The GetDPI Photography Forum

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

Capture Integration IQ180 + Schneider Lens Tests

dougpeterson

Workshop Member
My personal apologies for taking so long to post this. Things have been extremely busy around Capture Integration the last two months.
----------------

Capture Integration is pleased to provide the community with downloadable files taken with the IQ180 with the following tech-camera lenses:
- Schneider 28mm Digitar XL
- Schneider 35mm Digitar XL
- Schneider 43mm Digitar XL
- Schneider 60mm Digitar XL
- Rodenstock 70mm HR

The tests are presented as 6-image stitches of around 280mp showing the entire possible range of movement for a Cambo Wide RS. Superimposed on the images is a grid showing the size of a single frame and the increments of 10mm of rise/fall/shift. This will allow viewers to evaluate what range of movements they will be happy with for each lens.

Questions, comments, violent dissent, personal evaluations, and completely off-topic poetic meanderings all strongly encouraged.

Link: Phase One and Schneider Lens Tests

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off
 

dougpeterson

Workshop Member
For anyone new to tech cameras. "Stitching" with a tech camera is different than the pan-and-stitch technique most dSLR users are used to. On a tech camera the lens is stationary and you use move the digital back to capture different parts of the continuous image circle.

This means you don't have to geometrically distort the images together (so pixels aren't being stretched/compressed), you don't have to crop any of the resulting image.

It also greatly increases the effective sensor size which changes how wide any given focal length is, and greatly increases the resolution of the final image.

By stitching a few IQ180 frames you can easily capture 200-400 megapixel images. Getting similar detail from a dSLR would take dozens of overlapping frames and many hours in post-processing software.

We'll be teaching, and using, this technique (from capture to final image) at our NE Landscape (Peak Color) in October.

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off
 

Shashin

Well-known member
and greatly increases the resolution of the final image.
Depending on the projection of the pan-stitch image, you may actually be getting more out of that method. There is actually no clear cut winner here as there are too many variables. At least with pan/stitch, I can use the center of my lens rather than the extremes of the image circle where resolution decreases.
 

dougpeterson

Workshop Member
Depending on the projection of the pan-stitch image, you may actually be getting more out of that method. There is actually no clear cut winner here as there are too many variables. At least with pan/stitch, I can use the center of my lens rather than the extremes of the image circle where resolution decreases.
Agreed. But you probably underestimate the number of lens/back combos that allow generous stitching. As an example download the 60XL example file with the IQ180 and examine the area outside of the image circle.

I completely agree that each lens has an area outside of it's center where loss of sharpness is meaningful. But for many lens/back combos and many stitches you will not enter unsharp areas of the lens circle. This test is, in fact, exactly the sort of test you should run with your lens to decide how far you can go with a flat-stitch with your particular lens/back.

Basically if you can stay inside the great part of your image circle flat-stitching is a clear winner (workflow/quality/ability-to-actually-compose). If the end-goal would require you to leave the great part of your image circle pan-and-stitch becomes the clear winner.

And of course for anything like 180 degrees or wider you can only do pan-and-stitch.

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off
 

Shashin

Well-known member
You lose me in the comparison of the methods. There are certainly benefits to moving the back in the image circle, but your conclusion that it is better just does not muster. There is no level playing field here. Stitch/pan is not limited to the number of frames regardless of the angle of view. As a photographer, I will do what I need to do.

BTW, I can compose in a stitch/pan just as well as moving a ground glass around. I need a nodal point to stitch/pan, so a bit more there. Work flow for both are simple. Your argument on quality is not proven.

Look, I don't want to get into an argument. I just wish you stopped at the benefits of working in the image circle. The comparison cannot be made to stitch/pan as giving less quality as there simply too many variables.
 

rhsu

New member
Capture Integration is pleased to provide the community with downloadable files taken with the IQ180 with the following tech-camera lenses:
- Schneider 28mm Digitar XL
- Schneider 35mm Digitar XL
- Schneider 43mm Digitar XL
- Schneider 60mm Digitar XL
- Rodenstock 70mm HR
[/SIZE]
wow... that's brilliant... good work!:)
 

Christopher

Active member
You lose me in the comparison of the methods. There are certainly benefits to moving the back in the image circle, but your conclusion that it is better just does not muster. There is no level playing field here. Stitch/pan is not limited to the number of frames regardless of the angle of view. As a photographer, I will do what I need to do.

BTW, I can compose in a stitch/pan just as well as moving a ground glass around. I need a nodal point to stitch/pan, so a bit more there. Work flow for both are simple. Your argument on quality is not proven.

Look, I don't want to get into an argument. I just wish you stopped at the benefits of working in the image circle. The comparison cannot be made to stitch/pan as giving less quality as there simply too many variables.
I all depends. There are moments when flat stitching is much better than any type of panoramic movment. If you have details in the foreground and use a wide lens it works much better and you don't have the problem of mismatching details.

