Visual Effects FAQ
Common question regarding visual effects
Common question regarding visual effects
There's a rule of thumb: "The cleaner the image the better the key".
While many VFX-heavy productions have successfully used ProRes 4444 and ProRes 4444 XQ for their VFX work, only ARRIRAW provides the full resolution uncompressed sensor data and therefore the best choice for VFX work.
If ARRIRAW is not an option, ARRI recommends to only use the highest quality codecs for VFX work: ProRes 4444 and ProRes 4444 XQ.
The ALEV 3 CMOS sensor has twice the number of green photosites than blue photosites; therefore the resolution of a green screen is higher compared to a blue screen. Also the green channel is less noisy than the blue channel when using tungsten light.
Please also see the next topic.
For a green screen shot neither white point setting has an effect on the green channel gain. Daylight and tungsten light deliver the same quality.
For a blue screen shot the blue gain used with a 3200K tungsten light is substantially higher than the blue gain used with a 5600K daylight, therefore a blue screen captured under tungsten will be more noisy than a green screen. Hence daylight is preferred.
The brightness of the screen should correspond to the foreground, which needs to be separated. A good aim is, when the foreground is properly exposed for skin tones 1/3 of a stop above 18% grey, the screen should read about 1/3 to 2/3 of a stop over 18% gray. Please take care that the background is evenly lit and not underexposed.
The short answer is: 200-400 ASA.
Here's the long answer: The "base" rating for the ALEV 3 is 800 ASA. When setting the camera to 800 ASA you get 7.4 stops of latitude above middle grey and 6.6 stops below. This gives you a very good range similar to that of film.
When shooting green or blue screens in a controlled studio environment, often the 7.4 stops above middle grey are not needed. In a situation like this it could be preferable to set the camera to a lower ASA rating. As long as the highlights don't get clipped a lower ASA rating will give you less noise in the image.
When looking at a ALEXA or AMIRA image, often the noise/grain is not visible to the eye, but it is visible to a compositing keyer. Having less noise in the image means that the footage will be easier to key.
Linear Color Gamut
Modern keyers best work when supplied with linearly-encoded images. Working in a linear color space will most often give better results than working with Log C encoded plates.
Most often it is useful to degrain/denoise ALEXA/AMIRA green and blue screen plates. Even when the eye cannot see any noise present in the image, there will still be noise present that will have an effect on the keying of these plates.
In case you are working with ARRIRAW files, you have the chance to set the sharpening value according to your wishes. This means that you can reduce the sharpening after the debayering and might find that you can pull a better key. You can use a second pass with increased sharpening for the fill or do a resharpen in your compositing system. Over-sharpened images create a bad key, less sharpened images a better key.
Both cameras offer three different encodings of the recorded image:
While ARRIRAW is encoded as linear data, it can be converted to the other two encodings using the ARRIRAW Converter software or 3rd party applications which support ARRIRAW. For ProRes clips, the encoding can be logarithmic or video.
ARRI also offers Transformation-LUTs to convert from the different encodings back and forth. This can not be done without loss of data. For instance when recording a clip in Rec 709 encoding to ProRes files, these files can be converted Log C, but will have lost detail in blacks and whites.
Linear Encoded Data
Linear data is mostly used in VFX processing because it is the natural encoding for computer generated elements. The ALEV 3 sensor has a dynamic range of 14+ stops which corresponds to a linear range of more than 15,000:1. In most cases floating point numbers are used to store this range (the OpenEXR image format, for example, is based on 16 bit floating point numbers).
Linear from ARRIRAW files
The most direct way to linear files is to record ARRIRAW and to process the data with the ARRIRAW Converter (ARC). Third party software vendors also support ARRIRAW in their systems.
Linear from Log C files
Another way to obtain linear data is to undo the Log C curve of images recorded in QuickTime ProRes 4444, ProRes 4444XQ (or DNxHD 444) files.
ARRI's online LUT Generator can create LUTs to convert Log C images to linear sensor data. When the Log C data is converted to linear sensor data, black (corresponding to zero exposure) will be represented by the value 256/65535. This sensor black level is the mean of all pixels. Because of read-out noise, single pixels may be above or below this value. The standard deviation of the read out noise is approximately 2.5 meaning that the offset of 256 is more than high enough to encode the full noise amplitude (usually one assumes a range of three times the standard deviation or ± 8 code values). Using the parameters for linear scene exposure will map the black value, as expected, to 0.0. With the noise, however, single pixels will come out as negative values. When those values cannot be preserved and one does not want to clip them, a small offset of 8/65535 should be added to the relative scene exposure factor. This is equivalent of adding flare to the image data. The amount of flare expressed relative to the scene white will vary with the exposure index. It ranges from 0.1% (for EI 200) to 0.8% (for EI 3200). The flare should be subtracted before the images are converted back to Log C.
Logarithmic Encoded Data
The Log C curve was first introduced with the ARRIFLEX D-20 camera. It's an encoding with a transfer characteristic similar to that of a scan from negative film. Because of the fundamental differences between digital cameras and negatives, however, the color characteristics remain different.
Display-Ready Video Encoding for Rec 2020/709 Monitors
The Rec 2020 or Rec 709 output of the camera is either used for on-set preview or for when the program is edited for television without extensive color correction. Those images are displayed without any further transformation. While this simplifies the post production workflow it reduces the possibilities in color correction. The images have been tone-mapped and transformed into the target color space.
The tone-map curve is applied to the Log C data. This transform is also available as a LUT for post-processing of Log C footage. While this provides a nice image on a video monitor, it also means that some information has been "squeezed out" of the transformed image. The matrix transform applied immediately following the tone-map curve results in an image transcoded into the target color space, but if the original image contained extremely saturated colors, this may also bring a loss of color. If a transformed color is outside the gamut of the display, it will be mapped or clipped to an in-gamut color.
Beginning with ALEXA SUP 3.0 until SUP 7.0 ARRI offered a film style matrix that could be applied to the Log C output. The same transform was also available as a 3D-LUT for post-processing of Log C footage.
The film style matrix makes the color characteristics of the Log C image similar to negative film scanned on an ARRISCAN. The matrix is most usefully applied when the data is previewed or converted with a print film emulation (PFE) . This is the common workflow in Digital Intermediate where the PFE is applied as a 3DLUT in the display path.
The film-matrix can in some situations achieve a bigger color separation and therefore make the keying of some shots easier.
Due to new developments in color grading and digital workflow our film style matrix has been discontinued.
The ALEXA or AMIRA cameras store per-frame and per-shot metadata. This metadata is present in all recorded formats - onboard and in-camera formats ARRIRAW, MXF/ARRIRAW, ProRes and , DNxHD. Per shot metadata is also written to an FinalCut Pro XML file and an AVID AAF when recording to ProRes or DNxHD respectively.
Please refer to our Metadata White Paper
Also have a look at ARRI META Extract, an metadata extraction tool for MacOS and Windows.