PDA

View Full Version : Robotic View Camera



max-crow
7-Jun-2016, 14:14
Hi all,
Does anyone tried to build a robotic view camera?

A 4x5 device would be bulky and super heavy I guess, but it would nice to have a small digital camera-compatible device.

Given that modern cameras may be controlled by smartphones, an app could make all needed calculations.
Just imagine that you connect a mirrorless camera and a 35mm-format lens to the robot, then take your smartphone, choose points that should be focused (or defocused). Robot makes measures, calculations and needed movements, and you need just to push the button :)
That wouldnít allow you to have a 4x5 slide on the light table as a result, but a camera with movements is better than camera without them and a small view camera on hand is better than kardan at home :)

What do you think?

Lachlan 717
7-Jun-2016, 14:35
Grab some popcorn; this should be fun to watch unfold!!

Jac@stafford.net
7-Jun-2016, 14:43
Would it climb a mountain for me?

Perhaps a Sherpa could release a few hundred of them, each paid by clients and we could have a race to the top in real-time.
.

metalsmith
7-Jun-2016, 14:45
I think the mechanism to pull the dark slide and put it back would OK, but how about flip the film holder for the other side?

ic-racer
7-Jun-2016, 15:02
Already been done as an enlarger. I suspect it could be made much smaller and focused with linear actuators on the 4 corners of the bellows. You'd need some type of reflex view so the user can see the scene without the focus sensor getting in the way.

http://b-29s-over-korea.com/aerial%20photography/images/E-4_sm.jpg

http://shop.educatec.ch/images/firgelli-l12.jpg

Bob Salomon
7-Jun-2016, 16:17
It was introduced at Photokina about a dozen years ago and was shown in prototype at that and the next Photokina but didn't end up being produced. Sinar introduced a computer controlled view camera about the same time but it was too expensive to be commercially viable. The one shown in prototype took backs from 69 cm to 11x14".

Daniel Stone
7-Jun-2016, 16:27
Grab some popcorn indeed :o

http://www.popphoto.com/gear/2010/09/if-only-matthew-brady-had-robotic-view-camera

http://www.capcam.org/capcam-components/camera/


https://youtu.be/NZtK9Q0mChg

Daniel Stone
7-Jun-2016, 16:28
https://youtu.be/CcVtFOSHnFs

B.S.Kumar
7-Jun-2016, 18:04
Actual user experience: http://www.getdpi.com/forum/medium-format-systems-and-digital-backs/54173-shooting-capcam.html

Kumar

stawastawa
7-Jun-2016, 22:04
that is a cool tutorial! a bit over kill but cool! and the guess and check is a bit much... now if you could select where the ultrasonic motors would drive your helicoid to align the focal plane with your desired AF focal point... =)

Daniel Stone
8-Jun-2016, 06:52
What kind of enlarger is that IC?

Drew Bedo
8-Jun-2016, 08:18
All that is interest ing,; now what about a robotic Grafmatic style film magazine?

It could have a smart phone sized imaging system for preview and then expose 4x5 films.

max-crow
8-Jun-2016, 11:37
Thanks for links! I haven't seen it before.

It's a cool device (especially the fully computerized version) but in the perfect world I'd like to see something more practical. That means:
1) Something more transportable.
1.a) Smaller. I canít find measures for CapCam but it looks pretty big. Iíd say that the device shouldnít be larger than a 4x5 field camera.
1.b) Lighter. "Tripod load for a fully equipped camera is 15kg.". So, 33lbs.. I believe all my Sinar F1 equipment (including tripod and film holders) is lighter. 4-5 lbs for the device body would be good :)
2) Something more affordable.
Lenses with electronic shooter + digital back are really cool, but the price!..
Iím ready to sacrifice the quality and get APC-S images only but for a reasonable price. If we use 3d printer components (they have moving/positioning parts too), we could expect similar prices (about $500-$700, for serial production).
3) Batteries + smart phone instead of power supply 230VAC/24VDC + PC.

Let there be Smaller, Better, Cheaper! :D

el french
8-Jun-2016, 12:36
I just need to add a back to this:
151602

max-crow
8-Jun-2016, 12:53
All that is interest ing,; now what about a robotic Grafmatic style film magazine?

It could have a smart phone sized imaging system for preview and then expose 4x5 films.

I see two ways to make a previewer:
1) To design a new full-sized 4x5 matrix. It could have a low resolution like 100-300 dpi, but it's a new device that would be quite expensive to develop, I guess.
2) To make an optical system a-la viewfinder that could redirect the image to the phone lens.