However if I do something like a 3:1 or longer panoramic I use a normal stitching with lens movements, its just much faster.

I thionk it is also important to point out that it can be a hassle trying to do flat stitching on sunrise or sunset pictures, because you should "always" shoot a LCC right after.
 

rhsu

New member
flat stitching with LCC is not as simply as single shot LCC application - esp what Christopher has just mentioned - due to lumination. I have worked with P45+ via C1 and 75VL via Sinar Capture and both needs personal skills to make it work nicely - as Guy mentioned somewhere about LCC stitching application.
 

Stefan Steib

Active member
Hi Doug

thanks for posting these samples. I took a deep look inside, with Photoshop CS5 and channels in BW at 200 % enlargement.
I have to say I was expecting this, but I´m astonished about the degree of degradation. LCC does a good job on the visual file but if you look at the channels (especially the blue one) I would say that if someone needs a high quality file you can forget the 28mm and 35mm lenses for movements. The noise starts to get apparently more even inside the original frame border of the 28mm and only some mm outside of the frame of the 35mm lens.
The effect is less apparent in the highlights and midtones, but (as to be expected) the dark tones are suffering mostly.
The 43mm is “OK” but still there are not much movements (maybe 10mm?) before the noise hits.
The 60mm is nice you are right, it is a question of the lightbeam hitting the chip, the angle is essential. This is doing more harm than any Retrofocus can ever do.
I guess that at the 28mm effective resolution will loose up to 30% maybe more, it starts even inside the unmoved frame , the 35mm is ok within the unmoved frame but starts to loose the same amount of resolution only some mm more outside.
This can be improved with a Centerfilter (physically ! not Electronically - again Schneider errs completely!).

I would be very curious how the Rodenstock23+ 32 would look like, I guess somewhere between the 35mm and 43mm - respectively 47-60 (tending upwards) which is also proofing it´s got nothing to do with absolute sharpness but with the lens to chip distance MOSTLY !

Now what does this tell about Superwideangles on the new 80MPix chips ?
I guess there is only one way to get usable pictures- the answer is retrofocus (see Rodenstock...;-) and - with all respect - for sure our HCam-B1.
We simply don´t have any Color cast, even with 17mm.
And with 24mm we have more movements than Schneider with 35mm. And compared to Rodenstock the Canon 24mm costs 2000 € - the 23mm Rodenstock costs over 6000 € !

Greetings from Munich
Stefan
 

Stefan Steib

Active member
Christopher

I have seen this and also posted in that thread, what I meant was a similar circular stitch and LCC applied as well to get a visual correct file.

Regards
Stefan
 

tashley

Subscriber Member
Doug,

Thanks for posting these examples: I will have a really good dig around in them. As you know, I have the 35XL and saw, as a result of your kindness, some of your files relating to that lens about a week ago before I took delivery of my IQ 180. This sort of testing is absolutely invaluable to the user community and I thank you for all the hard work that goes into it. It makes hard and complex decisions so much easier for people to make in a world where various combinations of very high-end gear are relatively rare and where user feedback and testing is therefore pretty hard to come by.

It would be really interesting if you have time to hear your own opinions about what kit you'd recommend for use with the 180 for tech camera use?

Best
Tim
 

cly

Member
Doug,

thanks a lot for these tests!

There is something going on in the Schneider 43mm files which I don't understand: What is the cause of this vertical 'color boundary' shown in the attached file which is a crop from your stitched image? It's a problem I have noticed myself when shifting the 43 but so far I haven't had time for proper testing, i.e., eliminating variables.

The lens is in a fixed position, you move the sensor within the image circle - but it seems that the image projected onto the center of the image circle is rendered differently if this rendering takes place somewhere close to the edge of the sensor or if it takes place in the center of the sensor. From a technical point of view, this doesn't make sense to me. I don't know if flare plays a role in this as the pattern of reflections caused by the sensor glass is definitely different when the sensor is shifted, and this could result in a seemingly different impression.

Or is there a trivial explanation for this line in the crop? Is it a C1 problem?

Chris
 

gazwas

Active member
I've found with my 43XL that the LCC Tech Camera correction in C1 sometimes shifts the white balance a bit between shots, especially when it does more extreme corrections. The above could be the result of that?
 

dougpeterson

Workshop Member
The sensor has multiple read-out points. Under normal situations the back calibrates the reading of those read-out points so perfectly that it's not visible even under a lot of post-processing punishment. This is true of most cameras I have tested (including dSLRs like Canon) and some special tests (completely impractical to photographic applications) can show this in even in very "straight" shots.

