No. Duplicating one exposure and “randomizing” it won’t give the same benefit as stacking real, separate exposures.

Stacking works because the true signal (stars, nebulae, etc.) is consistent from frame to frame, while noise varies randomly between exposures. When you align and combine multiple real images, the repeated signal reinforces and the random noise averages down.

With only one image, you don’t have independent samples. Any software trying to invent or randomize noise is just guessing—it cannot know which pixel variations are real faint stars and which are noise. If it suppresses low-level variations too aggressively, it can remove actual stars (“star eating”) and reduce dynamic range.

So the practical answer is: no software can turn one exposure into the equivalent of a properly stacked set of multiple exposures. The solution is still to capture multiple frames and align them accurately, including rotation if needed. That separate data is what makes noise reduction by stacking possible.