DSLR Scanner: Light Sources

[marfa boomboom tx]

thanks...

[nonuniform]

I would love to see some more photos, especially of your film mount system for wet mounting. I don't really get the spacer concept. On my v750 wet mount, I just squirt the glass, lay down the film, add mylar. No taping for that. I was thinking I would go with a similar setup for my own scanner, so, I'm wondering how the spacer helps, or, well, whatever!

Thanks!
Andrew

I will create a video of it in action once I complete my most recent modifications, new motors and upgraded belt clamp. I took a wild leap and attempted to make an X/Y stage from two stacked IGUS Drylin rails. They are anything but precise in their build, in fact they are sloppy, but seem to be precise enough in the resulting travel. Peter is using rails and bearings that are capable of far greater precision. Still a lot of tests to perform as he mentioned. I no longer own the V700, I feel a curse has been lifted being free of it. I know lots of folks are happy with it though.

I am using Capture One at this stage, tethered to a D800E. I find it's sharpening to be excellent, the main reason for using it right now.

I'm wet mounting with lighter fluid for now (I have plenty of less than stellar film exposures to test with : ). I lay a spacer on the glass, squirt the glass, lay the film against the spacer which aligns the film edge and centers it at the same time, no taping, squirt the film, lay a sheet of mylar down. That takes a minute or so. A 9 shot 6x7cm scan takes about 45 seconds give or take with delays programmed for settling the carriage motion and another for a 2 second exposure delay mode. For the image I posted I inverted in C1 with the levels slider and then adjusted the color by pulling the white and black points for each of the three channels until they touch the histogram curve. That's all for the Raw files, which takes maybe 30 seconds. Processing for these Raw files takes less than a minute on a new build with a 500MB/s SSD and i7-4770 CPU. Stitching the Tif output in PTGui using Sinc36 takes another minute. Using Sinc256 takes a few minutes. No intervention has been required as of yet to get a good stitch but that's undoubtedly coming up with featureless areas.

I'll post details of the build in a few days, still ironing out some things. I will say that wet mounting makes the impossible possible. All of the tests I've done so far are on a roll of film that was in a tube for a year and has a terrible curl which was easily tamed by wet mounting. DOF is so incredibly shallow that it may even be mandatory.

Resolution with this camera at 1:1 is a little more than twice the V700 if one takes the widely accepted V700 2100 spi as a given. 7cm turns into 47.5" @ 300 dpi with this camera.

I'm eagerly waiting for comparisons myself! Drop me an email if you'd like to arrange for me to scan a sheet of your film, this seems a good way to get a range of comparisons. All I know at this point is I'm over the moon with my color neg scans for the first time ever, and I'm amazed at how easily they were obtained. All those hours I spent with different programs tweaking tweaking tweaking. God.

[Peter De Smidt]

Daniel only uses the spacer to keep the negative lined-up properly. We want to avoid having to do any image rotation in post production, as that leads to a loss of quality.

[Daniel Moore]

Exactly. I've found that with care, it's not hard to prevent bubbles being formed. When they are you need to be extra careful to push them out keeping a couple fingers on the mylar to prevent slipping around once the spacer is removed. I'm remaking the spacers to allow for more accurate indexing along both axes, an L shaped spacer. I'll be posting more detail images soon.

[DannL]

Attachment 101442 Attachment 101443 Attachment 101444 Attachment 101445

A simple, inexpensive, portable, bright, uniformly lit, cold-cathode light source for countless projects. I've been using these off and on for a number of years. They work great as a small light-table for viewing negatives up to half-plate size (4.75"x6.5").

HP made a scanner that came with this ScanJet XPA Transparency Adapter. This particular XPA (pictured) was purchased off eBay several weeks ago for $5.00 minus shipping.

To adapted the XPA to work without the scanner . . . You will need a DC power source. I used a 12 volt DC, 4 amp power supply from an old Sony video camera. Cut the connector from the end of the XPA's cable. Attach the power source to the red wire = (+12 VDC) and black wire = (ground). Seal those connections with electrical tape. Twist the orange and brown wires together to make a separate connection and seal with electrical tape. Seal all your connections with electrical tape.

I've been using mine with a 12 Volt DC power source. So far, so good. But, if anyone knows a different voltage requirement for this unit, please let us know.

Measurements: XPA case is 8-3/8" L x 7" W x 1-3/4" H. The screen is approx. 5-1/2" x 6-1/2".

Enjoy!

PS: If you decide to invest in one of these, my recommendation is . . . "The Newer, the Better". The screen can get scratched up if handled roughly.

[Peter De Smidt]

That an interesting light source. Thanks for posting!

[Daniel Moore]

Nice catch.

[DannL]

Attachment 101449 Attachment 101450

The last picture in my previous post doesn't do a good job of depicting the screen brightness. This is probably a better representation. Internally there are fluorescent type tubes around the edges. They require a minute to warm up to reach full brightness.

[Peter De Smidt]

I'll pick up one of DannL's light sources when I get the chance.
It'll be interesting to see how it's spectrum does with colors.

I really like the ability of our current light source to adjust the RGB composition of the light. I've scanned a few bw negatives developed in a staining developer (Pyrocat). Being able to roughly equalize the capture of the three channels is useful.

[Daniel Moore]

Here's another example from this light source. It was shot with an Arca Swiss F Classic and Schneider 58mm lens. I dialed in a nearly neutral film base with the RGB controller and got it closer in Capture One. Besides inverting the black and white points, correcting the black point was the ONLY editing done to this image:

https://www.dropbox.com/s/y5qakswy1h...patio%20sm.jpg

Here's a 100% crop from the best focus point in that image with only default sharpening:

https://www.dropbox.com/s/dfvmq7tf93...%20%20crop.jpg

The result file as captured with my D800E is 80" long at 300 dpi. using almost exactly 1:1 magnification.

Color me happy.

I deleted the previous post as the images were being resized too small. I'll make a note to leave them in my dropbox.