3
$\begingroup$

I am creating a video using the VSE where I show a sequence of screenshots of two devices side-by-side.

Unfortunately, one sequence of screenshots is from a Nexus 7 in portrait mode (resolution 800x1280) and the video I am making is 1920x1080. This means that the screenshot must be scaled down to fit on screen.

I have a sequence of the screenshots and some overlay elements grouped into a meta strip. Each screenshot has the Image Offset box checked so it is pixel-for-pixel to start with.

The meta strip has a transform effect strip applied to it and I apply a uniform scale of 1080x1280 to fit the meta strip on-screen. I do this so that all the screenshots in the meta strip benefit from the same scale and transform, and I can adjust the position of this screenshot sequence in the canvas without having to edit 10 different transform strips (and a growing number of overlays). Unfortunately, the top of the screenshot gets chopped off.

How do I prevent the large images in the meta strip from being cropped before the transform effect can shrink them onto the canvas?

screenshot of .blend file illustrating problem

(now has copies of the images packed into the .blend, but you'll still need to save them and point the image strips at them)

$\endgroup$
  • $\begingroup$ Could you provide some example Screenshots of your setup or a small example blend file? $\endgroup$ – Samoth Feb 27 '16 at 20:35
  • $\begingroup$ Done, Samoth, although I don't think the images will be embedded in the .blend . $\endgroup$ – Mutant Bob Feb 29 '16 at 18:20
  • 1
    $\begingroup$ @MutantBob Load the images into a UV/Image edtor window. Then pack them ( File->External_Data->Pack_All_Into_.blend .Resave and Re-Upload your file. $\endgroup$ – cegaton Feb 29 '16 at 18:33
  • 1
    $\begingroup$ maybe related: blender.stackexchange.com/questions/42094/… $\endgroup$ – cegaton Feb 29 '16 at 18:33
  • $\begingroup$ I did refer to that question while I was trying to solve my problem, cegaton, but none of the techniques in the answers appear to be relevant to scaling down a strip or meta-strip that has more pixels than the scene. $\endgroup$ – Mutant Bob Feb 29 '16 at 19:19
1
$\begingroup$
  • Consider a preprocessing pass where you just create new Scaled images at the intended final size with a possible transparent border.

  • Suppose the Blend User had 100 images at dimension (X,Y) where for example (X = 1280, Y=800) and their final render size was .5 times (X,Y). The User might just create a [2nd new] VSE scene to quickly create a new sequence of images at the new size for final use. VSE is used to batch scale images into a new file. GIMP might do the same.

  • I might even place the images in the upper right hand corner and leave a transparent border in PNG format. Suit borders to your needs. Your already use overlays. Thus no further scaling need take place because of a transparent border. Then many or all of the files would have the same size ... the offset would be different.

  • Since I cannot see your VSE strips I am not addressing them directly. So we cant see limitations, if any, of the strip manipulations. They may be some more space efficencies possible if all that were known, yet its not a requirement.

  • I have not considered file names.

  • One could claim their is some waste of hard disk space or RAM. Space may be cheaper than time.

$\endgroup$
  • $\begingroup$ I don't understand this answer. Should I conclude from the first two points that Blender just cannot do what Mutant Bob is asking, and therefore first one needs to scale down the screenshots? I don't see what your other points have to do with the question. $\endgroup$ – Daan Michiels Mar 28 '17 at 3:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.