Film 'structure' appears very evident in Camera scans

[Photodom]

Hi.

I have just begun trying to scan some old SLR slides and negs from a few decades back. I am using a Nikon D5100 with a Sigma 150mm f2,8 lens stopped down to either f16 or f20.

In the scans I did yesterday, it would appear that the structure of the film (not the grain,) has become very evident. This has happened on both slides and negatives which have been stored in protected boxes and sleeves.

I am not sure what is happening, as even using noise reduction techniques in Lightroom and Photoshop does not make a great difference and often softens the image too much.

I was using an old slide viewer I had, made by Hanimex call a Simon Slide Viewer as the backlight. It wasn’t that bright and I just wonder if this could be the issue?

The upload demonstrates the problem, especially in the sky.

Any hints or tips would be much appreciated.

Thanks

[FatihAyoglu]

Don’t know about the film structure but with f/16 you’ll have diffraction problem.

[Digitizer]

High ISO film has coarse grain. High ISO camera settings and underexposure can reveal digital noise.

Hard to say what the cause of your issue is. Please tell us what the original film is and at what settings you took the image, maybe post a screenshot of the histogram (of the unaltered camera scan) too.

[Photodom]

The original film was Fuji 100 and the exposure was pretty good. The copy camera was also set to 100 ISO. I took a number of images at varying shutter speeds but they all show this.

[Digitizer]

@Photodom

I took a number of images at varying shutter speeds but they all show this.

okay, you have changed exposure with the same backlight. All images have the (same) structure, all images have the (same) backlight.

Try something else as a backlight (like an overcast sky or a white wall lit by an incandescent light bulb) and make sure that the backlight is far away from the negative.

If the structure remains, the cause could be in the negative, check it with something like this:

[FatihAyoglu]

Also, what shutter speed are you getting with a dim light and f16-f20 aperture. It can be long exposure noise on the camera. Another question is, how is the raw file? equally distributed histogram? I am asking this because highlights of the negative are the shadows in the RAW file and I see noise in the highlights of the positive image, meaning shadows might be crushed on the RAW file and NLP or you are trying to recover it way too much for the sensor of D5100

Try to scan at Aperure mode with f8, highlight controlled EV at +1. My D750 gives very consistent results without any crushed shadows or highlights on the actual unaltered RAW file

[Photodom]

Thanks for the feedback Fatih.

I think I’ll set up a test rig today and scan a slide at a variety of aperture/shutter speeds and do the same with a negative and see what I can get. I will also scan each option with -1 and +1 exposure compensation to see what the effect is and maybe create an HDR version in Lightroom. This should provide me with enough options to find out which one works the best.

Annoyingly I have an older Hewlett Packard Photosmart scanner than can scan Negs, Slides and Prints (up to 7"x5") and obtained great results with it in the past. It cost me about £800 when I bought it (compared to the £2,500 that Nikon wanted for their film scanners at the time.) The problem is it is a SCSI scanner and Windows 10 does no support it.

I might try and install Windows XP on a separate Hard Drive, boot from that and see if I can use the scanner that way.

I have 2 PCI SCSI cards so the other option might be to buy a cheap case, install my old motherboard (I’ve recently upgraded,) stick in superfluous IDE hard drive I have and then try it on that. It just means another computer case taking up space on the floor though.

[FatihAyoglu]

It’s really important you don’t have clipped shadows or highlights on the unaltered RAW file, you might even need +2 maybe so give it a try

[Digitizer]

@Photodom

scan a slide at a variety of aperture/shutter speeds and do the same with a negative and see what I can get. I will also scan each option with -1 and +1 exposure compensation to see what the effect is and maybe create an HDR version in Lightroom. This should provide me with enough options to find out which one works the best.

With color negative films, contrast is not the problem for a digital cameras. You’ll be able to fit the histogram. Underexposing is not a good idea as @FatihAyoglu explains above.

I’ve found that I can bracket between 0 EV and +2 EV with most of my negatives and that the shot in the middle of the series yields the best conversions.

HDR composites have not given me an advantage that outweighs the effort. It’s certainly okay to see what you’ll find though.

