A7r on A/S M-Line Two?

[chiek]

All articles are very interested. Thanks for many opinion for me.
I'm thinking about our own sinar systems for many studio photographers to breath new life on a7/a7r like this articles.
Actually if studio setup, camera size does not matter.

[Rod.Klukas]

Posts: 95


View Profile WWW Personal Message (Online)
   
   
Re: making compact technical camera for A7/A7r
« Reply #37 on: Today at 12:18:58 PM »

Arca-Swiss has just released the bellows for their DSLR2.  And we have redesigned the mounting to make it fast and easy to use.
The mount will allow the use of wide angle lenses down to at least 35mm and we are testing to see how short we can go with movement.  It appears the 23mm and 28mm Rodenstocks are out. The rear of the lenses, especially on the Rodenstock's is so long that it contacts the body of Nikon's and Canon's below 70mm and even on the Sony A7 may not work.
The Sony A7 series, being mirror-less, allow much shorter lenses to be used, so more soon on this.
I will have pictures soon this week.

Additionally to make it easy to use, the bellows mount works as follows. Mount is pulled off from bellows.  That's right, pulled off.  It is a magnetic rear lens mount with a finger cutout to make it easy to rotate to connect to the body.  Mount the bellows to the the front frame of your DSLR2 and reconnect the two rings via the installed magnets. As the two rings nest inside, a perfect seal and light trap is created.  Turn on your body and you are ready to make images.  If you want to switch to vertical or vice-versa, you merely need to reorient the body and the bellows mount will rotate in place.  You now have an extremely precise camera capable of using wide angle lenses on your DSLR with ease.

All three mounts will be supplied in this form in the future.
Enjoy,
Rod

[Chris Barrett]

Rod was kind enough to send this over for testing.  Putting it through it's paces today.  Will report back with IQ260 images for comparison.  I gotta say, I love LiveView for swings!  SK 35XL focuses no prob.

[GregShapps]

That Arca looks very similar to my Cambo set up that I've been using for 3+ months now.

[Rod.Klukas]

Thanks for taking a look Chris.
Let me know if any concerns arise.
Be well,
Rod

[klane]

Chris how much tilt/swing can you achieve with the 35mm?

[klane]

Any results to post?

[cyron123]

Wow. This is a very interesting setup. Is there a sony mount attached directly to the bellow?

[Chris Barrett]

Here are some jpegs from a test...  the sun was in and out so you can't really judge DR very well, unfortunately.  If you look at the ceiling fixtures, they blow out on both cameras.  I always bracket for this stuff, so in actuality any DR difference (within a stop or less) is really inconsequential to my workflow.  I always have to bracket and do tone blending.  I'll include some quick retouches of the images.  I did use an LCC with the Sony / SK 35mm XL and that worked quite well.

One thing I noticed is that, while the Sony feels sharper (in camera sharpening???) the Phase has much nicer hi-light transitions (look at the orb fixture reflecting in the wood ceiling).

Arca Swiss M-Line 2 used with both sensors.

IQ 260 / Rodenstock 55mm / f11 / 0.5s / About 20mm shift and rise (never could have done this shot with my Rm3d!)  UNRETOUCHED



Same shot with some retouching and exposure blending



Sony A7r / Schneider 35mm XL / f11 / 0.5s / 10mm shift and rise / -1mm swing  UNRETOUCHED



Same shot with some retouching and exposure blending



Full size images available here:  http://christopherbarrett.net/forum_images/Camera_Tests/Tifs/

Overall, I still think the IQ260 makes nicer files.  They're very very close and with the retouching that I typically do, the Sony is more than capable as an architectural camera.  If I was shooting with it alone, I would carry the Cannon 17mm TS-E, 24mm TS-E II and then the Arca for lenses from 35mm to... well I wonder how long that bellows will focus?

I'm working on a full review of the Arca M2 / A7r rig and will link to my blog post when it's up.

