filmscanners: Re: Emulsion flaws (was dust in SS4000)

[Roger Smith <rsmith@unb.ca>]

        Another interesting take on the "emulsion bubble" phenomenon. 
I recently compared a few rolls of Fuji Provia 100F and Kodak Elite 
Chrome 100. Each of the pairs of film was shot with the same lens and 
developed in the same 2-reel Nikkor tank. At the time I was helping 
Rob Geraghty (of film grain fame) with comparing various films for 
their appearance of grain upon scanning. He has long maintained that 
Fuji Provia 100F is almost grainless when scanned, unlike most other 
films, and I agreed.
        Upon re-examining some of these slides under the microscope, 
I now see a huge difference in the bubble appearance of the two 
films, as shown in the attached clip. The two frames are taken from 
the same relative area of the two slides - a portion of a white 
building.
        Not only is the Fuji slide sharper, it shows relatively few, 
large bubbles compared to the Kodak slide. The Fuji bubbles appear as 
a few fairly obvious spots on a scan - easy to spot out in Photoshop. 
The Kodak slide when scanned shows a gritty, grainy appearance which 
is very hard to clean up.
        I compared slides from three other rolls of Provia 100F and 
they all showed the same very low bubble count. I wonder how much 
this contributes to the "grainless" appearance of Provia 100F scans?
        I haven't been able to compare many other films, but slides 
from two separate Kodachrome 64 films showed a very high level of 
bubbles. Negatives of Kodak 100 Royal Gold also showed a high number 
of bubbles.

At 8:46 AM -0400 9/22/01, Arthur Entlich wrote:
>I am now more sure of my theory.  Either these bubbles occur during
>>  manufacturing or processing, and are, I suspect, either introduced in
>>  the base plastic or as the coatings are layered onto the film, or, the
>>  processing creates some gas which doesn't fully migrate out of the
>  > emulsion during washing and drying.

        I checked a piece of undeveloped Elite Chrome 100 and saw no 
bubbles, so they are definitely a product of film development. As far 
as I can tell, they are in the base plastic. A more obvious place 
would be between the emulsion and the base plastic, but there is a 
large difference in the focal point of the microscope between the 
film grain and the bubbles, so there appears to be a distinct gap 
between them. (At 400x depth of field is only a memory <g>). In fact, 
a piece of genuine dust on the surface was at almost the same plane 
of focus as the bubbles.

At 10:35 AM -0400 9/22/01, Owen P. Evans wrote:
>My theory is that the slides themselves not only attract dust
>electrostatically, but degrade and crack over time. These micro-cracks
>aren't seen when we project the slides but under the microscopic scrutiny of
>the scanner, become larger than life. Once the cracks open to the
>atmosphere, all sorts of things can get into the emulsion layers of the
>slide. Art & Roger are taking the manufacturers to task in their theories;
>I'm just accepting the chemistry & physics.
>Again, when we blow these slides up in Photoshop, they are magnified beyond
>the 8 x 10" or 13 x 19" prints that we make so the spots become exaggerated
>on the screen not in the prints. I've printed a lot of scans from the
>SS-4000 to an Epson 2000P printer in the past 10 months and I don't see the
>crud in the print; only on the CRT of the computer.

        Owen is correct in saying that these spots are much more 
visible on the screen than on prints - fortunately. In fact, they are 
usually only visible in smooth, light areas like clouds and skies. 
His other comment is only partly correct - slides may indeed 
deteriorate with time, but most of the slides I've been examining for 
this experiment were developed a couple of days ago. The bubbles, at 
least, are present immediately after developing.
        Does anyone know how effective ICE is against these flaws? 
The bubbles have a high refractive index but presumably are 
transparent to infra-red.
        The investigation continues....

Regards,
Roger Smith

Attachment: F&Kbubbles.jpg
Description: JPEG image