Lens vignetting and flat field scanning

Digital Camera Scanning
#1

Hi guys, for details my scanning set up is a skier sunray copy box with 120 and 35 mm inserts. I’m having trouble with some of my scans I believe this is a result of vignetting from my lens. I have a nikon 55mm ai macro lens attached via converter to Fuji Xt-2. Perhaps the converter is the issue (K&F converter)?

I trying to get flat field correction working. For some reason I get these weird textures appearing after I complete flat field correction in lightroom.

Here is an example of the textures:

r34_1
r34_16000×4000 1.41 MB

If anyone can give me any advice as to what I might be doing wrong I would greatly appreciate it. For reference I scanned all my negatives. Then I removed the film holder and shot an image of the source light. I focused near to infinity so that the frame was out of focus. I’ve also tried making it out of focus in the other direction (i.e. focusing too close).

No idea what the issue is here! Thanks :slight_smile:

#2

Hi and welcome to the forum.

My first guess is that what we see are reflections picked up somewhere in the setup or process.

  • Do you also get the textures without flat field correction? <- maybe the flat field shot is off?
  • Is there any way for light to enter between the negative and the lens? <- shoot in dark room?
#3

Thanks! :slight_smile:

That’s what I thought, but the original image without flat field correction does not have these textures. And yes it’s shot in a room with no lights on and I’ve taped over any other light sources (i.e. side lights on the skier copy box).

Here is the calibration frame:

r34f_1
r34f_16000×4000 449 KB

#4

Hmmm, in this case, I’d not use the flat field correction… If it makes things worse instead of better, better not to do it seems the easy choice.

#5

Thanks for your thoughts. Is my best option then to just use the in built lens correction feature and work out appropriate vignetting amounts after processing through NLP?

These images should show a bit of the issue. The snowy one has noticeable colour change and light change in the top corners of the image.

r48_2
r48_23813×4766 801 KB

#6

Try another conversion with the top third of the image cropped off and reset the crop afterwards.
You could share a few raw files for us to better help you. Use Dropbox or any other sharing platform.

#7

Thanks will give this a shot, good suggestion. Here are two other examples in raw. I think it’s particularly noticeable with black and white photos that have a large shadowed area close to the film border. Let me know what you think :slight_smile:

#8

Use Lr’s lens correction/vignetting slider to correct the corners.
In B/W, simply press “V” in Lr to make the image B/W and all your color issues are gone.

Well, you know, for B/W, you don’t even need NLP. Flip the tone curve from / to \ .
The real benefit of NLP is most noticeable when color negatives are converted.

If we wanted to stay within the scope of the title of this thread, the raws of your examples would be of greater interest…

#9

I recently encountered a similar issue with ghosting like this after applying flat field correction. What I discovered was that Lightroom was picking another negative to use for the flat field correction rather than the calibration frame. My issue was that my calibration frame included a little bit of area beyond my light source. Your calibration frame looks fine, so I’m not sure why it is behaving this way. My recommendation is to try applying flat field correction to only one image at a time to see if it works correctly (assuming you are applying it in batch).

1 Like
#10

Brilliant, that seems to be the issue. Getting a much nicer image now!!! I try selecting one photo at a time and the calibration frame and it seems to be working perfectly! (albiet a bit of a fiddly process to do one at a time)

Two questions you might be able to help with:

  • does the calibration frame have to be white balanced before flat field correction?
  • Can i retroactively apply ffc to a frame that has already been cropped, rotated and put through negative lab pro? I’ve tried some tests and I can’t seem to tell if there’s a difference or not… There’s a back catalog of photos which require some FFC and I want to just use a calibration frame I already have as I normally leave aperture at f8 and ISO at 200 so those should be consistent… and scanning set up has been the same.
1 Like
#11

Ah, yes, if you select a group of images for flat field correction, Lightroom assumes that the calibration frames are interleaved. So, for instances…

Calibration Frame > Shot > Calibration Frame > Shot > Calibration Frame > Shot…

I wonder if this would work?

