Getting rid of gremlins on scans

[Stew]

I'm using a Creo iQSmart scanner with a wet mount attachment for 9x12 and 6x7 to produce 100 megabyte scans that give a print size of 20"x24". I started last week, and it is my first experience with scanning. I'll be using this scanner quite a lot in the coming months, and alread I have some questions. Despite the oil, there are gremlins on the scans. I am using Photoshop to deal with these. The healing brush does a grand job on the spots. However, there are a few other miscreants, specifically:

(a) lines an inch or so long (when the scan is at 20x24 size) that usually have some curvature to them and that look like fine hair or string;
(b) somewhat thicker lines that occasionally appear along the border of the scan; and
(b) small round blotches on the occasional scan that are the result of bubbles caused by the oil mounting.

I am using the healing brush to deal with all of these. Should I be using some other Photoshop technique in some of these cases? Am I the only person who is not sure, at times, whether the "spot" on a wall is dust or part of the image and, if the latter, whether it should be left alone?

One of the things that has become obvious in this process is that Photoshop can be used to create an image that has very little to do with the original. If I want to get rid of the air conditioner in the flat above a butcher shop housed in an historic building, I can do it rather easily. Whether I should is a different matter. Does anyone have a view on the role, if any, of authenticity in photographic prints in the digital age?

[Kirk Gittings]

"Does anyone have a view on the role, if any, of authenticity in photographic prints in the digital age?" If documentation is an essential part of your art aesthetic, then the air conditioners stay. You need to work that kind of stuff into your composition. My personal work is more about "a sense of place" and leaves me a bit freer in that regard, but I have never removed anything more than a distant telephone pole in a broad landscape that stuck up over the horizon. Stuff I used to spot crudely anyway. Commercial work is much different of course.

You will find that scans pick up flaws in film that would never show up any other way. Fine scratches, lint that dried in the film etc. I scan with a 20x24 print in mind and spot everything at 100%. It is one of the wonders of this approach that one can so amazingly clean up an image so easily. The bubble thing is a pain in the butt. When I get that in a scan I reject it and make them do it over again. But cloning and healing are the primary tools. Do it at 100% or you will create allot of small visible artifacts that will show up in areas like smooth tones in skies.

[Roger Hein]

"If I want to get rid of the air conditioner in the flat above a butcher shop housed in an historic building, I can do it rather easily."

When I worked in a commercial studio in the 70's 'retouching' things 'in/out' of a photograph was occassionally done. Of course back then it was done using very sharp blades and an airbrush with a skilled hand. I think much will depend on your own convictions and what the photograph will be ultimately intended for - as a 'record'? fine art? or commercial?

[Janko Belaj]

Jon, it isn't easy to imagine those lines you are talking about. probably some dust, but who knows. If you want, you can send me an JPEG to examine what that is. If I wouldn't be able to resolve problem, probably some of people from Scitex division will be (flour under me is a in-house service for Scitex) and few doors away are few such scanners where are my friends working... (100MB file in JPEG medium q. would be a 2MB mail, 50MB file will have enough info to see what a problem might be and the m.q. compression will produce 1MB for mail what isn't much)
I'm not talking about finding way to get rid of those lines in Photoshop, rather to find a problem prior to scanning. We (firm where I'm working) are using several solutions for scanning on flat-bad scanners (we have several Scitex scanners and 2 Topaz scanners and 2 drum scanners...) For me, the best way isn't oil (doesn't remove dust that much) but some liquid solution. I'm sure there are several types on the market, I like "KAMI 2000 Filmreiniger". When I was at Heidelberg's demo center in Wiena several years ago, they used their own mixture of benzine (petrol) (I'm not sure if I'm using appropriate word) and alcohol. Both of the medicine purity. You wash your slide in a tray or simply pour solution over it, than you place slide (or negative) on the glass and you pour some more solution on it, than you place something I will call "clean thin scanning film" with the paper above it(!), with the small force you wedge (? displace?) liquid out of your glass-slide-foil-paper "sandwich", remove paper and you are ready for very clean scan. After such treatment I usually need no more than 5-10 minutes (on 300MB 4x5" scan) to remove dust with cloning tool.
Hope this helps... (ouch, my english... ;-))

[John_4185]

Jon: One of the things that has become obvious in this process is that Photoshop can be used to create an image that has very little to do with the original. If I want to get rid of the air conditioner in the flat above a butcher shop housed in an historic building, I can do it rather easily. Whether I should is a different matter. Does anyone have a view on the role, if any, of authenticity in photographic prints in the digital age?

It's entirely your call if you want to modify the content of the scene so that it doesn't reflect what was in front of the camera. If by the exceedingly slim chance someone wants a picture of that building as it looked at the time, and they have integrity, they will be looking for negatives or transparencies, not digital output.

