Post your negative processing actions

I'm trying to scan negatives, and I'm having a hard time progressing beyond "poke about with photoshop curves till it looks less crap."

I use Epson Scan and a Epson 4990 scanner. I'd like to take out as much 'art' as possible and replace it with easy to get right procedure.

If I were to start with a scan of a negative with the scanner set to "positive film", what should my processing steps be?

What is a good curve that mimics printing a negative onto paper?

what is a curve adjustment that mimics adjusting the magenta-yellow contrast filters and when would you apply it?

What adjustment mimics keeping the enlarger on for longer?

What is a good curve that adjusts for under-exposure and when would you apply it?

What is a good curve that adjusts for over and under-development and whne should it be applied?

Are there curve adjustments they neutralise the response of particular films?

I am aiming towards a pre-set action that sets up curves that I can tweak. I've attached a somewhat lame attempt to show what I am asking for.

Post your workflow and corresponding actions and pre-sets.

Hi

if you look at the range of exposure in my histograms on the first part of this page I think it answers some of your questions. Essentially its nice to know what and where the histogram for clear film base is and what is the densest you can get with your negative.

This will then let you know when you scan it if you have a good density range or contrast range in your exposure.

From then I usually invert and then make the curves dependent on the image / scene contrast range. I recently cross checked this with some prints done at different grades of filter ... sadly I didn't get around to writing this up yet ... and wonder if any are interested anyway ;-)

pellicle said:

Hi

if you look at the range of exposure in my histograms on the first part of this page I think it answers some of your questions. Essentially its nice to know what and where the histogram for clear film base is and what is the densest you can get with your negative.

This will then let you know when you scan it if you have a good density range or contrast range in your exposure.

From then I usually invert and then make the curves dependent on the image / scene contrast range. I recently cross checked this with some prints done at different grades of filter ... sadly I didn't get around to writing this up yet ... and wonder if any are interested anyway ;-)

Click to expand...

Your suggestion to use a digicam is quite a good idea. I often have a digicam with me, and for at least some photos having an identical digicam photo will help me make a good set of curves.

Buying a standard negative might be a good idea too.

I have been searching a bit, and found this:
http://www.marginalsoftware.com/HowtoScan/tutorial_page_1.htm
which looks worth reading in detail.

So far I think I'm settling on scanning as if the negative were a positive, I'm going to set the levels in the scanner to be from the film base to the maximum highlight in the whole roll or sheet, and then adjust for exposure and development errors as a final step in software.

I'll use a well exposed and well developed negative taken alongside a digicam photo to make a paper response curve, that I will attempt to use as an initial first step.

I hope we have others posting their workflows and actions to this thread.

Thanks,

g'day fellas

admittedly i don't have very much, read virtually none, experience scanning film, but from my 20 years darkroom experience and my 5 years digital experience i'm thinking maybe your trying to make it too complicated/technical/formulaic.

at some point you have to create an image/print that looks good to your eye.

surely all you need to do in either analogue or digital imaging is record enough information so that you have detail with which to create the image you perceive as "good"

Ray Heath said:

g'day fellas

admittedly i don't have very much, read virtually none, experience scanning film, but from my 20 years darkroom experience and my 5 years digital experience i'm thinking maybe your trying to make it too complicated/technical/formulaic.

at some point you have to create an image/print that looks good to your eye.

surely all you need to do in either analogue or digital imaging is record enough information so that you have detail with which to create the image you perceive as "good"

Click to expand...

I agree with the sentiment, and the last step will always be a 'make it look good' step but with with scanning the first image out of the scanner looks really crap. After some poking about I can make it look okish, and with a lot more I can make it look good.

From experience, the more you have to fiddle with the colour of a photo, the less likely it is going to look good in the end, and so I am looking for ways of reaching close to good before I start fiddling.

A lot of what makes thing look good or bad is the contrast at a particular light level. The goal is to get the contrasts within a fixed range at all levels and make the curve that gets you there as smooth as possible

I'm striving toward this same goal, but with a much lower experience level, both from a darkroom and a digital post processing perspective. Like Ray, I tend to go for what delivers the most pleasing results, but one thing I've noticed after several months now of scanning slides and negs -- both color and b&w -- is that as my skill level improves, I often have to go back and redo images. This is almost always because I've overdone something in the processing, rather than underdone it. I have concluded that it just takes a certain level of experience to get to the point where I can accurately render images in scans.

I've found negs to be more difficult, generally, to get good scans of, because if they're underexposed even slightly, the shadow detail blocks up and noise, i.e. grain, is usually much more predominant.

I've used the Epson scan software a lot, though, and I've gotten to where I mostly use the gray eyedropper tool in the histogram window to establish both the most optimal color and density/contrast in an image. I might tweak the edges of the histogram after that, if it looks like it needs additional contrast adjustment. I have also realized that the more image adjustment I can do prior to the scan, usually the better the result will be. This is similar to the results I've found when adjusting digital images. There seems to be greater adjustment latitude working with the raw digital images before they've been converted to .tif or whatever than afterward. As for the Epson software, I haven't seen much point in setting up custom configurations or templates or whatever because it seems like just about every image poses its own unique set of problems or challenges.

Best,

Michael

Hi Ray

Ray Heath said:

i'm thinking maybe your trying to make it too complicated/technical/formulaic.

at some point you have to create an image/print that looks good to your eye.

Click to expand...

exactly .... the issues only come when you wonder why it is that you can't get on your scans what you see when inspect a negative. Its really only then that you need to examine what it is that the scanner is doing to / for you.

but yes, over complication of things like curves and stuff is a trap. All too often what is simple gets bogged down in the explanations ... try explaining how to play a G chord on your guitar using text only.