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!

New - Cambo Actus View Camera for Sony A7

Stefan Steib

Active member
I just had a good laugh....because in the whole thread nobody even slightly mentions about this not being medium format nor a digital back ?

I also think it is OK and I am glad this is normal now, but do you remember what kind of discussions were going on here when the D800/E arrived ?

Welcome to the new world of Digital fusions.....

Greetings from Germany
Stefan
 

foster_jb

Member
Interesting that a helical mount "is not required." Is the focus mechanism much more fine than an Arca Swiss ML2?

Foster
 

Jorgen Udvang

Subscriber Member
This is the coolest device I've seen in a long time. It also helps making the A7 family of cameras one of the most, maybe the most, versatile camera systems ever made, in spite of an almost total lack of native lenses. I might have to revise my shopping plans.
 

mbn

New member
I bought an digiback for the movements the cameras will give me.
Medium format is all about movements for me.

If this cambo works well, my alpa and digiback will be for sale soon.
 

Shashin

Well-known member
The A7r had a few problems with adapted lenses. Adding shifts, swings, and tilts might prove more problematic. I assume the A7r has microlenses.
 

tmh

New member
The Rodenstock HR and Schneider Digitar lenses can be mounted in the Cambo board for the camera (very affordable, usually less than $100) - they don't need to be mounted in the Cambo Helical LensPanels.
So... rather than going off in search of more lenses, if I have a very decent Cambo WRS mounted lens that I want to use on this Actus, I'd have to remove the helical and then fit the copal and lens into the basic Cambo copal board?

I don't have my WRS in front of me, how difficult is it to remove the helical? I've put lenses into copals and boards before, but never had to get it out of the helical.

Pity Cambo aren't offering (or maybe they will later?!) a board where a WRS/WDS mounted lens can slot straight in - then I can pick between Actus or WRS body depending on the job and not have to worry about remounting my favourite lenses...


First though, will have to wait to see how the A7R performs at the edge of the image circle...
 

torger

Active member
Chris Barrett has already tested the A7r with an SK35 over at Lula (using the MF-two) and showed the result, it's there for anyone interested.

The deal with these Sony sensors is this: they are extremely crosstalk prone, ie poor angular response. Due to the smaller pixel size the A7r is worse than the IQ250, and the IQ250 is not particularly good with the wides. I've attached an image that shows what crosstalk is if you don't know. It's light getting registered in the wrong color channel.

Just like the IQ250 the A7r has offset pixel (in relation to microlenses, popularly called "microlens offset" but it's actually the pixels thet are offset) at the sides which is to handle wides better, but that is only when they are not shifted. So if you don't shift you can probably go quite wide.

But when you start to shift you get immediate problems with the wides, due to the massive crosstalk you get color issues (in the best case just desaturation, but color shifts can occur too) and eventually demosaicing artifacts. Unlike regular color casts, a normal LCC algorithm such as the one available in C1 won't correct crosstalk.

How much crosstalk artifacts you can accept is a matter of taste though, and I've noted that tech cam users have a tendency to care a lot about sharpness and not that much about color stability, so I guess many will be pleased with quite wide lenses.

Also note that the result will vary greatly if you have the sensor in horizontal or vertical orientation when you shift, as crosstalk is a lot more in one direction (forgot which one it is with the A7r, with the IQ250 I think it's horizontal that's worse if I remember correctly)

If you want to shift and don't want any crosstalk you'll need to use longer lenses than say 60mm, or use regular strongly retrofocus MF-SLR lenses. For wide angle tilt-shift photography with the A7r the Canon TS-E II lenses is still the way to go, and for the 30-50mm range I guess you could find some decent MF-SLR lenses. If you care about color tech cam wides will not be a good choice. At least not until someone comes up with a crosstalk cancellation algorithm that works with this large amounts of crosstalk (I might just do that... so please do buy into this system ;) )
 

GrahamWelland

Subscriber & Workshop Member
So, the real question is how wide will it go to support SK or RS lenses?

Torger's point about crosstalk is relevant for colour fidelity but I'm sure that post processing could solve many of these issues albeit with the loss of some real data.

I'd buy an A7R for this if it works with something reasonably wide.
 

fotografz

Well-known member
Thanks for posting this Steve.