In this case the captures that take place at the outside of the image circles of lenses which Phase One does not recommend (but which I think the tests show are usable provided you understand the limitations of movement that are implied) seem to present a problem with that calibration resulting in these artifacts. I could not explain the engineering reasons beyond the fact they're likely related to the same phenomenon as the LCC.

Bottom line the color purity, sharpness, light-falloff, and any such artifacts must be considered in evaluating how much movement you're comfortable with for this combination of back/lens, and in making any decisions about which lenses you wish to couple with this back.

I would expect some further modest improvements to the LCC algorithms, just as we have seen some significant improvements already in the addition of the "Wide Angle Tech Lens" version of the algorithm vs. the standard algorithm. This might expand the usable image circle several more mm - only time will tell and if you're purchasing I'd do so based on what is possible today (illustrated by this test and your own testing) and depending on your needs swap out the wider of the Schneider XL's for Rodenstocks (or possibly a Hartblei solution - we're arranging to test one ASAP) where possible, practical, and desired.

The IQ180 is absolutely fantastic - I'm obviously a biased source of info on the topic, but hopefully have built up creditability here by openly stating when I think things are not great. However, all tools have their advantages and disadvantages and large movements with the 28XL/35XL/43XL is not one of the IQ180's advantages. Each of those lenses is great lenses for this back, but each within a range of movements. A range you can determine pretty easily (with your own criterium/needs) based on this test. That's the whole point of the test :).

Doug Peterson (e-mail Me)
__________________

Head of Technical Services, Capture Integration
Phase One Partner of the Year
Leaf, Leica, Cambo, Arca Swiss, Canon, Apple, Profoto, Broncolor, Eizo & More

National: 877.217.9870 *| *Cell: 740.707.2183
Newsletter | RSS Feed
Buy Capture One 6 at 10% off
 

cly

Member
Bottom line the color purity, sharpness, light-falloff, and any such artifacts must be considered in evaluating how much movement you're comfortable with for this combination of back/lens, and in making any decisions about which lenses you wish to couple with this back.

[...] However, all tools have their advantages and disadvantages and large movements with the 28XL/35XL/43XL is not one of the IQ180's advantages.
Just for the record: I have noticed this color problem when shifting with the 43mm on a P40+. So it's not a problem which is characteristic of the IQ180.

Chris
 

chrismuc

Member
Hi Doug

thx for the test. The result is very interesting.
Especially that the color cast is neither round symmetric to the center nor it shows horizontally and vertically the same symmetry.
This would mean the square pixel of the iQ180 back behave different towards out of orthogonal angle arriving light in horizontal or vertical direction.
Any explanation for this?

Another request: Could you please do the same test with the iQ160 to see how different (hopefully better) the back performs with the same Schneider lenses and could you do the same test with the iQ160 and iQ180 back with the Rodenstock 28, 32 and 40 mm lenses in order to have a complete comparison how the two backs work with the two lens families.

And ... btw ... Stefan Steib could add the same shift + stich tests with the 60 and 80 MP backs on the Hartblei cam with the Canon TSE 17 and 24 mm and the Hartblei Zeiss 40 mm lenses:)

Then we would have a substantial overview of wide angle shift lens + back performances.

Thanks for your efforts,
Christoph
 
I'll add that I shoot with the Cambo wide RS and both the new 28mm Super Digitar and 60mm apo. I use the Leaf Aptus II 10 - as most of you will know 56mp - is a similar Dalsa sensor to the Phase IQ 160 that Guy has. I have had great results. I use the 28mm for about 50% of my architecture work and it works nicely with the LCC on capture one software. And yes, use shifts and rise with it. I have had no issues with my 60mb back so hopefully, there are still options for many shooters and we will avoid "the sky is falling threads." I can get away with 10-11mm of shift without losing color saturation after fixing with LCC. Any more than this like up to the 17mm limit, and the edge of the frame loses most of your color saturation and looks somewhat monochromatic. Green leaves look brown. Blue sky looks like you desaturated 30 points. Yellow painted interiors go to neutral beiges. So hopefully you can know this ahead of time and compose your images to work with this if you need more movements. I find the sharpness of the lens like going from canon wide's to Leica M wide glass. It is like the same night and day from my zeiss 35mm lens on the contax to the schneider 28mm super Digitar. And the zeiss 35 is outstanding. So there you go. The 60mm is stellar no matter what you do to it.

I won't be shifting the 28mm to the extremes unless I am looking for a black and white image. Most of my work with the 28 falls in the zero to 10mm of rise anyway where it excels nicely (with my back.).

Hopefully in the next few weeks, guy and I can do an abbreviated but similar test and involve Doug with the 28mm on the phase iq160 to see if it has similar results to what I am seeing on my Leaf.

Robb Williamson
 
Top