Closed isben closed 11 months ago
I finally figured out the answer to my question above. For those wondering, you need to use an Empty Latent Image and a Noisy Latent Image and combine then via a Inject Noise node together with the strength provided by a Get Sigma node. This can then be sent to a KSampler Advanced with add_noise disabled.
Hello @BlenderNeko,
I tried injecting noise created by your
Noisy Latent Image
node into aKSampler (Advanced)
node and (naively?) assumed that by setting theadd_noise
parameter to disabled I would be able to obtain the same effect as by using anEmpty Latent Image
withadd_noise
enabled (using the same seed). This is at least what logic suggests to me. To my surprise, whenadd_noise
is disabled, I always get the same image result, whatever Latent I use as input to the KSampler. The result seems to have no relation with what is usually generated whenadd_noise
is enabled and remains the same if I use different seeds in theNoisy Latent Image
. I checked the generated noise by decoding and previewing the Latent and it obviously varies.The example you provide is making use of
Inject Noise
andKSampler (Advanced)
in a manner similar to what I'm trying to achieve. Can you explain why it seems to work in one case and not the other?