Teeny Tiny Little Lines

Sean Michael Robinson:

Greetings everyone,

This will be a brief one, as I'm still trying to catch up on Church & State work, while making progress on the CAN3 files, namely, getting Dave's remarkable notes for the project illustrated and laid out.

In light of the recent discussion of the photo negatives and how well that end of the project is going, I thought it might be instructive to do a bit of comparison and reiterate in what ways the project benefits from access to original art. As I've mentioned a few times, it would be an easy thing to batch-convert the negative scans and create a new digital "negative" virtually identical (functionally, anyway) to the original negatives. 

Putting aside pages that were underexposed initially, or where the negative has been damaged in some way-- how does a well-photographed neg compare to a high-res scan of the original art for the page? Is it really worth the hassle of dealing with the original art, of having to seek it out, correct tone shrinkage and other problems?

Here's a rare example of a page where I have an ideal scan of both original art for the page and negative. The original art in question was (most likely) scanned by Scott Dunbier as part of the work on the upcoming Cerebus Covers book from IDW, and passed on to me by Justin Eisinger. The negative was expertly scanned by Funkmistress Karen last week.

Other than a little bit of shrinkage of the border tape and a little bit of tone shrinkage on the left leg that will need to be repaired, this is about as perfect as you get. The page was in great shape, the scan was clean and easy to work with, with really sharp optics (Epson Expression 10000XL? I'm guessing here, but I wouldn't be surprised if I'm correct. Truly a "Cadillac" flat-bed scanner).

Okay, so how does that compare to a truly excellent scan of the original negative?

Take a look at the skew of the panels in the top tier-- the sides are still straight, but the tops and bottoms of the page have been skewed upwards to the right. This is most likely a result of the art being mounted at an angle to the camera during the initial photography. Strangely enough, it even makes Cerebus' expression in the top tier seem a bit different to me.

Perhaps more noteworthy are all of the missing lines.

Less visible in this particular close-up is the amount of plugging on the fine detail information, i.e. white space between black lines. As this plugging happens normally during the print process, it's important to preserve as much of this as possible before you get to that stage.

This might seem like fairly minor losses here-- but how much linework is actually absent from the page really depends on how much fine-linework was there in the first place. What got lost in photography? Very fine lines. Lines drawn with watery ink. Lines drawn with technical pens. Pages with tons of fine-line information had, no surprise, more to lose, whereas a page like this only really has fine hatching and feathering in a few places.

Last noteworthy difference, more interesting than good or bad-- the tiny Cerebus figure in the lower right hand panel has has his exposure "spiked", along with the rest of the panel, to turn his tone from a 25 percent to closer to a 20 percent. This was no doubt intended to combat the inevitable dot gain from the printing process, exacerbated by the newsprint the book was printed on, and by the desire for rich black on the part of everyone involved. (The unintended consequence of this-- a general thinning of the lines in that panel, and blowing out some of the more watery lines.)

Before I leave, one more plea for feedback--what would you like to see (or not see) in this space? Any suggestions welcome.

Church & State I: Progress Report

Sean Michael Robinson:

Okay, so my first mission here is to displace any and all medical photos from the top of the blog. Sheesh, Dave!

This week I thought I'd give you all a brief status report on the Church & State I restoration.

On Monday I completed my portion of the work on all of the pages sourced from original art. This includes 150 pages of original art scanned by Dave from the Cerebus Archive, plus an additional 18 pages contributed by Cerebus art collectors-- thank you very much Oystein Sorenson, Gregory Kessler, Dean Reeves, Dagon James, James Guarnotta, Jean-Paul Gabilliet, Nat Gertler, ComicLink, and Justin Eisinger of IDW! This does not include the twenty pages of "Mind Game III/IV," which will be reconstructed from a combination of original art and negative scans, the original photocopied elements having become dirty and speckled with age.

All of this means that yesterday I got to start into working with the negatives.

Last week, after Funkmaster John and I finished working out the routine they'd be using to scan the negs, the actual scanning was started by Funkmistress Karen. FMK started sending me scans, and the three of us tackled a few other outstanding issues.

Namely-- grit, "schmutz" and other Mystery Ailments on the surface of the negative.

As I've mentioned before, these are negatives, meaning, anything that appears black in the finished product is actually the clear acetate. Anything that appears white/paper-colored in the finished product is actually film emulsion on the clear acetate.

So-- here's some grime for you--

This is the raw scan, pre-adjustment. Most of these flaws end up significantly worse post-adjustment, as the sharpening and contrast adjustment bring out these "details" as much as the wanted ones.