[Bruce Watson]

There are three main sources for junk in scans. First is junk on the film at the time of exposure. It blocks the light from the film and leaves a clear spot on the film. Second is junk on the film surface from processing. Third is junk on the film and/or scanner from insufficient cleaning. There are others of course, but these are the main ones IMHO.

My favorite tool for spotting scans is the clone tool, working at 100% as Kirk says. I'm working with negatives only, and find most of my spots are white meaning they come from processing and/or scanning. I've got a cat in the house - need I say more?

I tend to remove all the spots I can see. I'll also remove things that are actually parts of the image if I think they are distracting. For example, when working with large landscapes, I often find that the sky contains birds - way off in the distance. On the prints these look like little black spots or smears. I clone them out. I've also been known to remove some distracting reflections, from water drops or mica reflections from rocks. I'll even clone out the occasional small twig that's in a distracting place that I didn't manage to find before making the photograph.

I do tell people that my work is *not* documentary regardless of what they might think. I emphasis that my job as an artist is to make the image say what I want it to say using whatever tools I find at my disposal.

To minimize spotting, you need every step of the process to be as clean as you can make it. I clean the walls and ceiling of the darkroom (can be a big source of dust on drying film) at least twice a year. I now use only distilled water for chemistry make up and washing. I now clean my scanner (and everything near it) right before I use it. I clean both sides of the film with film cleaner right before I mount it. I treat everything I can find with anti-static spray. I run air cleaners. I keep the relative humidity at least 50%. All this helps, but it's not a battle that's actually win-able. Sigh...

[paulr]

The only difference the digital age makes is that it's now easier to mess with an image. But it's always been possible. I remember heated debates about spot tone, and what the implications were if you used to not just to remove dust but to subdue a specular highlight, or remove a piece of litter.

How true you want to stay, or must stay, to what was there has always been a choice determined by the photographer's beliefs, not by the tools.

Personally, I'll use tools to diminish the specular highlight, but not to remove the piece of trash. That's what feels right to me. It has nothing to do with what I expect from another photographer.

[Ed Richards]

Get a good magnifier and look hard at the negative to make sure the stuff is not on the negative. Could be processing problems or the emulsion crazing. If the negative is clean, do a scan with no oil, emulsion down to minimize Newton's rings. If you do not see the stuff then, you know it is an artifact of the oil mounting so you can concentrate on that.

[George Stewart]

I too used the Creo for LF scanning and found that, although more work, oil mounting is the best. My recommendation is to not economize when it come to the oil. Be generous with it and use a roller to eliminate/reduce the bubbles. Are you using optical mylar on top of the films? I'd also recommend spending a little time with the film, ensuring that it is clean or as clean as possible.

It should also be pointed out that retouching LF film is exponentially more work than smaller film due to increased file size, hence slower computer response time.

[Stew]

Thanks everyone. It sounds like I'm on the right track at least. I clean the scanner glass before use, clean the film, use lots of scanner fluid and a roller and put a clear overlay (mylar?) over the film. I'm even using the occasional blast of compressed air from a tank, although I wonder whether that kicks up more dust than it removes. As for the dust-busting, I'm doing it with the image at 100%. As mentioned above, I'm concerned that my fixes will become noticeable if I do it at 50%. At 200%, I suspect that the time spent on this could increase significantly with no obvious increase in quality, perhaps even a decrease resulting from the additional messing around on suddenly visible "flaws".

As Kirk says, the scanner seems to pick up flaws that would otherwise be unnoticeable. It isn't that there are that many of them, it's just that it's a bit of an education to see them, and something of a marvel that they can be removed fairly easily. Being new to Photoshop, I wanted to make sure that there aren't techniques that I'm missing. I'll also have a close look at the transparencies and negatives to make sure that there aren't processing issues.

I rent time on the machine, and I can use either the Creo iQSmart or an Imacon 848. The rental rates are the same, but the iQSmart takes more time due to the wet mounting and a slower scanning rate. On the other hand, one can do a bunch of scans on the iQSmart in one shot, and I'm not sure if the same is true of the Imacon. In any event, people who are supposed to know about these things keep telling me that I'll get a better scan from the Creo. I may put a negative through both machines just to satisfy my curiosity.

As a result of the decision to scan and work with Photoshop, I'm also suddenly facing this issue of where to draw the line when it comes to retouching. It's one thing to know intellectually that Photoshop enables one to radically alter an image, it's another to see the potential implications for one's own images. For me, it's partly an aesthetic issue and partly a time issue. On the aesthetic side, the books I've seen on Photoshop all seem to assume that the objective of photography is to create some kind of perfect world. If Photoshop enables one to take 10 years off a person, or to remove a piece of debris on the pavement, then obviously it should be done. The other premise is that one should be willing to spend a good part of one's life making these "improvements". Of course, I suppose that there are worse problems than having to work through where to draw the line.

Thanks again for the responses.