But in any case a previewer won't provide a lot of propositions. Just a preview + kind of light meter. Is it so valuable?
To calculate tilt/swing angles we need to know points' coordinates (on the ground glass) and bellows extensions. We could try to set a touch sensor to ground glass to select points and input bellows extensions into the phone manually. Then, an app could calculate optimal angles.
It's not a "single button" solution, but the app could calculate angles much faster than most people can (especially if you have more than 2-3 points).

Drew Bedo
9-Jun-2016, 06:51
This is a bit like chasing Unicorns in the mist. I believe that these concepts amd many others are well within the grasp of today's technology—that it is just a matter of investing in the design, engineering and production of any of these "Gee-Wizz" ideas. Marketing and return on investment is the limiter.

My father-in-law was a successful business man with a lifetime career in production management and industrial sales. After hearing me spinning out some really great sounding idea (seemed good to me) he matter of factly said, "Sure Drew, many ideas are great. The limiting aspect is always the money."

I'd like to see a lot of products re-engineered or brought out new using the new design and production tools available now, Sadly, I am not in a position to act on these things due to a lack of talent, skill and resources. I'm hust not going to learn the CAD/CAM software necessary to access the 3D printing world (I dread the day when I will have to give up my flip phone).

Yet I'd like to see a re-issue of Mido-type holders, a plastic Grafmatic, extension backs for my TravelWide (or a re-designed model), and a host of other similar things. In electronics, I'd like to see a shutter tester (there are a couple of threads on this: Lets not re-visite that discussion here), and the digital sensor based view camera pre-view/TTL light meter.

Heck, I think a view camera body for a micro 4/3rds or APS body that isn't thousands of dollars would be neat too.

I am sure that all these things are possible, but not commercially viable.

maxotics
10-Jun-2016, 08:28
I agree with Drew, a digital scanning back for the Graflex is technologically doable today. Very risky that a device would ever sell enough to earn back its investment, let alone a profit!

It is on my list of things to do. I would try first with a using the Spy Camera for Raspberry Pi (https://www.adafruit.com/product/1937), The cable length is 300mm, well above the 127mm needed for 5 inch width. My guess is that the sensor is about 6mm, so for a 50% overlap, you're talking 3mm captures.

Let's talk 4x4 to keep things simple. We'd need 33 captures for 33 rows, or 1,089 captures. The problems with those little cameras is they aren't fast at capturing full-resolution images. Video at 30fps is misleading about their still capabilities.

We could create a row of 4 of them (there is a camera multiplexer out there). That would drop our scans quite a bit.

If the capture is too slow, could use video mode, and find some way for the controller to flash a light just as the stepper motor paused, and use that flash to indicate we'll have a still video frame.

The biggest obstacle is a lack of low-cost, and various, linear actuators. Well, lots of obstacles, but again, the tech is out there.

max-crow
12-Jun-2016, 11:04
Thank you for sharing your thoughts. And you are right when saying that money is the main limit.
However, technology is not the only domain that developed during the last years. Product development was improving too. Such approaches as open source, crowdfunding, lean start-up, etc. can help to reduce risks and initial investment needs.

For instance, I could describe the aim of the project as to make view cameras (cameras with movements regardless of the frame size) accessible to mass customers. Where "accessible" includes: a) affordable price (comparable with good lenses), b) easy to use. Yes, it sounds like chasing Unicorns in the mist, but we could split it to a several stages:
1) tilt/swing angles calculator for LF users. A free web page/smart phone application that allows users to input coordinates of focused points (using ground glass grid) and bellows extensions to calculate optimal angels.
In the beginning of the stage we need only a few users (out of the development team) that would be interested to play with the tool just "for fun". The goals are: a) to prove a theory that computer can calculate needed angels faster than a man can, b) getting usersí feedbacks, to improve usability. Success can attract new users, with more practical motivation.
2) Angle meter. A problem is that most field cameras donít have angle marks, so the tool would be useful for monorail camera users mostly. A cheap solution could be to attach smart phones to standards and use their accelerometers. Again, in the beginning we would need only a small amount of tech geeks (people, who like to research new technologies), who would use it for fun. Having a working soft, we could move to compact independent accelerometers, which could be valuable for practical-oriented users.
3) Further automation.
3.a) Bellows extension meter. An LF camera-attachable device that could measure bellows extensions automatically.
3.b) Ground glass with touch screen. IR touch frames could be used, for instance.
In that moment we would have a quite valuable solution for LF users, but the more important is that we would have a full stack of working software and working prototypes, so we could raise investment (through crowdfunding, for example) to move to the mass digital market.
4.a) Small manual field camera with mirrorless camera as a digital back.
4.b) Robotic view camera.

Thus, the main idea is to increase product value step-by-step without revolutions. Every step should not be very complex. So, we donít need to attract dozens engineers and start massive marketing campaigns immediately. Yes, itís not a super-fast way, but it is realistic and starting investments are the time for pleasant activities only :)