My main interest would be for table-top product photography … hopefully using my trusty Rodenstock 90mm, and Schneider 120 Macro with Copel shutters previously used on a Rollie Xact-2. Objective being to gain more DOF front to back using the optimal f/stops for each.

Selective focus portrait work is also a prime interest.

I assume that this will work, right?

- Marc
 

torger

Active member
First to make one thing clear: crosstalk applies to normal to wide tech lenses (say 60mm and lower) in shifted positions. Table top photography with longer lenses will most likely work extremely well with this camera. But this camera looks like a portable field camera rather than a studio camera, which means that people will be interested in wide angle performance. For studio photography with longer lenses one is probably better off with a camera that has rear movements as well, such as the Rollei X-Act 2 which some users already use with the A7r (it costs twice as much though, $4K).

Back to crosstalk;

Capture One LCC algorithm does not correct crosstalk at all. It's a quite complicated thing to do. However for mild crosstalks you will only get mild color desaturation, so you can ignore it or make a radial saturation increase. For more heavy crosstalk the color will be so dead and/or shifted you cannot just increase saturation, and for severe crosstalk you start to get mazing artifacts in the demosaicing.

Crosstalk with tech cam wides is nothing new. I actually have it even with my large pixel Aptus 75 when combined with the SK35 and shifts, and even with the new SK60XL at the edge of the image circle, so tech cams users have lived with it for long and been satisfied -- mild crosstalk is not too bad, and if the scene does not have very saturated colors you may not even see it. However with the Dalsa 6um sensors (P65+, IQ160) crosstalk became worse and in response SK35 became less popular and many upgraded to the (weak) retrofocus Rodie 32 for better performance with those new sensors.

Then we have IQ180 which have even higher levels of crosstalk, and IQ250 worse than that, and then after that the A7r comes in. So there's a whole new level of crosstalk which we haven't seen before, and thus we haven't really had time to investigate how much post-processing can solve these issues. If we're lucky it can solve a lot.

I just want to warn those that intend to buy into this system early that it may not perform well at all with tech wides, and new post-processing techniques may need to be developed before it will, and it's uncertain at this point how successful that will be.

Back in the days there were many upgrades from P65+ to IQ180 without that the users have been informed about the increase in crosstalk and what effects that would have on the lenses they owned. A7r / Actus is coffee money in comparison of course, but still I hope people get information about this issue this time around, before they buy rather than after.

A simple sanity check test to see if a lens/sensor combination has crosstalk is to shoot one normal white LCC and then one more LCC with a polyester color filter stuck inbetween, say a red Wratten 29 filter so you get a red LCC (other colors work too, but red is best for testing). Then you generate an LCC from the white one and apply to the red shot. If there is no crosstalk you should see a flat red surface with the same saturation over the whole surface. If there is crosstalk you will see desaturation in the red, and if there is heavy crosstalk you may see a color shift, even that red becomes green. Do test both horizontal and vertical orientation of the sensor as the result can differ greatly due to how pixels are wired on the sensor.

The thing with crosstalk is that it will depend on what colors you have in the scene you are shooting what the result will be. If you have colors that are low in saturation the result will be better, as crosstalk will have a lesser effect on those colors. This means that a single scene shot may not give you the whole truth of how the system will perform in various conditions.
 

Sunchai

Member
So, the real question is how wide will it go to support SK or RS lenses?

Torger's point about crosstalk is relevant for colour fidelity but I'm sure that post processing could solve many of these issues albeit with the loss of some real data.

I'd buy an A7R for this if it works with something reasonably wide.
The image have showed the distance from Flange of Sony to infinity is 18mm.( Rodenstock 45mm 4.5)
the lens that can be use with Sony A7 in this case have to have the distance from the lens end to infinity (d-e more than 18 mm.)
From the Rodent lens description the HR Digaron-35mm is the shortest lens 53.5-29.2 = 24.3 when focus to infinity and will have the balance space for adjustment of 24.3-18 = 6.3mm



 

Digitalcameraman

Active member
Marc,

Those should work as long as they are Copal #0. They certainly can focus the same as they do on the Cambo X2 which I am a big fan of.

I just wanted to add some excitement I have for this new product that my associate, Steve Hendrix has already outlined.

Having many MF back customers using the Cambo X2, I became very interested in the support of Leica R lenses.

We have used many Hasselblad V and Mamiya RZ lenses in the past with the Cambo X2, but the Leica R lenses are out there and available. Being able to use those in addition is a huge plus, once we know how this chip reacts to tilts and swings.

