Hi,
Overall, I achieve good scan and transfer results. Only grass and leaves from bushes and trees often appear too green, even autumn leaves shine like fresh spring green. Red values also often appear too intense.
The negatives were scanned with Vuescan on a Reflecta RPS 10M film scanner (settings according to Nate’s recommendations). I am using NLP 3.1beta (the same problem also occurred with 3.02, so beta is not the issue).
Here is a link to the raw file and the result after conversion: scanning - Google Drive
Thanks for any help.
Al
Hi Al, welcome to the forum.
I can’t replicate your results because Lightroom won’t white-balance the negative based on the overly bright bit of unexposed film edge supplied in your download. As well, everything I say below is based on a sample of one, and could well be invalidated by examination of a larger sample of different photos.
It looks to me that there is either unevenness in the illumination of the film (a scanner issue) or excessive contrast introduced in the conversion (NLP settings issue) as the upper left corner looks too dark relative to the rest of the photo and how a normal sky would look, and toward the center the photo looks somewhat washed-out; the tonal gradation of the sky is probably a bit more abrupt than it would appear in reality; the L* value of the sky on the unconverted negative ranges from 81 upper left down to 64 in the center. With some over-exposure in the center, the orange rooftops are weaker than they probably look in reality for those kind of roof tiles. The greens are probably lacking a bit in yellow content, which would also be consistent with the weakness of orange. The sky-blue is somewhat unnaturally Cyan-ish - probably should have a bit of added magenta and yellow, also consistent with the overall weakness in the orange content.
Much of that can be fixed by editing the colours and the brightness/contrast, but you also have another apparent problem - it looks to me, judging from the definition of the “grain” (in reality dye clouds) that the sharpness is well-maintained in the upper sky, fuzzy in the center and again better at the bottom of the photo. This would indicate that the negative was not completely flat and the focus is toward the outer areas rather than the center - an unusual kind of result. You can see this quite clearly magnifying the converted photo to 100% and scrolling down.
So it looks to me that you may have some issues with the scanner and/or perhaps with the colour model used for the conversion (I suggest trying “Basic” rather than “Frontier”), and/or perhaps with the White-balancing in Lightroom. Before you convert the file in NLP make sure that in Lightroom you have implemented a proper white balance and eliminated any margins around the photograph itself. Within NLP’s edit panel, also have a look at the Tone Profile you are using. I find Lab Standard generally the most satisfactory, but occasionally one of the others suits the photo better. I normally leave HSL at “Natural”, but periodically when stronger effects are needed one of the others could be helpful. I tend to leave WB at “Auto-Neutral” unless it is way off.
Had a look at the scan with RawDigger which shows a heavily blown red channel and partially blown green channel. How well RawDigger can handle scanner output, I can’t say, haven’d looked at the manual too deeply.
Nevertheless, I suppose that a
scan with lowered exposure could help.
I tried several conversions with NLP 3.1 b10 - and got the following output:
From left:
- converted the scan as is, with Basic and pre-saturation at 1 - and Border Buffer at 25
- same as above, this time with Border Buffer set to 10
- same as above, this time with automatic WB
- same as above, without setting WB, but with automatic tone control
- same as above, but with automatic WB and tone control
- converted according to “rules”, but with WB picked up from the shadow left of the car
While ETTR might look like a good idea, it’s counter-productive if pushed too far.
Also, some negatives (and positives) require special care and a few trials and/or postprocessing, be it with NLP’s second tab tools or - with a positive copy - in Lightroom.
Thank you very much for your super quick and detailed replies. These are several very good suggestions that I will now try out.
Unfortunately, the negative dates back to my teenage years when I used a very cheap camera, which probably explains the discolored sky in the upper left corner, which can also be seen in other negatives. The negatives are also so badly bent along the longitudinal axis that the scanner cannot flatten them properly and the image is not sharp throughout.
