"Brightening" of processed Ektachromes

greenspun.com : LUSENET : Large format photography : One Thread

Well, it's not often that I expose color film, but once in a great while I do and over the years I've collected a handful of slightly underexposed EPN's. A friend showed me a kit he bought from Edwal that was designed to brighten underexposed Kodachrome. He said they also made a kit for Ektachrome, but I have not been able to find info about that on Edwal's web site. I don't need more that about a half a stop of "brightening" in most instances. Any of you chrome experts out there have any ideas? Is there some secret supplier who sells this stuff? Is this one more environmental "worst nightmare" that has been banned? Your input will be appreciated.

-- Robert A. Zeichner (razeichner@ameritech.net), September 24, 2000


You could scan and modify electronically. It's easy to "undo" mistakes on the computer.

-- Larry Huppert (Larry.Huppert@mail.com), September 24, 2000.

Another vote here for scanning and, possibly, re-outputting to film.
The 'reducers' for colour slide are a bit hit-or-miss. They consist of bleaches for the three CMY dye layers individually, so the mix of chemicals is pretty critical. I know that one of the chemicals used is common household bleach; sodium perchlorate; for the magenta layer if memory serves. The other two aren't particularly exotic as I recall, but getting the mix right needs a lot of trial and error experimentation.

-- Pete Andrews (p.l.andrews@bham.ac.uk), September 25, 2000.

Ed Degginger invented the Kodachrome formula that lightens underexposed chromes. He spent his career as a chemist and was always an avid 35mm photographer (lots of wildlife, though much general stock too). It is/was a two-bath system that 1)bleaches and 2)stops the bleaching. He sold it to Edwal and now sells stock out of Morristown, NJ. I found that it always made chromes magenta. It hit successive layers of dyes in the film, the magenta being the last. You had to stop the bleaching fast to save it, but that left a preponderance of magenta and increased contrast. Some of my attempts were quite good, others not so good. I heard that Edwal came out with a similar chemical for E-6 but I don't know this to be true. Digital manipulation would seem to be the way to go these days. But Edwal is part of Falcon Safety Products in NJ, and they probably have a Web site and do have an 800 number. They are helpful and responsive.

-- Rob Tucher (rtphotodoc@juno.com), September 25, 2000.

Modifying electronically an underexposed slide means that you have to use a very high performance scanner with a D-Max of 4.2 or more. Otherwise the image you will get is likely to be lacking depth in the shadows. I also have a bunch of nice but underexposed slides (mostly Velvia). My scanner is not capable of reaching into the darkest areas to restitute enough details. I wondered if the best way to salvage them would not be making duplicates on low contrast film? A friend of mine showed me how he pulled color from a nearly black Kodachrome duplicating it with a sandwiched contrast mask. I was amazed! I used to make the masks for lowering the contrast of Ilfochrome prints, with Kodak Pan Masking, but I am not familiar with duplicating slides.

-- Paul Schilliger (pschilliger@smile.ch), September 25, 2000.

That's a bit nonsensical Paul. The Dmax of both Kodachrome and Velvia are ~3.7D, underexposed or not, so why would a scanner need to go to 4.2D to scan them?
Anyway, the question was about Ektachrome, with a modest Dmax of around 3.3D.

-- Pete Andrews (p.l.andrews@bham.ac.uk), September 26, 2000.

Deviating a bit into scanner-land...

Pete Andrews' point (that Dmax for EPN is fairly low) is technically correct, but from a practical standpoint I think that Paul Schillinger's advice is still good: If you want to lighten, you would do well to find a scanner with the widest dynamic range possible.

Although many scanners have rated Dmax well in excess of 3.3, very few can actually extractl noise-free detail from something that dark. My experience is that scanner Dmax ratings tend to be inflated by 0.5-1.0 units of density relative to what they can actually handle witho0ut perceptible noise. If you really want to fetch all of the shadow detail possible out of a piece of film with a Dmax of 3.3 or 3.76, then a scanner with a rated Dmax of 4.2 would probably be the appropriate tool for the job.

-- Patrick Chase (patrick@sdd.hp.com), September 26, 2000.

Pete, thanks for correcting me, although I think you are right and you are wrong. I'm not a specialist but I have soon discovered the limitations of the electronic tools I'm using. Often scanner constructors have a questionable way of putting up the specifications for their products. For one it would be that the scanner is capable of producing sharp details and differentiated colors into the D-Max and for others that they will be able to catch some light and produce a muddy gray in the same density. Both would boast a D-Max of let's say, 4.0 but when you compare the resulting scans from a scanner costing 3 K and from one costing 30 K, you will soon discover what makes the difference. I am personally convinced that it is not necessary to spend a fortune to get a good scanner, but then there are some limitations one has to put up with. Another precision is about density range, which is not to be mixed up with D-Max. A correctly exposed slide has a density range of 3.3 or 3.7, which is the density range most good scanners can analyze. But when a slide is underexposed, the density range can be the same, but the overall and Maximum Density are considerably augmented and this is where you have to use a "beast" to get the colored pixels out of the film. Otherwise, the dark areas will just be "dull black".