I know when doing a lot of testing with Canon 5D Mark II and MF backs, the Schneider and Rodenstock digital lenses out performed the Hasselblad V and RZ lenses.

Mamiya M645 lenses are also supported but not Mamiya/Phase One AF lenses due to the fact of not being able to control the shutter speed and aperture. I have even taken Sinar DB Digital lens like the 80mm and mounted it into Copal shutters and got fantastic results. So many lens options for devices like this one coming.

Copal #0 shutter boards look like the largest being supported mainly because of the size of the footprint. And it would be slick if we could take a Cambo WDS lens mount and adapt that to this device for all our DS and RS customers.

Stay tuned Capture Integration will have all those answers for you soon. Go Steveeee.


Chris Snipes






Thanks for posting this Steve.

My main interest would be for table-top product photography … hopefully using my trusty Rodenstock 90mm, and Schneider 120 Macro with Copel shutters previously used on a Rollie Xact-2. Objective being to gain more DOF front to back using the optimal f/stops for each.

Selective focus portrait work is also a prime interest.

I assume that this will work, right?

- Marc
 

Paul2660

Well-known member
I also wondering if you do shift, with the Sony which I would want to do, will C1 process the LCC's. C1 seems to be very camera specific on LCC's and will not for example allow you to use a LCC from a IQ160 on a IQ260.

I had wondered a few years ago, if I shot LCC's with my Canon 24 TS-E, would C1 see them and allow me to apply them to images taken with my 5d MKII. I was not then worried about color cast, as it's not an issue, but with the light fall off, which you still get. C1 does a very good job with light fall off in LCC processing.

Question to Torger, Crosstalk.

How much is crosstalk, vs, just light fall off. You can see for example with the SK35 when shifted past 8mm, a huge amount of color saturation fall off and detail smearing. This also happens with the SK43 past around 14mm on the 60MP backs. I have always assumed this was due to lens quality, at extreme edges of the image circle, but I guess some may also be due to crosstalk. There is a definite line with the SK's where you hit the wall and the image past that spot suffers too much. You also pick up the rippling which C1 sometimes can fix on a SK, but is hard on for example a pure blue sky. C1 does better with the Rod lenses with rippling. On the color/sat fall off, C1 does not fix this at all, but again I have assumed this was due to the image circle.

Quesiton to Steve:

I am assuming that the Rodenstock lenses will have to have a Cambo mount? Which would be cost prohibitive (at least to me) for folks using a system other than Cambo, i.e. Arca or Alpa.

Paul
 

torger

Active member
Question to Torger, Crosstalk.

How much is crosstalk, vs, just light fall off. You can see for example with the SK35 when shifted past 8mm, a huge amount of color saturation fall off and detail smearing. This also happens with the SK43 past around 14mm on the 60MP backs. I have always assumed this was due to lens quality, at extreme edges of the image circle, but I guess some may also be due to crosstalk. There is a definite line with the SK's where you hit the wall and the image past that spot suffers too much. You also pick up the rippling which C1 sometimes can fix on a SK, but is hard on for example a pure blue sky. C1 does better with the Rod lenses with rippling. On the color/sat fall off, C1 does not fix this at all, but again I have assumed this was due to the image circle.
LCC shots is not only camera model specific, it's specific per specimen, each sensor has their own "fingerprint" and can very quite much chip to chip, at least the MF CCD chips, not sure about the Sony CMOS sensors, might be more uniform.

Color saturation falloff is a pure crosstalk effect.

Concerning smearing, crosstalk does not actually cause that much smearing as it's only say 20% of the light that travels to the next pixel and much less to the next past that which affects color but not detail very much. Smearing at the edge of the image circle is much more than that, so that's a lens sharpness issue. However when you see mazing, ie strange micropatterns when you zoom in to 100%, you see a crosstalk effect - the demosaicer gets confused when red leaks into green1 and blue to green2 thus separating the two greens, this cause mazing. Should be said that C1's demosaicer is one of the more robust demosaicers so you need pretty heavy crosstalk before it starts to produce mazing.

If you use the center filter on the SK35 the optical vignetting is virtually zero. However there's pixel vignetting too, ie light that hits "the walls" (not all light hits a photodiode), and I think there's a quite large amount of that with the 6um Dalsa sensors, but I'm not sure as I've not investigated it closely.

