Can you copy panorama auto align transform edits between one stack of layers to another?

Auto-alignment relies on overlap of the individual images and recognisable detail i.e. shapes in that overlap that the application can use to align the images. Star-fields with no foreground detail, and similarly water with no land detail, tend to fail on that second element i.e. insufficient detail in the overlapped areas. Individual stars tend to be seen as noise and ignored in the auto-alignment process.

If you can recognise the overlap and the detail, then you could manually align the remaining images using transform>distort, and c.pfaffenbichler’s method might get you close to a starting point for that. 

Dave

Astro/star panos are a challenge.  It is probable that the geography and the movement from the automated-pano head is part of the issue but it is still something that you can correct.

Frankly, I do not see one satisfactory or specific way, in PS or LR, that would quickly ‘fix’ this.  The ways I do know require using a bit of “Transform” and “Perspective” on some frames/layers - it can be tricky.  But I can offer a couple of suggestions:  Firstly,  I would refer to Peter Baumgarten at CreativeIsland photography.  He knows everything about astrophotography and how to manage it in post to get your stars aligned.  

Secondly, I would refer to the pano-head mfg website for that specific model/version.  They usually have some valuable tech-tips.

Thirdly, try looking deeper into your camera-brand for their own software and options.  Some brands have exclusive built-in settings and exclusive software tools that help immensely.

Photographer to photographer - I completely understand and wish you great success with the pano!

I suspect the images might simply not provide enough »detail« for Photoshop’s auto-alignment. 

Photoshop just isn’t specifically targeted at astro-photography. 

Copying the warping parameters from another montage is not possible, as far as I can tell. 

As a longwinded manual work-around you could place the images of a successful panorama as Smart Objects, transform them individually according to the successfully transformed Layers and then replace their contents with the corresponding images from the series that did not get processed successfully.