That being said, this isn't particularly bad, as it's mostly confined to large black (currently visible as white) areas with little detail underneath, so it could probably be fixed relatively easily by Mara (or myself) in the cleanup stage, using the burn tool.

Here's a better example of when the schmutz is a real problem-- when it's sitting on top of detail or tone.

I sent John and FMKaren an email asking-- what does this look like in person? Like a milky fluid on the surface of the negative, came back the reply. What to do?

As always, the answer is improving the source and the initial capture as best as possible.

Kodak had this bit of wisdom for us regarding cleaning photographic materials-- 98 percent pure isopropyl alcohol.

We recommend isopropyl alcohol that has a purity of 98 percent or higher as a good, general-purpose cleaning solvent for photographic materials. Isopropyl alcohol (also known as 2-propanol or isopropanol) has several benefits. It is available in small volumes at a reasonable price; it has been successful in cleaning tar, streaks, processing scum, and opaque from photographic products; and it had no detrimental effect on the image stability of the emulsions we tested.
Use only isopropyl alcohol that has a purity of 98 percent or higher. Alcohol with a lower purity, such as rubbing alcohol, will cause streaking and take longer to dry. Also, the higher water content of rubbing alcohol may cause the emulsion of the photographic materials to swell, resulting in physical damage and possible deterioration of image-forming dyes.

And what do you know, it worked like a charm. There is a certain type of blemish that's resistant, and there are persisting problems with other wear, namely rips that have been repaired with clear tape-- but everything else has been smooth sailing from here.  How smooth? I prepared 47 negatives yesterday, and will likely be faster at it next week.

All very good news, seeing how many pages we have to go!

Silverfast Negatives

Sean Michael Robinson:

Greetings everyone,

Up until this week, all of the work on the Church & State I restoration project has been on the original art still in possession of the Cerebus archive-- a few hundred pages, all told. But the past two weeks we've started into the next stage of the work, namely, scanning and adjusting the several hundred pages represented by photographic elements.

These are at-size negatives, shot from the original artwork while the monthly books were being produced, and ganged up eight to a flat for use in creating the plates for printing the book. 

When I was contacting prepress professionals of my acquaintance at the start of this project, several people tried to warn me off from scanning the negatives at all. Too difficult, they told me, not worth the pain. One person actually suggested finding a facility that could make oversized Iris prints from the negs, and then scan those prints! 

Why are these materials such a pain?

Well, anyone who's ever tried to scan conventional film format negs or positives knows that shooting light through them with a scanner is simply not the same as shooting light through them on a projector for development. Trying to recreate a natural exposure level, having your light source so close to the carrier film. And unlike dark room projectors, most scanners don't have the same kind of physical separation between the light source and the "projected image."

Here's a scan of a Church & State neg on a conventional flatbed scanner.

Do you see the sort of "shadow" under the tone and near the border of the blacks? Here's a close-up.

Several months ago I was referring to this as a "negative shadow", but if you see it in reverse, how the negative actually appears, it might make a little more sense what's happening. I suppose it's more of a negative halo.

So what's happening here? 

This is one of the hazards of working with materials I'm not scanning myself-- I'm not totally sure what it is that causes the phenomenon. But my prior darkroom experience tells me that it's most likely this-- the scanner is picking up the halo of the light escaping through the edge of the emulsion. In other words, the light passing through the carrier (clear) portion of the negative is bleeding through the edge of the emulsion (black) portions of the negative, causing a halo around every element. 

As you might be able to imagine, this is a tremendous pain to deal with, and causes a lot of additional work and guesswork to make the image reproduce accurately. Want to shave off the halo with a curves adjustment? Careful you don't shear off any of your fine detail! Want to use sharpening to bring out a maximum amount of detail? Careful the sharpening doesn't grab the neg halo! Pain, pain, pain....

The solution?

Here's a scan of the same neg, using the Epson V700 and Silverfast, an amazing little piece of software from LaserSoft Imaging.

                              

        

So, where did that halo go?

It's hard to believe it, but the majority of the above difference is due solely to the software.

Simply put, Silverfast enables you to select the type of material you're trying to scan, and then applies a predetermined density curve at the point of capture, adjusting the exposure to adjust for the optical density of the film.

As I'm typing this, John is scanning negs using Silverfast and sending them via Google Drive to my desktop, while Dave Fisher and John's nephew Rolly cut new negs off of the flats and store them in folders for safe-keeping. It's an amazing mixture of high-tech and low-tech, of multiple generations, and an exciting time to be on a project like this.

So, TLDR-- I would never think of scanning a negative without Silverfast! Thank you, LaserSoft Imaging, for such an incredible product. We might have done it without you, but it would have been a heck of a lot more painful.