Microlens rippling is not a crosstalk effect but a microlens shading effect, but if you see rippling you have crosstalk too for sure. The ripple pattern is affected by the crosstalk which makes it very difficult to clean up, as the rippling on the LCC shot and in the real scene will not match 100%. C1 does a good job concerning the challenge.

When you know what to look for you can produce visible crosstalk artifacts also with the SK60XL far out in the image circle. It's obvious that the lens designers have chosen to accept some level of crosstalk, as it works in practical image making and a symmetrical design is an advantage in many other ways.

SK lenses are more symmetrical than Rod lenses, hence more crosstalk issues with them, more microlens shading etc.
 

RodK

Active member
I know that Chris Barrett reported using the S/K 35mm with the A7 on an A/S ML-2 (probably unshifted), so I'd imagine the same deal here.

For shifts with W/A lenses I'd imagine you'd need to go with the R/S wides* as these have the longer flange-sensor distance. Not sure how much the recessed nature of the sensor in the A7 is going to vignette the effects of shifting though - time (and testing) will tell.

Jim

*With no outrageously expensive helical and lens mount to buy, this isn't as much of an eye-watering proposition as it used to be.
The Arca-Swiss DSLR2 with the Sony A7 series or Nex can accept as short as 32mm lenses with lots of movement. Nex body may even go down to 28mm. Haven't had a chance to try yet.
Nikon and Canon are limited to 70mm and above.
We have a new magnetic bellows connection, which makes it very easy to mount and move.
You might enjoy the added precision of the Arca-Swiss.
And the increased movement that the shuttered Rodenstock and Schneider lenses can afford you, plus the ability to change up when you want to the benefit of 16 bit color a MF DB can bring you.
Rod
 
Last edited:

Shashin

Well-known member
...plus the ability to change up when you want the benefit of the 16 bit color a MF DB can bring you.
Rod
???? You need to do more than put a 16-bit A/D converter in a back to make it 16-bit. The sensor needs to be able to record 16-bits of information.
 

Steve Hendrix

Well-known member
I also wondering if you do shift, with the Sony which I would want to do, will C1 process the LCC's. C1 seems to be very camera specific on LCC's and will not for example allow you to use a LCC from a IQ160 on a IQ260.

I had wondered a few years ago, if I shot LCC's with my Canon 24 TS-E, would C1 see them and allow me to apply them to images taken with my 5d MKII. I was not then worried about color cast, as it's not an issue, but with the light fall off, which you still get. C1 does a very good job with light fall off in LCC processing.

Question to Torger, Crosstalk.

How much is crosstalk, vs, just light fall off. You can see for example with the SK35 when shifted past 8mm, a huge amount of color saturation fall off and detail smearing. This also happens with the SK43 past around 14mm on the 60MP backs. I have always assumed this was due to lens quality, at extreme edges of the image circle, but I guess some may also be due to crosstalk. There is a definite line with the SK's where you hit the wall and the image past that spot suffers too much. You also pick up the rippling which C1 sometimes can fix on a SK, but is hard on for example a pure blue sky. C1 does better with the Rod lenses with rippling. On the color/sat fall off, C1 does not fix this at all, but again I have assumed this was due to the image circle.

Quesiton to Steve:

I am assuming that the Rodenstock lenses will have to have a Cambo mount? Which would be cost prohibitive (at least to me) for folks using a system other than Cambo, i.e. Arca or Alpa.

Paul

Hi Paul -

The issue would be that if you had a pre-existing tech camera (Arca Swiss, Cambo, Alpa, etc), that the lenses would not be cross platform because they would have to be dismantled from the helical focus mounts of those cameras. There is a proprietary Cambo lensboard for the lenses, but that is not as important a factor as the dismantling of the helical focus mount, regardless of what board it would be mounted on from there.


Steve Hendrix
Capture Integration
 

Don Libby

Well-known member
Hi Paul -

The issue would be that if you had a pre-existing tech camera (Arca Swiss, Cambo, Alpa, etc), that the lenses would not be cross platform because they would have to be dismantled from the helical focus mounts of those cameras. There is a proprietary Cambo lensboard for the lenses, but that is not as important a factor as the dismantling of the helical focus mount, regardless of what board it would be mounted on from there.


Steve Hendrix
Capture Integration
So in other words a whole new set of lenses will be required?
 
Top