Models & Parameters

Denoise Settings: Denoising Strength and Sampling Steps

Also called denoising strength, sampling steps, denoise strength, image strength, steps

Denoising strength decides how far your input image is pushed back into noise before a diffusion model starts to denoise it again, which sets how much of the original survives. Sampling steps decide how many passes that rebuild takes. Together they are the two dials behind every image to image edit.

What the process looks like from the inside

Generation runs backwards. The model starts from noise and denoises it in stages, and at each stage it predicts what the picture underneath would look like with a little less noise. Text to image starts from pure noise, so there is nothing to preserve. Image to image starts from your picture with noise added on top, and how much noise gets added is exactly what denoising strength means.

At 1.0 your input is buried completely and only the prompt survives. At 0.3 a faint version of the original is still visible through the noise, so the model rebuilds around the shapes it can still see. Every practical decision about image to image is a decision about how much of the source you want to remain legible to the model.

Picking a strength

StrengthWhat survivesUse for
0.15 to 0.3Everything, subtly alteredGrading, texture, film grain, small cleanups
0.35 to 0.55Composition and poseRestyling, changing medium, repainting
0.6 to 0.75Rough shapes onlySketch to render, loose layout reference
0.8 to 1.0Almost nothingNew image with a hint of the old

The failure modes sit at both ends and look nothing alike. Too low and the output is your input again, which people misread as the prompt being ignored. Too high and identity goes: the face is a different person, the product no longer matches the packaging, and the pose you were trying to keep has drifted.

Steps, and the trap of multiplying them

Steps are how finely the path from noise back to image is sliced. More slices means each one is a smaller, safer jump, up to the point where the result has converged and further slices change nothing you can see.

The trap is the interaction. In most image to image implementations the model only runs the fraction of the schedule that strength asks for, so a low strength job silently gets very few real denoise passes. Twenty steps at 0.25 strength is five effective passes, which is why gentle edits so often look soft and unresolved. The fix is not more strength, which would change more than you wanted, but more steps: 40 to 60 at low strength costs little and resolves properly.

What no amount of either will fix

Neither dial adds knowledge or corrects layout. If the model cannot draw a convincing hand it will denoise its way to a bad hand at every strength and every step count. If the composition is wrong, the fix is a different seed, a structural reference, or an inpaint pass on the offending region.

One habit that saves a lot of credits: when a result is close but not right, decide first whether you want more change or cleaner change. More change is strength. Cleaner change is steps. Reaching for the wrong one is the most common way an editing session stalls.

The prompt for this

A starting point that reliably produces the effect. Adjust the subject and setting; keep the technical clauses.

Repaint this photograph as a 1970s oil painting, visible brush texture, warm ochre palette, keep the original composition

Try Denoising Strength yourself

Open the generator with a starting point already filled in.

Frequently asked questions

What denoising strength should I use?
Match it to how much you want changed. 0.2 to 0.35 for a grade or texture pass that keeps every detail, 0.4 to 0.55 to restyle while holding the composition, 0.6 to 0.75 when the source is only a rough layout, and 0.8 upward when you are effectively generating fresh. Inpainting inside a mask usually wants 0.75 to 1.0.
What is the difference between denoising strength and sampling steps?
Strength sets how far back toward pure noise your image is pushed. Steps set how many increments are used to walk it forward again. Strength decides how much changes, steps decide how cleanly the change is rendered. They are unrelated dials that people routinely swap.
How many sampling steps do I need?
For classic samplers 20 to 30 covers almost everything, and past about 40 you are paying for differences you cannot see. Distilled and turbo models are trained for 4 to 8 and get worse above that, not better. Raise steps only when output looks noisy or unresolved.
Why does my output look identical to the input?
Strength is too low. Below roughly 0.15 the model does not have room to change anything, so the prompt appears to be ignored. People often rewrite the prompt at this point when the fix is one number.
Why does a low strength result look mushy or unfinished?
Because effective steps are strength multiplied by steps in many interfaces. At 0.25 strength and 20 steps the model only runs about five real passes, which is not enough to resolve detail. Push steps to 40 or 50 when working at low strength.
Does raising steps fix a bad composition?
No. Layout is decided in the earliest passes, so adding steps at the end only refines what is already there. A wrong composition is a prompt, seed, or structural conditioning problem, and more steps just makes the same mistake sharper.

Related terms