CB

[Ken R]

Here are some jpegs from a test...  the sun was in and out so you can't really judge DR very well, unfortunately.  If you look at the ceiling fixtures, they blow out on both cameras.  I always bracket for this stuff, so in actuality any DR difference (within a stop or less) is really inconsequential to my workflow.  I always have to bracket and do tone blending.  I'll include some quick retouches of the images.  I did use an LCC with the Sony / SK 35mm XL and that worked quite well.

One thing I noticed is that, while the Sony feels sharper (in camera sharpening???) the Phase has much nicer hi-light transitions (look at the orb fixture reflecting in the wood ceiling).

Arca Swiss M-Line 2 used with both sensors.

IQ 260 / Rodenstock 55mm / f11 / 0.5s / About 20mm shift and rise (never could have done this shot with my Rm3d!)  UNRETOUCHED



Same shot with some retouching and exposure blending



Sony A7r / Schneider 35mm XL / f11 / 0.5s / 10mm shift and rise / -1mm swing  UNRETOUCHED



Same shot with some retouching and exposure blending



Full size images available here:  http://christopherbarrett.net/forum_images/Camera_Tests/Tifs/

Overall, I still think the IQ260 makes nicer files.  They're very very close and with the retouching that I typically do, the Sony is more than capable as an architectural camera.  If I was shooting with it alone, I would carry the Cannon 17mm TS-E, 24mm TS-E II and then the Arca for lenses from 35mm to... well I wonder how long that bellows will focus?

I'm working on a full review of the Arca M2 / A7r rig and will link to my blog post when it's up.

CB

Thanks for posting this.

The IQ260 images look much better IMHO. The Colors are just much cleaner and distinctive. The greens and blues look much better. Color separation looks much nicer. The A7R images have an overall yellow (warm) cast that seems hard to get rid of. You reduced it a lot in the retouched image but it is still there. The blue wall behind the TV makes it obvious that the IQ260 deals a lot better with color. The green tint on the glass on the 2nd floor also looks much better on the IQ260 image.

[torger]

I would expect the A7r have some crosstalk issues with the 35xl causing color accuracy problems in the shifted areas. Having the sensor horizontal as here is likely better than vertical though. I'd love to do some crosstalk cancellation tests with my new algorithm with this combo :-)

For that I need two LCC shots though, one normal and one with a red filter (used to measure crosstalk).

Edit: just looked at the images on a proper screen, and yes you see that the A7r has much less saturation of colors on the green glass and the blue wall behind the TV. A primary symptom of crosstalk is desaturation. In the lens center it should look okay, but when you go towards the edges you'll have desaturation issues, which can't be fixed by regular LCC. You need crosstalk cancellation.

If you shoot a red-filtered LCC shot (just put a red color separation polyester filter between lens and LCC card) with the same settings and send me the shot, LCC and red LCC I can send back a crosstalk-cancelled shot to play with. Haven't a production ready algorithm yet, but hopefully it should work on this.

[torger]

The main symptom of crosstalk is as said desaturation of colors, some color shift can occur too. This occurs because the leaks between color channels bring them closer together, ie desaturation.

There's another symptom too though which is mazing in the demosaicing. I've cut out a 100% crop of the top right corner of the A7r image to show how it looks, both in original and increased contrast that makes it more visible.

The mazing occurs because crosstalk will pass from red to green in one column and from blue to green in another, which means that the both green pixels that without crosstalk would have equal values (for a solid color ceiling) will get separated. Separated greens means that the demosaicer will see and try to render detail there, and you get mazing.

It should be said that Capture One's demosaicer is quite robust against green separation, so the result here is better than it would have been with many other raw converters. The demosaicing algorithm designer must make a tradeoff between being robust and extracting detail, and it seems like Capture One has focused more on robustness than maximizing fine detail.

