Drum scans of Adox CMS 20, 4x5 sheet film

[terrywyse]

Decided to start a new thread revolving around this new film I'm testing for a colleague of mine.

Film is Adox CMS 20 developed in Adotech II. Shot at f/11 @ ISO 20. I don't know what lens was used. 4x5 sheet was tray developed.

Film was wet-mounted and drum-scanned on a Screen 1045ai, maximum scanning resolution is 8,000ppi.

First image is of the 30 x 30mm section of the 4x5. This area had the best focus and well reveal texture in the blouse and sweater.

I'll be pasting links to screen shots from my dropbox....hope this works.



(Image insert didn't seem to work so I'm saving the image as an attachment)

Terry

[terrywyse]

Here's a series of scans at 1,000, 2,000, 4,000 and 8,000ppi. Also had scans at 3,000 and 6,000 but left them out for this comparison.

These are images viewed @ 100% without any re-sampling. There was no sharpening applied to these images either on the scanner or in Photoshop.



Terry

[terrywyse]

Same thing as previous...only with the view scaled to the same relative size...again, no re-sampling.



Here you can start to see the detail difference between the different scanning resolutions. There is a slight improvement in REAL detail @ 8,000ppi with no visible grain.

Terry

[AFairley]

Thanks, Terry.  Curious to see what the pattern is with the Epson scans.

[terrywyse]

Here's the last one. On this one I included the 3,000 and 6,000ppi scans. Here I took all the scans and re-sampled them to 10,000ppi in Photoshop so I could have an equivalent view % without pixelation. This, to me, represents the real detail available at these scanning resolutions and from this film/lens combo.

Starting from the left and moving right, the scan resolutions are 8,000, 6,000, 4,000, 3,000, 2,000 and 1,000.

Even at 8,000ppi, I'm seeing not just enlargement of grain (there's virtually no grain with this film) but I can see an increase in real detail despite the fact that some may say that this film scanned at 8K would have long out-resolved the lens. I don't know what lens was used but I don't believe it was anything special. I could ask if it's important.

This film is claimed to have a resolution of 800 lp/mm when processed in Adotech II....which equates to about 40,000ppi (!). In any case, it appears the scanner and camera lens were capable of extracting just a bit of extra detail at scanning resolutions beyond 4,000ppi.

At a scanning resolution of 8,000ppi, final print size @ 240ppi would be 11' x 14'.



Terry

[Chris_Brown]

That is remarkable film! If/when possible, please post an image that shows its full dynamic range.

[terrywyse]

At AFairley's request, here are three scans from my Epson V750 of roughly the same area. Scan rez was, left-to-right, 9600, 6400 and 3200ppi.

What I'm seeing is a slight improvement from 3200 to 6400 but none to 9600. In fact, I'd say there's a slight degradation (result of some interpolation?). I think someone said that 6400 was the maximum optical resolution so this makes sense.

Next up will be a comparison of the Epson 6400ppi scan and the drum scan @ 8,000ppi....I'll be upsampling them both so the viewing scale matches.



Terry

[terrywyse]

Alright, this should be the last one.....in this screen shot, I've got the 8000ppi drum scan flanked by the 9600ppi (left) and 6400ppi (right) scans from the Epson V750. These were all sampled up to 9600ppi so the view magnification would be the same. The view here is at 200%. Despite the slight lightness difference (I tried to quickly match the "density" of all three scans), it's pretty clear that the drum scan is holding a superior amount of detail, especially in the areas where the spot of black dye in the fabric meets the fibers.

I didn't take any particular pains with the Epson scan so I'm not saying this is the best it can do (I didn't take the time to wet-mount for example).....but one concern I had was focus....so I scanned the 4x5 sheet both emulsion-up and emulsion-down. Contrary to how you're supposed to mount these ("right-reading" means scanning with the emulsion-up, facing you), the focus was much better with the emulsion down. Gives you an idea how narrow the depth-of-field is on the Epson if the thickness of the emulsion plus some slight curl is enough to make-or-break detail resolution. I mean the detail was pretty much GONE when I scanned with the film flipped over. Scarey!

Anyway, drums rule the day as far as I'm concerned.



Terry

[Chris_Brown]

Anyway, drums rule the day as far as I'm concerned.

+1

Unfortunately, parts & service have become very rare.

[terrywyse]

+1

Unfortunately, parts & service are become very rare.

Tell me about it! My scanner has some issues with scanning reflective (won't calibrate) and I think it could use a going over. The techs that can still work on these are getting pretty scarce these days.

