Flextight workflow; TIFF scan utility confusion

Flatbed and Film Scanners
1

Hello, everyone. I know NLP is primarily a converter for those who ‘scan’ their film negatives from a digital SLR/camera, but I am seeking help on improving my workflow and/or settings, and to anyone else who uses (traditional) scanners.

I shoot medium format (6x7), almost exclusively on Kodak Portra 400. I scan on a Flextight X1, in FlexColor software (4.8.13 on macOS), as 3F raw scans. (Meaning they’re 16-bit, no settings applied, and are scanned as positives since I want NLP to do the negative conversion.) I then rename the extension to .tif from .fff (since the files are wrapped TIFF’s).

I bring the files into Lightroom, and have had a difficult time with conversions in regards to color — especially with skin tones.

I’ve experimented using two methods:

  1. A straight conversion, where I use the white balance picker on the film border, crop said borders, then convert (Control + N). These conversions are never quite to my liking; the colors always appear unnatural, even after adjusting color settings.

  2. A conversion using the TIFF Scan Prep tool (File > Plug-in Extras). I take my file, and run the tool without doing any white balancing or cropping. I then convert using this newly-created file. However, this tool is a bit of a mystery to me. I’ve experimented with many of the Gamma settings: Linear (1.0), Flextight Mac (1.8), and Custom Input Gamma of 2.2.

Some settings are better than others, but I can’t get good results with any of them. (I would have assumed the Flextight Mac option would be the way to go, but I’m not sure I’m using it appropriately.)

I’ve included outputs of all the methods below as reference, including one from a while back that I did with Epson Scan. (Please don’t suggest I go and use Epson Scan. It won’t work with the Flextight, and I want to make NLP work).

All methods/conversions use the following settings:

  • Source: TIFF Scan
  • Color Model: Frontier
  • Pre-Saturation: 3 - Default
  • Border Buffer: 10%
  • Tones: Linear + Gamma
  • Brightness, Contrast, Lights, Darks, Whites, and Blacks are all at 0.
  • Soft Highs: Checked
  • Soft Lows: Unchecked
  • WhiteClip: -2
  • BlackClip: -3
  • Color (on the film tab) is set to AutoColor 2.0 Warming.
  • Every other color setting is at 0.

Scan comparisons
Scan comparisons4679×2847 2.73 MB

As you can see, the color is all over the place, none quite right - especially when viewing the skin tones (which seem quite depleted in reds). The standard conversion is too cyan. Linear really depletes the reds (her lips especially). Flextight Mac is too yellow. Custom also has red desaturation issues.

I know I could tinker with color settings to get to a decent place, but all of these feel like such a departure as a starting point. The Epson Scan - though a tad bit on the magenta side - is at least a good starting point where adjustments are kept to minimum. I feel like I must be doing something wrong. Is the above the correct method? Anything I’m doing incorrect or can be improved upon?

Thanks in advance!

(Output images converted to sRGB. I’m using Negative Lab Pro 2.1.2, in Lightroom Classic 9.4, on macOS Catalina 10.15.6.)

2

Hey @jessyel !

Thanks for the detailed write up and example!

Since you are working with a TIFF here, I would recommend NOT white balancing off the film border. Just use Negative Lab Pro directly on the cropped tiff file. (The white balance thing is ONLY for true RAW files). That’s probably what is causing some of those weird cyan tones that are hard to get in check. Try that and try adding some red to the midtones and you should be close!

(Also, I’d love to experiment with more Flextight files to find ways to improve that workflow. I can also see if I can find a better starting point for this image - just email me at nate@natephotographic.com).

Re: the Tiff Prep Utility - I would still consider it somewhat experimental and it breaks down in some scenarios. I haven’t shown it that much love because I’m not sure it is the best approach… with the upcoming v2.2, there is loads more accuracy and nuance to the controls, so the Tiff Prep Utility is probably only necessary in extreme circumstances (like if there are so few levels of blue that there isn’t anything for NLP to convert in that channel).

Cheers!
-Nate

1 Like
3

Goodness, do I feel like a dope. I had assumed to always white balance the border, not realizing only to do so on negatives captured by a digital camera. After converting without doing so, everything looks amazing. See below. Some good news for a change — thanks so much, Nate!

(I’ll be emailing you some Flextight scans, as well.)

Re-convert
Re-convert2000×1601 806 KB

4

Awesome! :clap:

And thanks for sending the original. Took a quick stab at it and converts very nicely in v2.2. (coming very soon).

flextight-negative-lab-pro
flextight-negative-lab-pro2602×1592 594 KB

1 Like