CyclesSessionOptions
Controls a Cycles rendering session (which is behind a CyclesViewarea).
Inheritance
Code
XML encoding
<CyclesSessionOptions progressive='TRUE' background='FALSE' progressiveRefine='FALSE' displayBufferLinear='FALSE' experimental='FALSE' threads='0' samples='65536' tileSize='64 64' startResolution='64' logFeature='' />
Classic encoding
CyclesSessionOptions { progressive TRUE background FALSE progressiveRefine FALSE displayBufferLinear FALSE experimental FALSE threads 0 samples 65536 tileSize 64 64 startResolution 64 logFeature [""] }
Interface
Filter: X3D only | Avalon only | All
id | Name | DataType | PartType | Default | ValueType | Description |
---|---|---|---|---|---|---|
progressive | SFBool | inputOutput | TRUE | Render progressively or in offline mode. Progressive rendering is geared towards fast feedback with few samples, offline rendering is geared towards achieving optimal results with a given sampling budget. | ||
background | SFBool | inputOutput | FALSE | Render in background without displaying continuous updates. | ||
progressiveRefine | SFBool | inputOutput | FALSE | Do not render tile after tile but the whole image at once, even in non-progressive (offline) mode. Slightly slower but can be interrupted any time (e.g. when noise is low enough). | ||
displayBufferLinear | SFBool | inputOutput | FALSE | Use linear display buffer internally. In fact this toggles between one-byte-per-channel and one-half-per-channel internally. (The result is still displayed in the Viewarea's target, probably a RGBA8 back buffer.) | ||
experimental | SFBool | inputOutput | FALSE | Use experimental feature set. | ||
threads | SFInt32 | inputOutput | 0 | Number of threads to use. 0 uses number of hardware threads available on current machine. | ||
samples | SFInt32 | inputOutput | 65536 | Number of samples per pixel to use for rendering. | ||
tileSize | SFVec2f | inputOutput | 64 64 | Tile size to use. For CPUs, choose a tile size that results in at least as many tiles as there are threads, because CPUs work on one tile per thread. For GPUs, choose larger tile sizes (>128), as they work on one tile per GPU. | ||
startResolution | SFInt32 | inputOutput | 64 | Resolution of the first 'pixelated' frame. Lower values allow smoother navigation and faster feedback. | ||
logFeature | MFString | inputOutput | state, child, parent, route, eventIn, eventOut | controls the logging of changes, state: log state changes (e.g. live), child: log child add/remove, parent: log parent add/remove, route: log route add/remove; eventIn: log receiving of events, eventOut: log sending of events: guiView, runtime system should create node-view, guiEdit: runtime system should create node-editeverything: log everything |