I picked up my scanner from a local printer that had literally shoved it out in the "boneyard" some two years prior to me finding it. Picked it up for $600 (they went for $65,000 back in the day I believe). It even survived the 150 mile trip in the back of my pickup truck. Had to build special ramps just to get it off the truck....thing weighs something like 350 lbs. and takes up about the same amount of space as my 24" HPz2100 printer.

So......I'll have fun with this thing as long as it lasts (or the old Mac driving it dies).....but until then, it's a blast seeing what this "antiquated" technology can do.

Terry

[Chris_Brown]

Tell me about it! My scanner has some issues with scanning reflective (won't calibrate) and I think it could use a going over. The techs that can still work on these are getting pretty scarce these days.

If you have problems or need to find tech support, this group is an excellent source of info. There's probably a few links to Screen tech papers, too.

....thing weighs something like 350 lbs. and takes up about the same amount of space as my 24" HPz2100 printer.

Yup. Those scanners were made to run 24/7 and were cash cows for color sep shops, back in the day.

. . . or the old Mac driving it dies . . .

I used an old PowerMac 8500 for about twelve years to drive a Howtek. Except when the hard drive failed (once), it was rock solid.

[Mark D Segal]

At AFairley's request, here are three scans from my Epson V750 of roughly the same area. Scan rez was, left-to-right, 9600, 6400 and 3200ppi.

What I'm seeing is a slight improvement from 3200 to 6400 but none to 9600. In fact, I'd say there's a slight degradation (result of some interpolation?). I think someone said that 6400 was the maximum optical resolution so this makes sense.

Next up will be a comparison of the Epson 6400ppi scan and the drum scan @ 8,000ppi....I'll be upsampling them both so the viewing scale matches.



Terry

Yes, that someone was me. Rather than "optical resolution", whatever that means, it is the resolution of the sensor according to ISO 14473 (the physical limit of the sensor array), and yes, if you select a higher resolution the scanning application is inventing information (up-sampling) which can have the same effect on apparent sharpness as other kinds of digital upsampling. Lasersoft Imaging has limited the ability of SilverFast to scan at more than twice the ISO standard resolution because beyond that degradation of image quality can become quite apparent. Astute observers may notice the beginning of deteriorating IQ before that two X point, depending on the image.

[Mark D Segal]

Alright, this should be the last one.....in this screen shot, I've got the 8000ppi drum scan flanked by the 9600ppi (left) and 6400ppi (right) scans from the Epson V750. These were all sampled up to 9600ppi so the view magnification would be the same. The view here is at 200%. Despite the slight lightness difference (I tried to quickly match the "density" of all three scans), it's pretty clear that the drum scan is holding a superior amount of detail, especially in the areas where the spot of black dye in the fabric meets the fibers.

I didn't take any particular pains with the Epson scan so I'm not saying this is the best it can do (I didn't take the time to wet-mount for example).....but one concern I had was focus....so I scanned the 4x5 sheet both emulsion-up and emulsion-down. Contrary to how you're supposed to mount these ("right-reading" means scanning with the emulsion-up, facing you), the focus was much better with the emulsion down. Gives you an idea how narrow the depth-of-field is on the Epson if the thickness of the emulsion plus some slight curl is enough to make-or-break detail resolution. I mean the detail was pretty much GONE when I scanned with the film flipped over. Scarey!

Anyway, drums rule the day as far as I'm concerned.



Terry

The way around that is to use the Fluid Mount Accessory, but instead of wetting the film, let a thin slice of Museum Glass hold it to the scanner bed. Doesn't cause Newton rings. I'm covering this approach in a forthcoming article.

[Ernst Dinkla]

Terry,

You can not expect that a wet mount drum scan done by a good operator can be challenged by a scan on a flatbed Epson. The point light source, the variable opening, the one or three PMTs, that whole sampling system is in a sense perfect for the task. Old and proven technology.

That said a flatbed Epson can be very satisfying, if I scan a B&W film wetmounted with PET film to the underside of  glass wet mount film holder. The emulsion facing the sensor, the holder tweaked for best focus at that spot, highest Sampling Per Inch set and selecting the green channel either in Vuescan or in ACR (semi RAW/DNG import) then add deconvolution sharpening (either in ACR or Photoshop) and reduce grain noise in Neat Image about 50%. Using a BetterScan holder BTW. There is a link on their pages to an article on wet mounting with Epsons that I must have written more than a decade ago.

BTW, there is a Digital B&W forum on LuLa now. Scanning analogue film should still fit that group.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots