Content-Aware Grain

Tom Schum wrote:

Truman Prevatt wrote:

Assuming one is trying to simulate film grain the S/W needs to account for some basic facts of where film grain arises and the fact that different film produces different grain profiles. In B&W the developer also has an impact on the grain. For example classic D76 contains "silver solvents" which smooth off the sharp edges of the grain while Rodinal has no silver solvents and the grain edges are sharp. The reason why film developed in Rodinal appears to have more grain than D76.

Grain arises through the development process where in the exposed areas, the silver halides in the emulsion is converted to silver. The higher the exposure the more silver (denser negative). In the shadows there is little silver and the fixer removes the unexposed silver halide leaving only the transformed silver. In the shadows there is little grain in film and in the highlights is where the grain appears. Grain depend on the emulation type. Thick emulsion films like TriX will have contain bigger clumps of grain in a mix of sizes. Think emulsion films (classic show films like Ilford Pan F) contains much smaller crystals in the emulsion so much smaller grain with no large clumps. The Tabular grain films, Kodak T-max, Ilford Delta and Fuji Neopan, the silver halides are flatter and tabular. They tend to be finer grained but have less exposure and development latitude than the traditional grain films.

Film grain is just the opposite of digital noise since digital noise is higher in the shadows than the highlights.

For me the Nik collection provides the optimal tool for film simulations in among other things its ability to simulate realistic grain based on the film selected. I find Silver Efex to be by far the best S/W for emulating classic and the newer B&W emulsions with a very realistic grain structure with the flexibility to tweak the grain structure to taste. Granted I am partial to B&W but the few times I have used Analog Efex - it provides a comparable capability for color as Silver Efex does for B&W.

Interesting post, and you describe film grain that is the opposite of my concept. I'm not saying you are mistaken, to be sure. I'm most likely to be mistaken, since it has been 50 years since I made enlargements from Plus-X monochrome negatives.

Here is my concept. If there is less light the emulsion will be thinner and the film grains will be more easily seen. If there is more light the emulsion will be thicker and the film grains will be closer together and harder to see individually. Fortunately the Exposure X6 grain tool sliders will support this concept as easily as they will support your concept.

There is not enough control in the Exposure X6 grain tool to allow grain sizes to be different in different luminance areas, unfortunately. This grain tool is definitely not perfect.

In B&W the chemistry is pretty simple.  Since color film is basically three layers of B&W emulsion covered by different color filters, followed by a dye transfer process it is a little more subtle but the principle is a same.

The emulation contains silver halide crystals - of all different sizes with high speed films predominately large crystals and low speed films small crystals.  High speed films are normally thick emulsion since the crystals are larger and low speed films thin emulsion.

When photons hit the crystals, they some of the molecules undergo a reaction forming the latent image.  This reaction is proportional to the number of photons, ie intensity.  The developer will react chemically to turn those exposed molecules into metallic silver.  The more photons (higher exposure) the more silver halide crystals are converted to metallic silver by the developer.  In the shadows there is little silver.  At that point the stop bath (acid) reacts with the basic developer stopping the action.  The fixer then dissolves the remaining silver halide crystals so the film is not longer light sensitive.  This leaves unexposed areas clear and highly exposed areas dark with a lot of silver crystals deposited.  The grain arises from the boundaries and the size of the crystals of silver.  These are proportional to the size of the crystals of the silver halide.  So while the shadows will have a little grain and the mid tones a little more, the greatest impact of the grain is in the highlights.  That is the opposite of digital where shot noise is going to be higher in the shadows and not noticeable in the highlights if proper exposure.

Some places grain is quite noticeable is in clouds in a black and white.

In color (negative), there is a second process where in each of three layers, dye is attached to the silver the density of the die proportional to the density of the silver.  This happens in each of the three layers with different colors of dye.  The final process is the silver base image is dissolved leaving only the dye blotches.  The edges of the dye blotches are smoother than the silver crystals they were attached to so the appearance of grain is soften.  That was one of the reasons for the development of the C41 B&W films which reduces the grain compared to a silver B&W film for the same ISO.