Hey there! I’ve been having some issues getting consistent colours and skin tones when scanning 35mm portra 400 on my Epson V600 and was hoping to get some advice on how to get better skin tones. I get all of my film developed at a lab but I always scan myself. Here’s an example:
Scanned with Epson Scan 2 at high quality, 2400dpi, no colour correction. White balance taken from the film border prior to conversion in NLP. All default settings in NLP, Noritsu profile.
I’ve got a friend who shoots on Kodak disposables, lab dev&scan, and she always ends up with superior skin tones and colour casts in her images and its starting to bug me. Could anyone help me out with this?
Note that I set Tones to Linear and film to something other than None and pushed the Hue slider to the right. You might also try with a lower pre-saturation value. Here, I left it at its default.
Note that I converted your jpeg file to 16 bit tiff before conversion and tweaking. Jpeg is simply too limited for the color corrections I wanted in this case. The image at the right is a tiff copy that I used to change hue and saturation of sky, skin and forest. All changes made are global. Skin tones could further be treated with local corrections.
I recommend NOT white-balancing prior to conversion when working with a TIFF/JPEG files. The reason is that this control works differently on TIF/JPEG files vs RAW files, and it often leads to very very intense colors (which works well for specific styles, but it generally less accurate.
As others have mentioned, it’s also very important to scan as a TIFF (and not a JPEG) so that you can get full 16-bit color depth. You’ll quickly run into posterization and other tonal issues if you’re trying to work with a jpeg.
Also, the lighting of the scene itself will have an impact on the skin. Direct, bright, mid-day sunlight with a strong shadow (as in this scene) just isn’t going to produce great skin tones.
That said, it’s still possible to get a more natural rendering from this jpeg… shown below using v2.2 (not released at the time of this comment, but coming very soon).
Thanks for the responses Digitizer and Nate. I have been scanning and working with my negatives as tiffs, I only uploaded the samples as jpegs as the forum wouldn’t allow me to upload as a tiff.
I’m quite new to the software and I’ve noticed that I often have to turn down the lights, darks, and brightness (or I adjust the white/black clips) to get images that aren’t overly saturated and contrasty. Could this be due to my scanning methodology?
You can find it out by changing the settings before you scan the negatives. Maybe you can influence brightness, contrast and saturation with your scanning software?
On the first tab of NLP, you can also change the pre-saturation. I found that lower saturation values produce images that are closer to what I like. If you create virtual copies in Lightroom, you can test the different settings without wasting too much disk space.
Please note that the settings you find with one negative might not necessarily work with other negatives and that different paths can lead to the results you want. Check out what Nate and I did for similar outcomes.
@nate, to your point about white balancing, does that include when the scans themselves have RGB exposure adjusted in vuescan that ends up neutralizing a lot of the orange mask? With the scans I have, saved as RAW from Vuescan, the RGB exposures were tweaked by Vuescan that end up producing negatives that do not have strong orange casts.
Per the guide on using Vuescan to create RAW DNG files, I recommend scanning in “image” mode in Vuescan rather than in “color negative” mode (which is the mode you are referring to). Vuescan isn’t consistent in the way that it applies color multipliers, and most times you will have better results if you make RAW DNG scan in image mode, and then white-balance & convert using Negative Lab Pro. This is what the calibrated profiles and Negative Lab Pro are expecting when it comes to a RAW DNG file.
Hi Nate, In my case, I have a huge backlog of already scanned RAW TIFF scans from Vuescan. I use Vuescan to convert the RAW TIFFs to DNG format to use in Lightroom / Negative Lab Pro. Attached is what the TIFF looks like out of Vuescan (sized and converted to JPG and in sRGB colorspace for web). The RAW scan, from how I understand it, is before any settings are applied besides focus, RGB gains and/or exposures and infrared cleaning in VueScan. In theory, then, I should be able to “scan” the raw file to a DNG output file to meet the requirements.
Sometimes I do have a difficult time with color balancing these files and I’ve always wondered if I should be white balancing off what remains of the mask along the edge or not before converting. That was why I asked.
I didn’t do much to tweak the image much - to my eyes it’s a little green and could have a little boost in the shadows, but it was just an example image. I’m thrilled with how close it got though. ColorPerfect required a lot more tweaking and edge cropping to get just so.
