again back to the market that will work well with our tech cameras. If this happens, I would say it is at least a couple of years out.
-
Thanks Joe,
Lucky I like using my 60 then...It's definitely got a number of years left in it!
;-)
-
As I could never have the money to get into DMF, I learned to love profusely my EOS-TSE system.
Still shooting with an EOS body while salivating each new Sony, but not yet. The GFX is immensely atractive but unless is below $5.5 K I don't think I'll jump. By the time the GFX is selling, Sony already announced and soon to sell another amazing FF camera.
Eduardo
I've been meaning to do a more in depth post about the state of camera platforms in relevance to architectural photography. While I don't believe Phase One or any other manufacturers have intentionally abandoned the genre, I do feel like the natural evolution of the industry has left us arch shooters somewhat handicapped.
The backs over 60mp do indeed have less usable movement when shifting for composition. I found the quality from my SK 35 so unusable when shifting over 10mm that I frequently dropped down to the DSLR and 24mm TS-e. Also, 35mm is just not that wide when you're shooting architecture.
If you really consider the degree of shift available on any 35mm lens with a MFDB, and I'll be generous here... Let's say you can get 12mm on an IQ 260 (the last back I owned). That represents about a 22% shift for a horizontal composition (54mm wide sensor). On a 35mm sensor you can get 13mm shift from the 25mm TS-e (and I can get 15mm when mounting it on the Arca) that represents a 36% image shift.
So, by using the smaller system, I can compose images that would be impossible to obtain on MF. Add to that the fact that the 24TS-e on a little sensor performs better at the edges than my 35mm SK did on the DB.
On top of that, there is the 17mm TS-e which will do things you could never dream of on MF.
All in all, I think the state of things has developed almost randomly. MFDB as a capture platform is much more attractive now than it was for the 10 years I used it... just not for the way I compose architecture. The little camera, though it certainly produces lesser quality files, is just a more capable platform for me.
I've shot the IQ3 100 and I love the image quality. I really wanted to try to find a way to make it work. I just don't see it happening.
#frustratedarchshooter
The OP asks about architectural photography. Chris Barrett response nails the issues for professional architectural photographers.
The majority of us who earn our living in this field have moved on.
Paul
-
Chris - I have an A7r and sometimes find the look a bit harsh compared to my P45+ and P25.
Shooting efficiency aside, do you ever miss the look of those big pixels?
-
It seems some manufactures may be looking to outside sources for the tech toys for these new backs. Seeing what Alpa is about to release? Maybe I'm wrong but now a days once the camera is released, the manufactures money and R&D likely goes solely towards the next generations of product. This leaves lots of further advancements, software upgrades and brand specific accessories to third party vendors.
Sony makes the chip, someones else probably makes the lenses, maybe the brand makes the body but for sure lots of development with third party software vendors to put this complicated chip and software together in a marketable package before its outdated is maybe the only way to survive in these times.
-
I think its important to test the equipment in your own work flow environment. I was considering upgrading my IQ260 to IQ3 100 and took (see attached screen shot) some test shots near the Phase One office in Hong Kong. All images are taken with the Rodenstock 32HR (no CF) with a vertical shift of 15 at f11, no work was done on them. Top row is the IQ100 with its respective LCC card, bottom row same shot with my IQ260 with its respective LCC.
My conclusion is that this back got less shift than the IQ260 and thats exactly the point of the original poster. The new back is IMO a trade off between remarkable quality and excellent life view versus limitations on movement.
Grischa
-
An excellent comparison. It would be nice to see how the IQ100 file and corresponding LCC processes out in C1 10. To see how much of the red cast remains after the LCC has been applied.
Paul Caldwell
-
Chris - I have an A7r and sometimes find the look a bit harsh compared to my P45+ and P25.
Shooting efficiency aside, do you ever miss the look of those big pixels?
Andy, while I did find the IQ3 100 to make nicer pictures than the Sony, big pixels really don't come into play. Both sensors have very similar pixel pitch. I've been doing a lot of wide stuff lately with the 11-24 and the 17 TS-e. These are working remarkably well on the Arca/Sony combo and I just couldn't achieve the same field of view with the tech cam setup.
-
Chris - maybe I have to change my processing habits for the Sony? Reduce sharpening, contrast etc. I used the word harsh before and that's the way I find it to be, compared to the more subtle renditions of the P45+
Any tips?
