< Previous | Contents | Manuals Home | Boris FX | Next >
Normally, when a shot is read, the image bit depth (8-bit, 16-bit, floating) is determined by the "Process Depth" setting on the Shot Setup panel (which is also found on the Rez tab of the image preprocessor). That turns out to be problematic for disk caching, where shots can be open from multiple places and settings can change without warning.
The disk caching system uses one of two different approaches, controlled by a preference (Cache only 8-Bit versions):
- Always caching a shot in its native bit depth
- Always caching every shot at 8-bit depth
Note that the setting does not matter if all your shots are 8-bit, for example, a typical AVI, MOV, Targa, or JPEG!
Using the native setting (default) preserves the full original content of the files for sure, and is highly recommended for applications such as stabilization or lens undistortion, where modified images will be output.
Using the 8-bit-only setting will cause the images to be converted to 8-bit, then stored in the disk cache. This will speed improve interactive performance and minimize the size of the disk caches. Do not use if for stabilization or shot undistortion, however.
Fine point : if you have half-float or floating-point images and 8-bit depth, the Range Adjust control affects how the images are converted to 8-bit format. Normally, floating-point values range from 0..1 and the range control can be left at zero. If you adjust the range setting, it will affect all images as they are
read and placed in the disk cache. If you change the value after reading some but not all images, you can wind up with a mixture in the disk cache, which is not good. If that occurs, you should flush the cache using Script/Flush Shot's Caches.
When you change the preference, all currently-open shots will be converted (and flushed in the process). Other disk caches are unaffected until the shot is reopened.
©2024 Boris FX, Inc. — UNOFFICIAL — Converted from original PDF.