DSLR Scanner: Light Sources

[Daniel Moore]

Quote from rbultman in the scans and comparisons thread: "I'm having a hard time visualizing how turning on each color for a specific duration works relative to PWM? Did you try different PWM frequencies? What about using a low-pass filter after the PWM output to reduce ripple or using a DAC output? How many effective bits do you need to acheive the color range? Are you using a custom LED setup of your own design or a commercial one? Are you just looking for a change in color temperture of the light source?"

We're using an Armacost Color Controller from Lee Valley. The frequency is hard coded and I'm not familiar enough with it's inner workings to try and change it. The workaround is to turn the LED's on and leave them on during exposure rather than pulsing them on and off with PWM, which caused the banding artifact. Need more blue?, leave it on longer than the others for example. I read a low pass filter could help but this solution seemed easier. The color temperature is not the real goal of having the ability to adjust the LED's, it more to dial in a neutral gray before capture for each film scan to optimize color and tonal separation relative to the putput in one's preferred Raw processing program. This made obtain pleasing color straight forward and well worth the minute or so it might take to find it through trial and error. Using the same values for different exposures of the same film made it even faster as they provide very good starting points.

[rbultman]

Thanks for moving this here. I now realize that there are dedicated threads and will try to keep to the subjects of each.

Is it this one?

[Daniel Moore]

The one we're presently using is not programmable. http://www.leevalley.com/en/wood/pag...=1,43349,70322
That's going to go away however and be replaced by mosfets as in this circuit which does not rely on PWM and has only Red and Blue controls as the Green channel was always used the most for all film types:

https://copy.com/AWBinZu4cPdUKorC

Scanduino V2 will also use a different display, an OLED similar to what ReallySmall is using for his Stackduino V2, but that's a way off. Most of this circuit is a derivative of Stackduino. The drive towards generating an actual PCB is likewise driven : ).

[marfa boomboom tx]

other options, not as high tec, could be:

hasselblad: http://www.bhphotovideo.com/c/produc...Drum_Lamp.html

aztek: http://store.aztek.com/servlet/-strs...mp-Drum/Detail


best,
rl

[rbultman]

Very cool stuff Daniel. Fritzing looks very cool. I've never used it. I'm looking forward to your progress on this.

Regards,
Rob

[Daniel Moore]

Rob, that circuit is fully functional, the code needs an update to support the mosfets though. I'm going to keep at the PCB version and if successful, supply the boards for DIY'ers to solder up, or perhaps offer completed boards if someone wanted one.

[A pickard]

I am just getting started using a DSLR scanner with a copy stand and a D800 with a 105mm f/2.8 macro lens. For my light source I am using an Artograph Lightpad http://www.artograph.com/lightboxes/lightpad-930/ I am using this for B & W MF & LF B&W negs and it seems to be appropriate. Not sure about color temp & response. Since its not as bright as a flash unit, I lock the mirror up and use CaptureOne tethered to trip the shutter to reduce vibration. The light source seems even and I move the neg rather than the light source. At some point I hope to make an X-Y stage to move the neg carrier around. Thanks to all for your great input and suggestions.

[Peter De Smidt]

Thank you for posting. One of the great things about this is that it can be as complicated or simple as you wish. Tethered capture is a good idea. Please keep us updated on your progress.

[Zndrson]

Hi all,

Though I have not read through every post in this thread, I did want to offer my solution for correcting evenness issues using a light box/other light source.

In Capture One, there is a little known tool used mostly by Technical Camera photographers with digital backs called the Lens Cast Correction tool (LCC).

Used properly, a calibrated piece of translucent plexiglass is placed directly in front of the lens of the Technical Camera. An exposure is taken for middle grey, or slightly under.




The resulting image shows all of the lenses faults (vignetting, color cast, etc.). When an LCC is created in Capture One and applied to that same image, the software corrects for all of the faults. Those adjustments can then be applied directly to images taken using that camera/digital back combination so that the image is correct from the beginning.

I use this same technique when scanning with my D800, though the piece of plexi is replaced with the light table. I take an image of the light table first, and create an LCC in Capture One from that image. Then I'll photograph the negative, and apply that LCC to the image of the negative.

The benefits are obvious, and provides me with piece of mind. The software works very well, especially with subtle gradation changes. It can't work magic and make black to medium grey without significant noise, but as long as you get pretty close, the LCC application will do the rest of the work for you.

Hope this helps someone. Sorry of everyone knows about this tool already and is already implementing it in their workflow!

[Daniel Moore]

A late reply (sorry, I was doing summer : )).

This is an excellent corrective technique for those setups employing a fixed light source as in Peter DeSmidt's first incarnation, a flash based unit.

I have to wonder what the real world difference would be now that both Peter and I (and anyone else out there) have moving, full film size light boxes, would this be enough to rule out the falloff in a result stitch? I'm not in a position to test that theory easily.

This could really make someone's day, depending on their approach.