Combine Pixaroma

What it does

Takes any two inputs and merges them into one. Images and video frames are stacked into a single bigger batch; latents are batched the same way; numbers and text are gathered into a list. If different-sized images come in, the second is resized to match the first.

What you can join

Both inputs must be the SAME kind of thing:

• image + image -> one bigger image batch (for example 3 frames stacked together)
• video frames + video frames -> a longer clip
• latent + latent -> a batched latent
• number + number -> a list, like 0 then 1 then 2
• text + text -> a list of words

Good to know

• You can't mix kinds. An image plus text, or a latent plus a number, stops with a clear message - wire two of the same kind instead.
• An empty input is fine - Combine just passes the other side through. That is why it works as a loop accumulator from round 1.
• Different-sized images: the second one is resized to match the first (whatever you wire into any1 sets the size).

Using it in a loop

Combine shines as the 'pile-up' node inside a loop. Wire the running total into any1 and the new round's result into any2, then carry the output back through Loop End. Each round adds onto the pile. On the very first round one side is empty - Combine just passes the other side through, so it works from round 1 with no special setup.

Outputs

batch

The two inputs joined together. For images and frames this is one larger batch; for numbers and text it is a list.