An interesting PDF on film scanners by Barbara Flueckiger

[Daniel D. Teoli Jr.]

Here is an interesting PDF on film scanners by Barbara Flueckiger and some others. 

 

Investigation Of Film Material Scanner Interaction B. Flueckiger D. D. Teoli Jr. A. C. : D.D.Teoli Jr. A.C. : Free Download, Borrow, and Streaming : Internet Archive

 

Barbara is pretty knowledgeable with vintage cine' film. She is a Professor in Zurich.

 

Tinted film base / Kodak Sonochrome | Timeline of Historical Film Colors

 

I had written to her years ago with some questions about Kodak's tinted Sonochrome film. Never got a reply.  

[Perry Paolantonio]

Interesting and deeply flawed. Most of what's in there is techically and factually incorrect and could have been proven as incorrect by someone who knew how to use the scanners they use. I contacted Barbara when this came out and sent a 2 page long list of easily verifiable errors. Some were fixed in a later revision, others were not. 

The biggest one is that thing about the tinted film, and how some scanners can't "see" the dyes. This is completely false. The reason those scans eliminated the color wasn't because the scanner couldn't see it, it was because the operator base calibrated the film, effectively wiping out the color. You can tell because the perfs are pink. They should always be white on a positive scan. If they're pink that means the color of the light (or some color correction) was done that eliminated the blue. This is purely operator error. They even told me privately that they knew some of the operators weren't experienced scanning this kind of film. 

Every machine they used was scanned by an operator/owner. They should have all be scanned by the manufacturers, who know best how to handle each kind of film. Instead, they left open a huge variable that invalidates all of their findings. This paper is garbage. 

Just because someone is a professor doesn't mean they know how to do everything. Frankly, they should have retracted this paper because it has so many problems. 

[Daniel D. Teoli Jr.]

Well Perry, you know many of these scanner companies charge for every little thing. And she probably had no budget for the scans. And maybe she got tired of waiting for them to never answer her emails?

As far as tinted film. Why would you have to adjust the color and not scan in color? Why second guess things? 

[Perry Paolantonio]

5 minutes ago, Daniel D. Teoli Jr. said:

Well Perry, you know many of these scanner companies charge for every little thing. And she probably had no budget for the scans. And maybe she got tired of waiting for them to never answer her emails?

If you're writing an academic paper that purports to be objective, it is your responsibility to make sure things are tested in the fairest way possible. Otherwise, expect to be torn apart. That's what peer-review is. I may not be an academic, but I would certainly be considered a peer, at least as far as the scanning part goes. What they did was sloppy. Their testing methodology was flawed, and there were pages of factual errors in the first few drafts (upon which conclusions were drawn), that could have been verified with a simple visit to the manufacturer's web site to look at the spec sheets. 

 

14 minutes ago, Daniel D. Teoli Jr. said:

As far as tinted film. Why would you have to adjust the color and not scan in color? Why second guess things? 

You're not understanding what they did.

When you scan color negative film, which has an orange cast to it, part of the setup process is that you do a base calibration. This is the case on any scanner. Think of it (in part) as a kind of white balance. The scanner looks at some unexposed film (between frames, or between perfs usually), and it sets that dark area to dmin at a specified level in the output file. Part of what it's doing is adjusting the brightness and color balance of the light so that the level is where it needs to be and so that the black is black. This also gets rid of the color, which is why you can see the color temp of the light in a scanner change, depending on the kind of film you're using. 

But if you do this on a tinted positive film, it's going to neutralize the blue. A side effect of this is that the white in the perfs will turn a color, in this case, pink. The perfs should always be white if you're scanning a positive image - the color of the light source because there is no film where the perfs are. If they're not, then there is some color correction happening that typically shouldn't be happening in the scan except in specific situations (not this one). 

Anyone who has operated a scanner on anything other than B/W positive film should know that. But the archives they worked with pretty much only dealt with B/W pos. And they screwed up.

The paper's authors also should have known this, since they have close ties to Arri. Instead, they came up with a cockamamie theory that the spectral value of the blue die in the tinted films was somehow invisible to the scanner and ran with it. (And yet, you can clearly see the same blue in a photo of the film on a light box taken with a $5 smartphone camera? Give me a break). 

It's sloppy, lazy, "research."

[Daniel D. Teoli Jr.]

OK, thanks for the explanation.

[Daniel D. Teoli Jr.]

In a case like this Perry, Barabra's paper was the foundation and should have been built upon and not destroyed.  If the corrections were not made, you should have posted your own rebuttal to correct what needed correcting. But maybe that would cause too much friction? Although you could have reviewed it is such a way to bring out good points and add the corrections that were wrong. 

That is something I like about Robert H's replies. He is to the point and does not make things personal and leaves the digs out. That is why I say he is more levelheaded than you in his replies. In archival work we preserve history and don't destroy it Perry, whether that history is imperfect or not.

 

 

I just looked over the paper quickly, but I didn't see she mentioned / thanked you in it for the revisions. Maybe she did and I just didn't see it. I'm not much for text, I like images.

Credit should always be given so you know what / who was involved in the development. Although in my case, I always ask if people want credit for help. A lot of people don't want to be associated with me and come up in a Google search. It is OK with me if they don't, but I still give them the courtesy of asking.

Edited March 1, 2022 by Daniel D. Teoli Jr.