How much grain is there on a camera negative? Let's find out!

[singhcr]

As a lover of movies and film in general, one thing that has always bothered me is the usage of DNR to reduce a film to a waxy, smeary mess. It is obvious when the grain is totally blasted away or looks strange in motion that DNR has been applied, but when we see what appears to be a natural looking grain field, has it still been reduced? What would the image look like if no noise reduction was applied at all?

When we see movies on Blu-ray and especially UHD that were originally shot on film, there are so many factors at play. Is this a scan of the camera negative, or an IP? Was it 16mm, 4/35, 8/35, 5/65, or 15/70? Was it shot with spherical lenses and cropped to 1.85:1, or was it 4-perf anamorphic 2.40:1? Or did they shoot with the Super35 process using spherical lenses and get our 2.40:1 frame by cropping it? What speed film was used? Was it black and white, classic three strip Technicolor, pre-1990s Eastmancolor stock, or the newer T-grain EXR or Vision stock? Was it pushed or pulled at all? Did the scene involve the use of any optical effects? How recent is the scan and at what resolution and color space? What scanner was used? Was the scan timed to an answer print?

All of these factors play into how much visible grain is apparent, and that’s not even counting our choice of TV technology and whether our display is calibrated or not!

Since we do not have access to an original scan of the negative before any DNR is applied, we are at best making an educated guess at whether the amount of grain we see in a film to video transfer is unadulterated or not. I have always been curious to see how much grain is on a negative if it was left alone, but as I do not have access to studio film archives it was always a dream.

Until today.

Kodak’s Double-X black and white film stock has remained relatively unchanged over the years unlike their color stocks. I discovered recently that companies sell this film in 135 format cartridges for use in still cameras. I am going to shoot 5222, which was used in Raging Bull, Schindler's List and the opening sequence of Casino Royale for example. So now I have a way to get very close to using the same emulsions used for my favorite movies and see how much grain there would be on the camera negative from a raw scan. (FYI- they also sell color stocks with the remjet backing removed so one can develop with standard C-41 chemicals. I may have to try this out soon too!)

I will be developing this film myself, and as there is no remjet backing to deal with that you find on color negative motion picture film, the development steps will be the same as I normally do for black and white film: developer, stop bath, fixer, rinse/hypo, wetting agent, and drying. I just need to look up the preferred settings for 5222. This film will be shot at its box speed of 250 EI.

Some caveats:

1. A 35mm still camera exposes 8-perf frames horizontally (8/35), whereas the vast majority of movies were shot 4/35 vertically. I am essentially shooting VistaVision. A larger negative means finer grain.

2. The choice of developer influences sharpness, tonality, and most importantly for the purposes of this experiment, visible grain. I use Kodak XTOL as it gives high sharpness and medium grain and is more environmentally friendly over classic developers like D-76. I do not know how XTOL compares to traditional black and white motion picture developers like D-96. I would venture D-96 wouldn’t give as fine grain as XTOL as advances in emulsions and developers have moved towards greater sharpness and less visible film grain, not the other way around.

3. The lenses, coatings, filters, etc used will affect how sharp and fine the grain appears. Thankfully all of the three movies I mentioned above used spherical lenses, as will I for this experiment. I can’t use the exact same lenses as the aforementioned films, but at least they are all spherical.

4. The subject photographed as well as the lighting and exposure conditions affects the film density and therefore the visible grain pattern. A well exposed image will receive more light and grain will be finer as a result. The grain will vary with the image density. A white cloud will look more grainy than a dark wood chair in the same shot, for example. As a still photographer I have the ability to use extremely short and slow shutter speeds that a cinematographer cannot. I am not limited by the fact that on a motion picture camera, images are exposed at 24 frames per second. This places limits on the shutter speeds. I suspect this is primarily an issue for my experiment when it comes to low light photography. For a still photograph at night, for example, I could take a ten second exposure if I wished. If I were shooting a movie, I’d almost always be aiming for exposure times of 1/48th of a second to keep a 180 degree shutter angle. So if I was shooting a dark scene where I didn’t have any artificial light available, I’d have to either push the film a stop or two or just shoot underexposed at box speed, both options increasing visible film grain. I will try long exposure times and intentionally shooting with shutter speeds that are appropriate for motion picture photography even if my light meter indicates the shutter speed is not slow enough for a properly exposed image. I will also try photographing subjects that are similar to screenshots available for the aforementioned movies to get a more even basis for comparison.

5. The choice of scanner, exposure settings (i.e. timing lights), and color space influences the image as well. The scanners movie studios use are leagues beyond what I have at home. I’ve found out that my Epson V600’s film holders are such where they aren’t even within the focal plane of the scanner’s fixed focus optics, and even if they were, the film holders don’t do a good job keeping the film flat. It’s an unfortunately common occurrence with consumer scanning devices and I have been very frustrated trying to find something that is up to task. As the image will not be in sharp focus, grain will be less apparent and finely resolved over a professional scanner used for movies or still images. It is also limited to SDR. Unfortunately for the foreseeable future I’m stuck with the scanner I have. I’m just going to have to use what I have available.

I got the film a few days ago and will be running around town taking photos. I’ll post them once I finish off the roll and get time to develop and scan them.