Check this out for best practices, it might help you find your way more easily:

Best Practices for Scanning Film Negatives with a Digital Camera (DSLR or Mirrorless)

Scanning your film negatives with a digital camera can be an incredibly rewarding – or incredibly frustrating – process, largely depending on your setup and know-how. While DSLR (or Mirrorless) film scanning holds many promising results – like RAW capture, faster capture times, and sharpness that rivals $10,000+ drum scanners – there are many potential stumbling blocks along the way that could prevent you from getting the results you want from your film negatives. The purpose of this guide is …

[pemcg]

I had exactly the same problem when I inadvertently left my scanning camera (Fuji X-Pro2) in Auto-ISO mode. Make sure you set your camera to the lowest ISO value (on mine it’s ‘L’ for “low”)

Hope this helps,
Peter

[FatihAyoglu]

It should be base ISO Not always lowest ISO is a natural iso, therefore it can produce artifacts. Like my D750 has a base ISO of 100. It then L1 which is lower than that and L2 which is even lower than L1. Both L1 and L2 are pull iso which are not natural

[pemcg]

That’s interesting to know, thanks

[Photodom]

Managed to try the scans again with a different setup.

I used an LCD screen as a backdrop set to ‘white’ or as white as I could get it.

The first few copies I could see the individual screen pixels coming through, so I moved the rig I had set up forward about 6 inches which cured that problem.

Camera was set to 100 ISO (D5100 base ISO,) and the setting were generally aperture priority at f8.0. I shot in Raw and jpeg.

Shutter speed was set by the camera and I bracketed to ensure I obtained a good image.

The couple of samples attached are jpegs straight from the camera with no adjustment in Lightroom.

I am fairly happy with them. The slide of the Battleship Potemkin in Leningrad (as it was,) is 40 years old, and the slide of my sister in law is around 30 years old.

I don’t see an evidence of granularity or excessive noise in the shadows.

[SteveA]

I downloaded your images and blew them up. Hard to be sure, but I’m wondering if they are really focused properly. I’m seeing grain and scratches, but they’re not really sharp. Either that, or the lens you’re using isn’t optimized for macro work.

[Photodom]

Hi Steve

Thanks for taking the time to look at the images.

I do use a macro lens (Sigma 150mm f2.8) and the slides are up to 40 years old so will have a few scratched on them. These are straight out of camera with no work done on them whatsoever.

I used manual focus as the auto focus was finding it hard to lock on for some reason (the Sigma is not the best autofocus lens in the world - very slow and hunts a good deal.)

[burkphoto]

• Use the lowest ISO on your digital camera.
• Be sure everything is physically locked down. Use a cable release, wired remote, self timer, or smartphone app to trip the shutter. Vibration is your enemy!
• DO NOT use small apertures such as f/11 to f/32. Your camera has diffraction limiting of sharpness issues that start at approximately f/8. Stop the macro lens down about three stops. An f/2.8 macro lens probably is sharpest around f/8. If your film is kept flat and plano-parallel relative to the sensor, that should be sufficient.
• Use a genuine macro lens capable of 1:1 (the image projected on the sensor is the same size as the image on the film).
• Use a well-diffused light source of known-accurate color fidelity. An example would be a photo quality LED panel with 96 CRI and high CRI (LE). https://en.wikipedia.org/wiki/High-CRI_LED_lighting
• Keep the light source masked off from the camera lens, to avoid flare.
• Keep the light source at least two inches behind the negative.
• Do an extreme underexposure test of the evenness of your illumination. Adjust positioning of the light source until there is no fall-off or hot-spotting.
• Record raw files.
• Use the same exposure for every frame. You want to be sure the histogram shows no highlight burnout, but does fill the graph.
  Use PEC-12 film cleaner as a last resort to remove stubborn stuff on your negatives. I use in this order:
  A Giottos Rocket Blower bulb
  A Staticmaster Polonium-strip brush (neutralizes static on the film, then brushes surface dust off)
  An Ilford AntiStaticum cloth
  PEC-12 with PEC Pads (go to photosol.com for their supplies).

I hope that helps…

[Photodom]

HI William

Thank you for the very comprehensive reply.

I do practically all of the above with the exception of using the Pec-12 Film Cleaner, which I will probably invest in.

The Sigma macro lens I use is also 1:1.

The only think I need to re-consider is the light source. As I said, the sample images were telen bakclit by an LCD Monitor screen with a white background.

Thanks

[FatihAyoglu]

You can use 99% alcohol to clean the films

[Photodom]

Fatih

I’m Scottish. That’s not what I’d be doing with 99% proof alcohol.

[Digitizer]

@Photodom

I’m Scottish. That’s not what I’d be doing with 99% proof alcohol.

… I wouldn’t use a single malt either