-
Hmm, in the camera I have the profile set to neutral. Sharpening, contrast and saturation are all zeroed out. In C1 Pro, I often do something like +12 Sat, +20HDR both shadow and hilight, +20 clarity. I feel like the phase 1 files are more punchy and colorful, so I add saturation and clarity to the Sony files. If you go through the front page of my website, the shots are mixed about half and half A7r / IQ260. I don't think you'd pick any out as harsh.
IMO,
CB
-
I've heard Capture One 10 has a greatly improved LCC algorithm for the IQ3100, reducing ripple and residual cast artifacts. Has anyone tested it first hand?
I'm most interested in if they manage to keep color fidelity and fine tonality in the corrected areas, which would require some sort of crosstalk cancellation. As a software developer I worked with that myself a few years ago but never got it to fully work. I imagine the extremely low noise Sony sensors would improve the odds to succeed though.
-
I've heard Capture One 10 has a greatly improved LCC algorithm for the IQ3100, reducing ripple and residual cast artifacts. Has anyone tested it first hand?
I'm most interested in if they manage to keep color fidelity and fine tonality in the corrected areas, which would require some sort of crosstalk cancellation. As a software developer I worked with that myself a few years ago but never got it to fully work. I imagine the extremely low noise Sony sensors would improve the odds to succeed though.
As someone who first noticed this with the IQ150 I am very familiar with the issue. Phase One made significant improvement with C10 for the IQ100. The rippling is gone and color cast correction improved also. The fix is enough to allow a shift across a solid blue sky at 12mm of shift with the 32mm and 15 mm of shift with the 40mm Rodenstocks.
Can it address problems seen by the OP or other architectural shooters? Don't know. But for landscapes the fix is impressive.
Capture Integration showed some testing in their blog.
Paul Caldwell
-
I've heard Capture One 10 has a greatly improved LCC algorithm for the IQ3100, reducing ripple and residual cast artifacts. Has anyone tested it first hand?
I'm most interested in if they manage to keep color fidelity and fine tonality in the corrected areas, which would require some sort of crosstalk cancellation. As a software developer I worked with that myself a few years ago but never got it to fully work. I imagine the extremely low noise Sony sensors would improve the odds to succeed though.
Marked improvement. LCC generation itself is also much faster.
You can use our IQ3 100mp Raw File Download Catalog (https://digitaltransitions.com/iq3-100mp-100-megapixel-raw-file-download/) along with the free-for-use-with-phase-one-backs Capture One (https://www.phaseone.com/en/Download.aspx) to test for yourself.
Combine this with the sensor based electronic shutter (https://digitaltransitions.com/iq3-100mp-electronic-shutter/) feature added recently and it's very clear this topic is misnamed. Phase One is doing far more for the tech camera market than any other maker.
-
Doug, do you think they will get ES working with faster flash sync? For interiors, I'm typically between 1s and 1/8s and am always using strobe.
-
Marked improvement. LCC generation itself is also much faster.
I know the answer is "no comment on future releases", but here's hoping the LCC improvements are brought to the 50MP chips as well.
-
Combine this with the sensor based electronic shutter (https://digitaltransitions.com/iq3-100mp-electronic-shutter/) feature added recently and it's very clear this topic is misnamed. Phase One is doing far more for the tech camera market than any other maker.
Here here! Now how about that split live view?
;D
Dave
-
I agree it's probably not fair to complain to Phase One about the tech cam support. They seem to do what they can, the LCC corrections they do today is already pushing the limit of what's really possible. I'm impressed. It's mainly the sensor technology that let tech cams down, and that sensor tech is just a response of what the majority needs.
I also think as resolving power is virtually the only reason most people are using tech cams (or put it this way -- as soon as the resolving power advantage goes away, the users tend to go away too), it's just a dinosaur waiting to fade away as the traditional camera systems improve. Few are willing to mess around with less than fool-proof LCCs when you can get almost as good quality with an ultra-wide and some cropping and keystoning.
I have myself more or less come to peace with that I'm probably owning the last tech cam I'll ever own, and I plan to use it for many more years as I like it as a creative tool. I just hope the digital back holds up for some serious age and Linhof stays in business for service and repairs.