StableZero123_Conditioning_Batched - ComfyUI Built-in Node Documentation

The StableZero123_Conditioning_Batched node prepares conditioning data for generating a 3D model from a single input image. It encodes the image with a CLIP vision model and a VAE, combines the visual features with camera embeddings built from elevation and azimuth angles, and produces positive and negative conditioning plus a latent tensor for a batch of samples. When batch_size is greater than 1, the elevation and azimuth angles are increased by their batch increment values for each item in the batch.

Inputs

Parameter	Description	Data Type	Required	Range
`clip_vision`	The CLIP vision model used for encoding the input image	CLIP_VISION	Yes	-
`init_image`	The initial input image to be processed and encoded	IMAGE	Yes	-
`vae`	The VAE model used for encoding image pixels into latent space	VAE	Yes	-
`width`	The output width for the processed image (default: 256)	INT	Yes	16 to MAX_RESOLUTION (step of 8)
`height`	The output height for the processed image (default: 256)	INT	Yes	16 to MAX_RESOLUTION (step of 8)
`batch_size`	The number of conditioning samples to generate in the batch (default: 1)	INT	Yes	1 to 4096
`elevation`	The initial camera elevation angle in degrees (default: 0.0)	FLOAT	Yes	-180.0 to 180.0
`azimuth`	The initial camera azimuth angle in degrees (default: 0.0)	FLOAT	Yes	-180.0 to 180.0
`elevation_batch_increment`	The amount to increment elevation for each batch item (default: 0.0)	FLOAT	Yes	-180.0 to 180.0
`azimuth_batch_increment`	The amount to increment azimuth for each batch item (default: 0.0)	FLOAT	Yes	-180.0 to 180.0

Note: The width and height values must be multiples of 8, since the node divides these dimensions by 8 internally when building the latent tensor.

Outputs

Output Name	Description	Data Type
`positive`	The positive conditioning data containing the image embeddings and camera embeddings for each batch item	CONDITIONING
`negative`	The negative conditioning data with zero-initialized embeddings	CONDITIONING
`latent`	A zero-initialized latent tensor with dimensions batch_size x 4 x height/8 x width/8, along with batch indexing information	LATENT

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 94fc53dace8f294a746c47f8aa0da145f3e7beeb77a95912a38f0037ac094292

​Inputs

​Outputs

Inputs

Outputs