Digitizing a Black and White Negative While Preserving Its Subtlest Textures

⊕ Alain Oguse
alain (at) oguse.frAlain Oguse - 2025 - 2026

WARNING: After the initial publication of my work, I discovered a fundamental error in my approach: I had underestimated the need for "surgical" focus precision.
This website now presents the corrected version.
ERRATUM:  Quick summary of the error and its correction

My Adventure with Point Light and Some Ensuing Ramblings

Trained in professional black-and-white printing in one of the best Parisian N&B labs in the 1960s, I sought to achieve the finest possible results when digitizing 135 negative films. However, I was disappointed with the results I obtained using the various processes I tested. Digitization with a DSLR seemed to be the most promising technique, yet something wasn't right...

I am not an optics expert. Therefore, one might legitimately question the validity of what I present on this subject.  My homepage was slow to get indexed. I asked Google’s AI why. Here’s its answer: "Google tries to determine whether a website’s author knows what they’re talking about. You write with a lot of humility: 'I am not an optics expert'. Instead, present your profile from the angle of your real strength: You have practical experience—highlight it!" Apparently, I’m hopeless at selling myself… Before I even got around to fixing my "flaw", I came across a fascinating interview with Kip Thorne, Nobel Prize winner in Physics for his work that helped prove the existence of gravitational waves. To my surprise: "[...] I've been proved wrong often enough in areas where I think I know the answer that  you shouldn't take my pronouncement seriously." I felt less alone and didn’t change my text. Too bad for Google. :-)
https://www.youtube.com/watch?v=PTs--eFrzGo (2:19:23)I have only my solid experience and the "eye" I gained from it to support my claims. I wanted to base my approach on the most enlightening sources I found and my ability to evaluate the results of my tests. Thus, through the story of my adventures, I hope to show how I understood the physical principles of light and optics and why they often lead to disappointing results when digitizing silver negatives.

Thus, my research on the web allowed me to discover that the cause of my disappointments was the diffuse light, which is nonetheless used, more or less discreetly, by all processes. Point light thus became the only recourse possible, even though it is well established that its use is not without drawbacks. Would I be able to master it?

Spoiler alert: Yes! But it's not a simple and quick solution. However, what a pleasure to obtain A3+ prints that I don't have to be ashamed of ;-) Here are the videos of the opening of two exhibitions of photographs by François Huchet using collimated point light.
- https://www.youtube.com/watch?v=DcqbX-MB-bU
- https://www.youtube.com/watch?v=IzIiosSaSm4

Full Version only in PDF

• PDF (en) - 58 pages - (39 Mo)

So here's a complete version of the story of my adventure, in which I recount my trials, the prototype, the results, my thoughts, and some personal anecdotes. I hope it will help or give ideas to some. Qptical connoisseurs will easyly improve and, above all, simplify this prototype to make it a more convenient tool.

You might prefer the technical excerpt below, which sticks to the technique and results in images.

Technical Excerpt

Partial version below in HTML or PDF

• In HTML - First Part: Towards a Prototype Towards a Prototype
  A disappointing first attempt
  Understanding the issues
  What role does diffuse light play?
  A prototype and its optical principle
  Theoretical and "pragmatic" schematic of the prototype
  Necessary equipment
  Settings

  
  

• In HTML - Second Part: Results and Improvements Results in images
      Initial results with a white LED
      Results with a green LED
      And yet there remains a problem
  Final assessment in a few images
      An example with white LED
      An example with green LED
  In conclusion

Post-production and printing
of a point light scan

When writing the above articles,, I remained focused on the core of my subject. It was only after publishing it that a reader pointed out that, once the RAW file is saved, it is not straightforward to achieve the desired print. I must admit that the scattered information I provided on this topic was quite fragmented....

• In PDF (fr) - 24 pages - (1,7 Mo)
• In HTML

I've often heard about "reticulation" of gelatin without ever encountering it...

Scanning with point green light reveals details that remain virtually invisible in diffuse light. In this way, a major alteration becomes clearly visible, identifiable and therefore understandable. The opportunity to find a corrective solution?

• In HTML - Does it exist? Can it be treated?