-- Paul Schilliger (pschilliger@smile.ch), September 26, 2000.

Yes. It's a complicated issue Paul, and I agree that a lot of low-end scanners have ideas above their station when it comes to dynamic range.
The ability to pull shades of dark grey up to mid tones is more a function of a scanner's bit depth and the programmability of its 'gamma' curve, as well as its inherent noise, rather than the actual Dmax that it can 'see' down to.
Most scanners could, in fact, see a higher density, simply by winding up the intensity of the lamp; but if it was that easy, then we'd have even more ridiculous claims from scanner manufacturers than we have already.

The problem's compounded by the fact that the film's RGB curves flatten out and separate widely after 3.0D or so, and this is another good reason to go the scanner route, as this can be partially corrected in sofware, but not by chemical bleaching.
Ektachrome appears to have less separation than Kodachrome or Velvia from its characteristic curves, but in real-life it always seems to get the blues in the shadow areas.

-- Pete Andrews (p.l.andrews@bham.ac.uk), September 27, 2000.

Pete, to illustrate what we are talking about, I have two scanners. One is a 7 years old drum scanner, dynamic range and D-Max are 3.6, 12 bits color. The other one is a recent flatbed boasting 3.7 dynamic range, 4.2 D-Max and 16 bits color depth. This is not a very expensive scanner (a little more than 3K) but you would think it is better than the previous. Not at all! I made comparison scans and the drum has far better picture quality in the dense areas despite it's modest specs. The flatbed is quite comparable when it stays within well exposed films, with perhaps a little more USM needed to achieve the same "optical" sharpness. It's 16 bits differenciates millions of colors but as they will be compressed to the 8 bits of Photoshop, you will hardly see the difference from the 12 bits scanner anyway.

But, I agree, scanning a dark slide is certainly the best an easiest way to correct it.

-- Paul Schilliger (pschilliger@smile.ch), September 27, 2000.

Pete Andrews wrote:

> Most scanners could, in fact, see a higher density, simply by winding up the intensity of the lamp

This is untrue for most scanners, unless you're willing to allow the CCD's analog output to saturate in mid- and light-tones. If you turn up the lamp, then you'll simply end up shortening the integration time (the amount of time between transfer gate signals) to stay under saturation. Provided that the integration time was reasonably short to begin with, this puts you right back where you started in terms of maximum density.

-- Patrick Chase (patrick@sdd.hp.com), September 27, 2000.

What I said about 'turning up the wick' is absolutely true Patrick. I didn't say it would increase the dynamic range, or that it wouldn't have any other deleterious effect. In fact, I went on to say that it wasn't that simple.
Why do you assume that I was only referring to CCD scanners? PM based drum scanners don't saturate easily, and some of them actually use lamp intensity to control exposure.

Paul, the bit-depth of a scanner is more important for maintaining a linear relationship between density/brightness and the digital output, than in capturing more colours. Greater bit depth should also allow more flexibility in mapping the tone curve to an 8 bit output, but if the scanner is CCD based then the limitation is probably with the sensors, rather than the rest of the circuitry. Most CCDs have a dynamic range of 4 or 5 thousand to 1, about 3.6D in terms of density, or 12 bits in digital terms. Feeding this into a 16 bit A/D won't improve this basic limitation, and that's why your high-end flatbed still can't beat your mid-range drum scanner.
The real bottleneck these days is that damned 8 bit per channel output. I think if we can make 16 bit the norm, then desktop scanner makers will be prodded into doing some catching up.
Anyway this is all way off topic. Sorry!

-- Pete Andrews (p.l.andrews@bham.ac.uk), September 28, 2000.

"I think if we can make 16 bit the norm, then desktop scanner makers will be prodded into doing some catching up." Pete, not long ago, this would have seemed a ridiculous expense of megabytes. But now, with the super-computers being brought to the desktop at a low price, new huge and cheap harddisks (Ultra ATA 66+) and affordable memory, this norm, already partly supported by Photoshop and some scanners is at hand. Still, output machines and monitors have to be updated.

-- Paul Schilliger (pschilliger@smile.ch), September 28, 2000.

Moderation questions? read the FAQ