Performance(パフォーマンス)

参照

Panel(パネル):

Render ‣ Performance

Properties that affect render speeds or memory consumption. There are several presets available to help choose between different trade offs:

Default(デフォルト):

Balances memory saving and faster rendering settings.

Faster Render:

Uses settings to render faster at the cost of higher memory consumption.

Lower Memory:

Uses settings to decrease memory consumption at the cost of slower renders.

Threads

Threads Mode

Method to determine the maximum number of CPU cores to use while rendering.

Auto-Detect:

Automatically chooses the amount of threads to match the number of logical processors on your computer.

Fixed(固定):

Manually choose the maximum number threads to use for rendering. This can be useful for example, if you want to use your computer while rendering, you can set the property to a thread count lower the amount of logical processors on your computer.

Threads

The maximum number of CPU cores to use simultaneously while rendering.

メモリ

Tile Size(タイルのサイズ)

Render high resolution images in tiles of this size, to reduce memory usage of render passes. Tiles are cached to disk while rendering to save memory.

Lowering the tile size also reduces texture cache memory usage, as only the image textures needed by each tile will be loaded into memory.

Tile size has an impact on performance, and the default 2048 pixels provides a good balance for most scenes. When running out of memory, lowering to 1024 or 512 pixels helps reduce memory usage, typically at a small performance cost.

Texture Cache

Reduce memory usage and startup time on scenes with many image textures, at the cost of a small rendering performance impact and increased disk space usage.

Uses .tx files that store images in a tiled, mipmapped format optimized for loading just the tiles and resolutions needed for rendering. By default, files are stored in the blender_tx/ folder next to the image file. This can be changed with the Texture Cache file path preference.

Auto Generate

Automatically generate and update .tx files for images when rendering. Files are regenerated when the source image is modified.

Generate All

Generate .tx files for all images used by materials in the blend file.

Generating Tx Files from the Command Line

Standard .tx files generated by other renderers and OpenImageIO's maketx are recognized, To generate Cycles native .tx files from the command line, the following commands are provided:

# Generate tx files for all images in a blend file.
blender scene.blend --command maketx
# Generate tx file for a specific image file. This requires specifying the
# color space if it can not be determined automatically.
blender --command maketx image.png --colorspace sRGB

Acceleration Structure

Use Spatial Splits

Spatial splits improve the rendering performance in scenes with a mix of large and small polygons. The downsides are longer BVH build times and slightly increased memory usage.

Use Curves BVH

Use a special type of BVH for rendering curves. The bounding boxes are not axis aligned allowing a spatially closer fit to the curve geometry. Disabling this option will reduce memory, at the cost of increasing curve render time.

BVH Time Steps

Split BVH primitives by this number of time steps to speed up render time at the expense of memory.

Use Compact BVH

Use a more compact BVH structure, which can reduce RAM usage but render slower.

注釈

The Acceleration Structure panel will not be present when using an Optix compute device.

Final Render

Persistent Data

Keep render data in memory after rendering for faster re-renders and animation renders at the cost of extra memory usage while performing other tasks in Blender.

When using multiple View Layers, only data from a single view layer is preserved to keep memory usage within bounds; however, objects shared between view layers are preserved.

Viewport(ビューポート)

Pixel Size (ピクセルサイズ)

Option to control the resolution for viewport rendering. Allows you to speed up viewport rendering, which is especially useful for displays with high DPI.

Compositor(コンポジター)

Device

The device used for compositing.

CPU:

Use the CPU for compositing.

GPU:

Use the GPU for compositing.

Precision GPU

The precision of compositor intermediate result.

Auto(自動):

Use full precision for final renders, half precision otherwise.

Full:

Use full precision for final renders and viewport.

Denoise Nodes

Denoising Device

The device to use to process Denoise nodes in the compositor.

Auto(自動):

Use the same device used by the compositor to process the denoise node.

CPU:

Use the CPU to process the denoise node.

GPU:

Use the GPU to process the denoise node if available, otherwise fallback to CPU.

Preview Quality

The quality used by Denoise nodes during viewport and interactive compositing of a render if their quality is set to Follow Scene.

High(高):

Produces the highest quality output at the cost of long processing times.

Balanced:

Balanced between performance and quality, typically processing in half the time as High, while retaining most of the quality.

Fast(高速):

Produces an output quickly at a noticeable cost of quality.

Final Quality

The quality used by Denoise nodes during the final render if their quality is set to Follow Scene.

High(高):

Produces the highest quality output at the cost of long processing times.

Balanced:

Balanced between performance and quality, typically processing in half the time as High, while retaining most of the quality.

Fast(高速):

Produces an output quickly at a noticeable cost of quality.