Output formats: Bit depth, “flat” scans, and “log” scans
C2D offers to scan in three different image formats or “containers”: AVI video files, DPX and TIFF image sequences.
Bit depth:
The output formats differ in bit depth, that is the amount of colour information they store per colour channel and per pixel, measured in bits. AVI is 8-bit, DPX is 10-bit log (explained below), and TIFF is 12-bit (saved as 16-bit, 4 bits are unused).
E.g. An 8-bit image will save the Red, Green, and Blue values of each pixel as 8 bits each, meaning there can be 256 different values of Red, and the same for Green and Blue. Combined, the values of Red, Green, and Blue can give about 16 millions possible colours for each pixel.
The camera is 12-bit, it can “see” 4096 different values per pixel. Cine2Digits works with this data in 12-bit, with R, G, and B colour values of 1 to 4096 (actually 0 to 4095).
Saving to TIFF allows to keep all that information. It is the best choice if you intend to do colour grading of your scans afterwards. But it also produces the heaviest files and the slowest capture times.
Saving the image in a 10-bit or 8-bit format allows faster capture and creates lighter files, but means some precision will inevitably be lost.
When saving to 8-bit AVI, the colour values for each pixel are linearly converted from a scale of 1 to 4096 to a scale of 1 to 256. All the colours are reduced in precision equally. This linear reduction in bit-depth can cause banding and limit the possibilities of post processing correctly, so applying colour and gamma corrections directly in C2D is recommended when saving to 8-bit.
When saving to 10-bit DPX, on the other hand, a (logarithmic) curve is applied to the original 12-bit values to reduce them to 10 bits preserving more details in the lower values (darker pixels) and compressing the information in the higher values (brighter pixels). This non-linear mapping mimic the response of negative film and matches the particularities of the human eye, which needs more nuances in the dark areas than in the light areas to perceive them correctly. It helps reduce the file size and capture time while preserving as much “valuable” information as possible. It is usually a good compromise.
“Log” picture recording generally describes a way to preserve more details in the darker values while not loosing valuable information in the lighter values, but giving the images a washed out look that needs to be balanced with some colour grading.
A “flat” scan should preserve all the available data without any precision loss or manipulation. Therefore you should save in the TIFF format if a true flat scan is desired.
The DPX format “log” scan does offer a comparable result in quality and perception while being a lot lighter. So if your policy is not absolutely strict about preserving the 12-bit information, it can be a good choice.
You may be used to applying look-up tables (LUTs) to convert the log response curve into a linear one for display or for conversion to other formats.
Be wary that LUTs can produce bad results, and an ad-hoc correction curve can be preferred to produce a result which looks right. The empirical approach is often better than a “standardised” one.
For film archives, the response curves of the original film is often unknown, and the output files “look” will depend on how the scanner was configured when capturing. So a standardised treatment is not imaginable and manual grading is required to get the best result.
If a flat scan is needed, do not apply Gamma correction during capture (Gamma must be set to 1.0).
See Gamma settings for more information.
Understanding the data format:
The camera sensor on the scanner captures information linearly.
Your colour and light settings in Cine2Digits will impact what information is captured by adjusting the LEDs on the scanner’s light source. Your goal is to maximize this information to be able to be as close as possible to the source material and have the most control possible on your image in post-processing. Cine2Digits is made precisely to maximize the dynamic of the captured image. How you set it is up to your needs and tastes, but in the end you may need to do post-processing to get a good-looking video for a monitor or projector. The images saved by C2D, whatever the selected output format, can be considered “raw”, unaltered, and linear. They can be considered in the sRGB colour space (even though the term colour space does not really apply before any digital transformation is made in post), which is equivalent to the Rec.709 standard except for some gamma adjustment.
For positive film, if you are not using the advanced colour matrix, there is no image processing happening in C2D (apart from Debayering, to render a colour image), what you get is the data the camera “sees” and it gives you the freedom to do whatever you need with this “raw” data.