  1. Select all the shots themselves (except the calibration frame), then “right click > stacking > group into stack”
  2. Then select the calibration frame, and your single stack of negatives. So your selection would look like this:

Calibration Frame > Group Stack

I’m not sure if that will work, but it works on Pano Merges and is really useful for batch processing panos.

#12

Thanks Nate, unfortunately no luck though, but nice idea! It treats the stack as a single photo, so when you try to correct with flat field it just does the first photo of the stack. (If I’m doing something silly please let me know!)

If you expand the stack and select all photos and then the calibration frame. Then it greys out flat field correction as an option.

#13

Hmm… Ok… one more trick to try…

  1. Make a bunch of virtual copies of the calibration frame.
  2. Then in the Library view, move the virtual copies to be interleaved (every other photo is calibration frame) with the main photos (you need to be in the image folder view for this to work, i.e. not the “Previous Imports” or “All Photographs” view.)
  3. Then run on the whole group
#14

Unforunately no luck there, I don’t think the virtual frames are treated as calibration frames as if you put a virtual one at the beginning and or end it requests that a you place a calibration frame in that location. Easier to just take the same photo multiple times. Though it’s a lot of extra raw frames to have stored, though I suppose you can delete them after calibration is done… Either way flat field calibration makes a noticeable and much improved difference, even if it’s slow to calibrate each frame! :slight_smile:

However, even with flat field calibration some of my photos are still having falloff at the edges. See attached. Playing with border percentage on conversion doesn’t fix it. And playing with manual lens correction doesn’t solve it as the light at the borders isn’t uniform across either side so one side will end up darker in order to give correct lighting on the other. Any other suggestions?

DSCF2686
DSCF26863216×4824 2.26 MB

#15

Having similar challenges with my setup as well. I am also using a Skier box (v3) a Fuji APS-C mirrorless and an adapted Micro-Nikkor lens. It is very interesting to see your images which have similar artifacts to mine.

Any other thoughts @nate?

#16

@iancawood out of curiosity what is the film camera you used to shoot the photos shown?

#17

Would love to hear any comments/suggestions regarding the issue shown in this thread. I have a few things to try out on my side, short of trying a whole new film holder. but any info is helpful.

I am experiencing the exact same kind of artifacts, see the pic below. This is just a white wall in flat light. I have adjusted exposure and contrast to exaggerate the issue.

white_wall_edge_exposure
white_wall_edge_exposure1500×1000 207 KB

And here is another shot - just 3 frames later on the roll. Same exposure (different, but equivalent SS/aperture). The vignetting has moved to the other side. I wonder if there’s something going on with the curvature of the the film interacting with the holder and casting a very subtle shadow of some kind.

white_wall_edge_exposure_02
white_wall_edge_exposure_021500×1000 205 KB

My process is much the same as the OP’s. Adapted Micro-Nikkor 55mm/3.5 on a Fuji X-H1. Film is on a Skier Sunray v3. What do you think?

#18

If the corners of your camera-scan are darker, this can produce brighter corners and color blobs on conversion.

My process:

  • Load film and focus
  • Remove film. Shoot a “blank frame”
  • Load into Lightroom. WB with eye-dropper.
  • Examine in the Develop module. Look at the brightness in LR points (0-100) at center, edges, and each corner.
  • Figure out a LR adjustment that will make it even within a couple of LR points. Might be radial adj for vignetting or a graduated adjustment. Save that adjustment.

Then do camera-scans. Apply the saved adjustment. Then convert in NLP.

#19

Thanks Richard, that is basically what I’m doing now, but it feels like a band-aid solution. And, as you can see, even two frames near one another on the same roll (the two frames i posted above are 2-3 shots apart, taken at the same time in the same conditions) have distinctly different artifacts. So, the flat-field correction, or a general vignette correction is not going to work. Basically the same issue as OP.

I’ll keep your process in mind for the next roll I scan though. I think I will also get a scan from the lab to compare against to see if the vignetting is on the emulsion or not. Cheers

#20

From your description, something is causing this, not random. Could be end of strip effect? Sometimes I’ll jam a piece of dark film adjacent to the end-of-strip frame.