TrueGrain is a pro-grade tool for accurately recapturing the aesthetics of black and white film with digital photography.

With TrueGrain, you can:

- Accurately recapture the aesthetics of particular film stocks—including “lost” films—while retaining an all-digital workflow.

- Creatively employ credible film aesthetics.

- Add high resolution film grain information to digital images to elegantly minimize the pixilation effects of upsampling.

- Match digital imagery to existing film images for restoration and compositing purposes.

The following example demonstrates a straightforward aesthetic application of TrueGrain at both the macro (9.5:1) and micro (1:1) levels:

Achieving the look of film, digitally

Other approaches exist that algorithmically simulate the generic “look” of scanned film by converting to grayscale and adding random noise to resemble film grain. These approaches differ little from what you could readily accomplish by using Photoshop’s built-in filters. Moreover, the results are not particularly convincing, because:

- real film grain is not random noise.

- real film grain looks dramatically different across different film stocks.

- real film grain expresses itself differently based on exposure.

TrueGrain’s uniqueness lies in its use of empirical data collected, sampled, and profiled under carefully controlled conditions. It draws from a library of historic film stocks, some of which have been out of production for some time, and are not likely coming back.

TrueGrain can actually adapt a digital image to match the measured dynamic range and spectral response of a specific film stock and then correctly incorporate that film’s actual film grain into the image, even respecting how that grain expresses itself relative to exposure. The result is an image that is basically indistinguishable from a carefully scanned film frame of the same scene, using the same exposure, stock, and development process.

Hovering your mouse over the 1:1 examples, below, will reveal the underlying digital source image, which is the same in all four cases.

Example 1

Example 2

Another example from the same conventional algorithmic approach (allegedly a Kodak T-Max 3200 effect). Note how the “grain” is just the same noise as the above example applied more strongly.

Processed by TrueGrain as Kodak Professional T-MAX P3200 with real, sampled T-MAX 3200 grain.

TrueGrain implements a variety of different film stocks, and more will be added as they are completed. Please see our Grain Library for comparative examples.

Aesthetics

The purpose of TrueGrain is principally aesthetic. Photographs created with film have a different character from those taken digitally, and sometimes, the film aesthetic is desirable. Although film grain is most consciously noticeable when examining a highly magnified image, it is still evident in images that have been drastically scaled down to Web size:

Looking at a detail at 100%—such as would be clearly expressed in a print—the aesthetic difference is not subtle.

Matching

What better way to prove the efficacy of our process than to compare, side-by-side, analog and digital equivalents of the precise same scene? Unfortunately, this is easier said than done. First of all, film stocks are simply not that consistent from batch to batch. Second, even if they were, there’s no way to insure that they develop exactly the same way. Third, different digital capture devices have their own peculiarities when it comes to exposure, and none behave like film. Less crucially, even if you use the exact same optics and set up special equipment for fine-tuning the depth of field, the images will not turn out identically.

Still, we got close. The film original was shot and developed conventionally, then scanned on a professional film scanner. The RAW processing of the digital original was on Adobe Camera Raw defaults, without any effort to “improve” the image. The only manipulation of the digital source material was adjustments to the dynamic range settings within TrueGrain in order to match the overall exposure characteristics of the scanned image. The only “trickery” involved is that the film original was shot and processed concurrently with the grain samples. (The point of this exercise was to validate our approach, not to create an experiment you could reproduce.)

Copyright © 2007 Grubba Software.

http://zonezero.com/magazine/articles/grubba/index.html