scanning negative film as negative vs positive (and invert)

[taner]

Hi there

 

I should indicate that this is a follow up on question I asked in the

digital darkroom forum.

 

Do any of the popular films out there yield better results with

either the scan-as-negative or the scan-as-positive-and-invert method

REGARDLESS of the scanner software? Or is this a question that is

ALWAYS related to the software?

 

These are the films I use mostly: Sensia 100, Provia 100F, Superia X-

Tra 400, Reala, Kodak TMY.

 

And here is where I am coming from: I applied the two alternative

methods to the image below (and some degree of levels, curves, color

balance, resize, USM), and ended up with the images below.

 

<p><img

src="http://www3.sympatico.ca/askintaner/photonet_post/scanning-

dilemma.jpg"></p>

 

Your comments and suggestions will be appreciated.

[taner]

Sorry, I listed Sensia and Provia too in my list since those are the two films I use most frequently, but obviously, my question is with regard to negative films.

[the_macman]

It would be nice to see a scan of the negative and an inverted positive scan side-by-side

without any modifcation of levels, curves and color balance.

 

Anyhow, very interesting. I'll try this myself.

[bill_tuthill]

What software did you use, and how, A. Taner?

A simple invert should not work as well as these examples show,

because of orange mask. "Purple haze, all thru my brain" -J.Hendrix

[silent1]

This <i>should</i> be strictly a software question. However, the software in question might in fact be firmware in the scanner.<p>

 

The light coming through the film is the same for either path, and the CCD sensor that detects the light is the same as well. But some scanners do the inversion to scan direct to a positive image in firmware inside the scanner, and such an inversion will likely be inferior, overall, compared to one matched to a particular film's base -- because the base color and density do differ for different color negative films. If your scanner lets you select a film type, the inversion it makes will probably be better than you can do by hand, though you might still find you want to adjust color balance afterward (just as you might make color corrections in printing from a color negative). If your scanner (or software) treats all color negative films as being the same, you can likely do about as well for yourself, especially if you can save a color correction that will render the unexposed base as a pure, neutral black (rather than as a color with zero brightness -- probably easiest by making the base white before inverting values).

[sandro_de_gregorio]

Think Donald is right.

I scanned a test film (Agfa Vista 200) by AGFA Duoscan T1200 just like you did, but with no colour or levels adjustments. Duoscan allows you set the film type by a pre-scanning of the unexposed zone on the negative and let you save it. This yelds to better results than without prescanning but in any case to different results from scanning the negative and invert it taking in account the orange mask of each film by Photoshop.

This suggested me that the software of the scan interpretates in any case the result when scanning as positive. All this matters ended when I had to buy a new scanner that makes it very well and let me forget all this stuff :). (Epson Perfection 3200 if you're interested in).

[taner]

I scanned these images using my flatbed Epson 2400 scanner. Auto exposure, 48 bit, 2400dpi. The image on the right was inverted in PS 7. The resulting all bluish image was applied auto-levels. That resulted in a 'usuable' state, and finally, a little tweaking (the usual steps).

 

If either image is tweaked more than a little - the image becomes unusable. So it was impossible - or let me put it this way, it was beyond my PS skills to start with either image and make it similar to the other one without screwing up other parts of the image.

[sandro_de_gregorio]

One question: before inverting did you fusion the layer with the original image with the one of the scanned unexposed zone of the negative ?

That's the step required to neutralize the bluish colour (i.e.: the orange mask of the negative)

[taner]

No, Sandro, I did not.

[kenneth_beatty]

If you take the time to scan as positive, invert, then ajust individual red green and blue levels you get better results most of the time. The down side is it is very time consuming to do this the proper way.

 

So for speed scan as negative. For ultimate control scan as positive and tweak the photo yourself.

[pete_andrews10]

All I can add is that I've found most scanner software will clip the shadows too severely when left to do the inversion automatically. I've always been able to improve on the results by scanning as a raw 48 bit positive and then using the curves tool to negate the mask and correct the colour and gamma.

 

Incidentally, I believe that Photoshop applies the Gamma of the colour-space in use during the inversion process. It's not just a straightforward two's complement of the RGB data. (Reminder to self to investigate this further!)

[sandro_de_gregorio]

Here it goes Taner:

1) Scan the unexposed zone of the negative

2) By the eyedropper with an average of 3x3 pixeles select the resulting colour

3) Create a new layer with this colour, invert, select opacity at 50% and fusion mode to "Colour". Now you've neutralized the orange mask

4) To this layer add an adjustment layer and select "Invert". Now you've got a "weak" positive.

5) Add an adjustment layer of "Levels" and select Automatic correction

6) Finally, by another adjustment layer, modify Hue/Saturation by increasing saturation.

In this way I obtained the closest results to the positive-scanned image.

You'll tell me.

[michael gordon httpwww]

With b/w films, I am almost always scanning as positive and inverting in PS. Better results, for sure.

[taner]

Thank You all very much

[biotime_biotime]

Using A Minolta Multi Pro soft V 1.1.4 I am not getting good results pos with invert. I also tried 16 linear with somewhat similar results. After inversion weather I auto level or manually adjust levels I get a flat greenish uncorrectable image. In particular neutral midtones are totally missing magenta ending up light jade. This is true amoung various neg films. I wonder however could there be an related issue as my Multi adds a significant magenta bias scanning chromes? Could this be robbing me of my midtone neutral magenta after inversion? The scanner recently came back from service with new lamps, cal, etc.

 

Do I need to do something different than just inverting and adjusting levels?