Re: Don't buy a film scanner
Shawn67 wrote:
Alan71410 wrote:
Seriously. Unless you are considering an Imacon or other high end scanner, forget about these things - you have a far better piece of equipment in the sd Q-H. Scanners (up to and including the unsupported Nikon Coolscan 9000ED have a fixed focus and a wide aperture, and invariably inadequate film holders which don't keep the film flat. Unleess you're only copying mounted slides, just about any film will bow to some degree, which means that either parts or all of the grain will lose focus. The camera has both higher resolution thanks to better optics and much better DR than a run-of-the-mill scanner as well.
I did have a Nicon CS8000ED (Nikon no longer support it) and now do have an Epson V700 Photo which is absolutely pathetic compared to the results I get with the pictured setup (sd Q-H, A4 LED light pad, Durst Laborator neg holder with masks for anything from 35mm to 5x4), and my old 35mm copy stand repurposed for the job).
The camera is dead square to the copyboard achieved by bouncing a laser off the sensor from the copyboard - the return dot coinciding with the sending beam ensures the rig is dead level where it needs to be. The neg holder holds film perfectly flat, the 70/2.8 Sigma macro (old version) offers both the possibility of critical focus as well as DOF via stopping down if necessary. If higher resolution is required it is a simple matter to take sections and stitch them - with everything true and level and in focus, the job is easy and extremely quick.
Your setup looks great. I've considered a setup like that. What software do you use to invert negatives? How long does it take to process one of the images?
You are incorrect about fixed focus in the Coolscans. They do focus and because of that handle film at different planes. You can hear them AF at the start of a scan. Case in point, I added glass to my LS8000s film holders (both sides of the negative) and the scanner has no problem at all dealing with that. You can even set the focus point in the scanning software. Biggest downside with the Coolscan is it is *slow* and the Nikon Scan software needs older OSes to run. VueScan runs the scanner but seems to need more fiddling with MF film sizes compared to the Nikon software. Esp. for 6x9 film. Think Nikon Scan had better colors too. I haven't tried scanning in raw from Vuescan and converting in something like Colorperfect though. For 35mm, I typically use my Pakon as it does the whole role in about 5 minutes. If I need more resolution I move to the LS8000.
Shawn
I use Photoshop to invert, but I stitch the negs before inversion in AutorPano Giga or PTGUI software, depending on my mood at the time (If using Bridge as my browser at the time it is far quicker to send the images to APG directly rather than PTGUI which requires drag & drop. Adobe Photomerge can also do the job, but isn't as good at times in matching the pieces, particularly things like sky areas. For a four-segment 120 neg it takes around 30 seconds to do the "scan". I batch process the 3XF files in SPP, and save as TIFF (I have a preset for this saved in SPP now) and APG does the stitch in around 30 seconds to a minute start to finish (i7, 64GB ram). Then a few seconds to open and invert in Photoshop, apply curves to suit and save.
It's so long since I used my 8000ED that I'd forgotten the AF bit at the beginning. A component in the circuit board failed in my scanner around 2009 and Nikon couldn't supply a replacement, plus the fact that they had stopped making and supporting the scanner as well meant that I went looking for alternatives, although I've only just started getting back into film (B&W only) so I really wasn't in a hurry to do so. In the meantime both the sd Q-H and those nifty light panels appeared and the solution became obvious to me.
Having recently done a few 6x6 negs using the 70mm macro at full 1:1 extension and 12 overlapping frames (around 1GB file size, 13,200 x 13,200 px), the resulting images were about as close to actually printing them in my point light source Laborator 1200 enlarger as I could ever want with 400 ISO film; with 25 ISO film that's still not enough to properly resolve the individual grain, though, so the resulting "scan" shows the grain appearing bigger than it would be on a bromide print via enlarger.
Having a macro lens with both manual focus and an adjustable aperture makes a huge difference in keeping the image focused all over at film grain level. The critical thing is getting the camera dead level to the light panel/film holder in the first place, after that it's plain sailing. At present I am constructing a jig to enable shifting the neg holder around for the stitch segments that will keep the tracking straight and therefore make stitching even more accurate. Keeping things true by eye alone starts to become a bit haphazard when doing 12 or more segments at 1:1.
Having pulled the 800ED scanner down a few times between 2002 and 2009 (the period I used it - the 45° angled mirror that directs the image to the lens faces upwards and collects dust, which both depleted contrast and encouraged flare at high contrast edges), the way it was put together didn't inspire confidence and definitely didn't justify its rather impressive price tag, either. Also there was the perennial problem of having to use glass in the carrier in order to keep the film flat enough for the fixed aperture lens' extremely narrow depth of field to keep the film in focus. Anti-Newton ring glass or not, the fact is that it adds four more surfaces into the scanning sandwich which just means more work retouching extra dust afterwards.
