9000 dpi effective DSLR scanner

[Pere Casals]

You can avoid the need for a flip out screen if you tether the camera. Most cameras support this.

Larry, yes you are right.

Time ago I was using DigiCamControl ( http://digicamcontrol.com/ ) connectig DSLR to a Laptop with a USB cable. It would be perfect to focus on the easel. The DSLR flip screen is small, good enough if focusing wide open and then stopping two stops, but it would be challenging to focus to see how the enlarging lens behaves wide open.

I've to say that I lended my dslr gear to a friend for 3 months so when another friend came with his D5100 I saw the opportunity . I was happy to see my dslr departing for a long trip but now I longing for it a bit.

The projection method would also be useful to check taking lenses, this is (horizontal) projecting the slide to some x30 the focal length distance on a DSLR without the lens, we would move the dslr to take the right groups and then we would focus conveniently thanks to tethering... This way does not evaluate field flatness but at least it removes the alignment problems in the measurements when evaluating corners.

idea to project the negative and this makes a lot of sense since the DOF would be larger this way and focusing accurately less of a challenge.

Yes... stopping two clicks, that it would be around the sweet point, I find that focusing it's quite easy.

Also I guess that it would be easier to displace the negative than the camera and the lens. In this way we may use a tinny illuminator covering only the region that ends in the sensor, we may move the negative carrier and then illuminator and the camera would stay in a fixture. At the end an X5 works a bit in that way, lens and illuminatin stays and film displaces.

To move the glass holding the negative we may use a plain CNC XY kit, around $150:



An arduino cnc controller its really cheap, we only have to write the G code for the movements , the camera triggers and waiting time in beetween. Those controllers have switches for the engraving laser activation that can be used to trigger the camera. This is simply writting the movement instructions in a text file.

Also cheap 3D printer kits can be used, with a RAMPS board, IMHO this is the easiest way to automate the glass movement and camera operation.

I'm proficient patching 3D printer firmware, last summer I spent two weeks understanding internal Marlin firmware structure, I was tracing all the code, it was a fantastic learning because it's a very well made code. Feel free to PM if you need something about that.

And I do believe that lots of older lenses are quite good

Yes... IMHO it's really interesting to evaluate enlarger lenses. Printing the projection of the target slide and inspecting it with a magnifier it's quite easy, and with a DSLR it's even straighter...

Just spending an evening to test our gear at different apertures and magnifications would deliver nice information. There are plenty of gossip around about enlarger lenses, some testing (say 2 hours) would improve a lot our own criterion.

What I'm realizing is that a refined darkroom technique it's the critical factor, as many said before.

[Drew Wiley]

Let's see... how many entire texts of the Bible could they fit on a microdot clear back in the 1930's. That should put some of this in perspective.

[calebarchie]

Let's see... how many entire texts of the Bible could they fit on a microdot clear back in the 1930's. That should put some of this in perspective.

Are you suggesting we should return to WW3?

[erian]

To move the glass holding the negative we may use a plain CNC XY kit, around $150:

I have been thinking about this solution.

I think that common solutions will easily work with the medium format. Perhaps up to 4x5 but I think thqt from there it would be difficult to manage negative holder stability due to lack of sufficient support.

I think that what we are looking for is a XY table.

Another solution is to move the camera (or both on single axis) but moving the negative carrier could be a more stable solution.

Perhaps existing affordable solutions could be adapted with additional custom components. For example some of the existing CNC drawbots already contain 2 rails for the both axes that could make the adaption fairly easy.

[Peter De Smidt]

You might want to check out the DIY part of this site. There are two versions of an automated xy table that people here made, a control system,....

[Pere Casals]

I think that what we are looking for is a XY table.

Yes... I've have been searching, and the solution I liked the more is hacking 3D printer hardware, because it has the right size. New driver chips make movement silent and smooth,

The proposition in this thread is using a reversed enlarger lens, with the lens back facing the film, and with lens way closer to the film than to the sensor. As noted by EH21 member this will allow more DOF in the Sensor plane, but we should critically ensure that the film is flat and in place.

With that configuration it's easier to move then film than the camera (having a long bellows draw), the camera fixture should be as rigid as possible.

Then we need flat film. An idea would be using ANR glass for the bed, the film on it with emulsion facing up, and placing on it a glass with antireflective coating, say True Vue brand.

I checked it again and the reversed enlarger lens is 10k dpi effective able, this is a plain overkill even for 35mm, for LF often we don't need more than 3000 dpi effective, IMHO.

But having an "unlimited system" is always nice, from that we may adjust the resolving power vs speed as we want. It may happen that if we change the lens magnification then we should use a duplication lens, a Rodagon D 1x or 2x.

I've only tested the EL-Nikkor at x4, I've pending to check how it works at x2 or x1.

[erian]

Yes... I've have been searching, and the solution I liked the more is hacking 3D printer hardware, because it has the right size. New driver chips make movement silent and smooth,

3D printers usually moving the head on the XY axes, so not very suitable out of the box. Some I have seen move the bed on X axis and the head on the YZ axes but this solution must move the bed on XY axes.

There does not appear to be many off-the-self products that do that. Closest to my idea are some XY plotter kits but these will require custom components for modifications.

Do you have something specific in your mind based on what you plan to do your modifications?

[Pere Casals]

Do you have something specific in your mind based on what you plan to do your modifications?

IMHO a suitable XY table can be made with two sets of Prusa i3 guides/belt/stepper for the Y axis, that one moving the bed. Using those small steppers I'd use good/cheap drivers that are silent and smooth (TMC2130) in the RAMPS control board.

IMHO motors/guides/electronics components from 3D printers match better the requirements to move a negative under camera than components from a CNC Router. Perhaps using CNC router components (post 21) may require less DIY workload. Both choices are good.

[Jac@stafford.net]

But having an "unlimited system" is always nice, from that we may adjust the resolving power vs speed

It might be necessary to include image stabilization somewhere in the capture phase.

[Peter De Smidt]

Jac is right that having a minimal vibration system is very important, and it becomes more so as magnifications rise. At 1x, shutter vibrations cause problems, especially with exposures from between 1/15th of a second to 2 seconds. An Electronic Front Curtain Shutter (EFCS) is highly desirable, as are very well designed support and movement systems. Wait till you tether your camera, go to a max live view, and lightly tap anywhere on your system....A lot of the systems I've seen people build are not anywhere near close to good enough to maximize capture at 1x, let alone 9x! I say this as someone who spent a couple of years building and testing these things.