I've got to study these issues in more detail when the first tech cam demonstrations with the IQ250 was released which really highlighted it. Since then I know that crosstalk has been there with tech wides also for CCD sensors, I can even get it with my old Aptus 75 and the SK35XL and even the modern SK60XL when I shift to the image circle edge, but with the Sony sensor like in the IQ250 and here in the Sony A7r the crosstalk is much more severe. When developing the crosstalk cancellation algorithm I've had my own camera system as test bed where it seems to work well, I'm not sure how it will handle the more severe crosstalk with the A7r and IQ250 though, so I'd love to get some test material if I can . If I can make it work it will make both the A7r and IQ250 much more usable with tech wides. It's that red LCC shot I need in addition to the normal white which make it a bit more cumbersome (it does not need to be shot in the field though, can be shot later if you just have noted the shift/tilt/aperture settings).

With these new CMOS sensors in mirrorless cameras and digital backs I think it's good to be aware of this issue so you know what can happen and what to look for to detect it (ie desaturation and mazing).

[cyron123]

Hi,

sorry for my silliy question. You could use a wide angle lens like the schneider 35mm with a Sony A7R??? I thought that there is only lenses from (something) 70mm possible, because of the mirror box?

And you could do movements with this package too?

--> Wonderful. OK the results are not as good as from the Phase One but the costs are different too.

Do you think the sony is better for that use, because of the small mirror box?
Thank you.

[torger]

sorry for my silliy question. You could use a wide angle lens like the schneider 35mm with a Sony A7R??? I thought that there is only lenses from (something) 70mm possible, because of the mirror box?

The A7r is a mirrorless camera, ie no mirror box. However there is space required for the lens mount so you can't get as close as with a digital back. Obviously from the demonstration the SK35mm does fit though, with at least 10mm shift movement. You need to relate to the crosstalk though, Sony CMOS sensors are not designed to cope with the angular challenge tech wides present, either correct it or live with it.

[cyron123]

Hi Torger,
Yes :-) a mirrorless box is not there! I mean the space between.
I think the A7 is a very good alternative. Especially the price!!! I know that you use a techno. Do you gave some experience with a A7 too? What board do you use to attach a A7 to Linhof?

I know that a German company called " just together" build boards for dslr cameras. But for the A7? I don't know?!?!
Cyron

[AlanG]

These tests look good to me. I use DXO as my raw converter and I certainly feel I could adjust the A7r files to look more like the 260 ones if I wanted to. The selective color adjustment, local contrast in fill light, and shadow radius are some of the tools I'd use.  These and other adjustments really can change the character of a file. So as long as the camera captures the dynamic range with good detail I think I am ok.

I also have a lot of experience shooting commercial interiors and know that the discontinuous spectrum of many lights often plays havoc on accurate color reproduction. One cannot expect any camera on its own to overcome this because if the color is not in the light in the first place it can't magically be restored when reflecting off a colored object or by changing the color balance.

[cyron123]

Hi guys, simple question: what is the crop factor of a Sony A7r with LF lenses like the rodenstock? Thank you..

[Chris Barrett]

Hi guys, simple question: what is the crop factor of a Sony A7r with LF lenses like the rodenstock? Thank you..

Well, think about it like this... if you use the Schneider 35mm XL, it's the same as having the Sony Zeiss 35mm on the A7r.  Thinking about crop factors causes unnecessary confusion, just look at the MM (focal length).

CB

[AS1]

Chris,
Thanks for posting these files; fascinating...!

I'm assuming both are from raw files, and if so, I'm amazed at the sharpness difference in the files? Is this the difference in the lenses? The 55 seems softer with more chromatic aberration (looking at edge with the toaster oven area)? Or does the Sony have some kind of in camera pre-sharpening being applied, even to raw files?
Did you apply any chromatic aberration corrections to either files? I'm assuming you didn't since they aren't available as presets because of the lens movement issue.


Either way, a very interesting comparison....

Alan.