Would it be possible to tweak the program so, that the overlay function could be used with and after the "switch to direct zoom"-function. The frames can be aligned with the zoom and the synchronization separately and the overlay saved with the "Save a sequence of images" function...
- joan
- Admin
- Offline
Interesting…
An addition on this would be the ability to drag the image around even when it's at normal view (zoom 100%).
With the combination of these two, one could directly adjust/align one image over the other in a very simple way.
Yes, that´s the idea 
. And also there would be a way of aligning painted tracking curves on top of each other. By using two copies of the same video there is also a way of building simple cinetograms with the overlay-tool. I just overlay multiple copies of pictures. If one could vary the transparency of the overlay, then the quality would be much better?
- joan
- Admin
- Offline
This has been implemented in the last version (0.8.12) - by allowing whole image dragging even at normal size (no zoom). (and application of the change for the copy on the other screen)
It is only active when blending between the two videos is active for now…
Also, currently you can't zoom out to less than 100%… maybe it is limiting the usability of the feature…
Thoughts for improvements ?
Thank you, Joan for this so far I will check it out !
OK. What would still be needed: the possibility to make a sequence of overlay photos. Now the sequence moves only in the other video... Would it be possible to increase or decrease the transparency of the overlay to improve visibility?.
- joan
- Admin
- Offline
Alexander wrote:
What would still be needed: the possibility to make a sequence of overlay photos. Now the sequence moves only in the other video...
Hum… Trying the various export buttons on the individual videos indeed reveal several inconsistencies.
- export image : both images blended.
- export sequence : sequence of the primary video, blended with a fixed frame of the other video.
- export video : just the primary video, the other video is not blended at all.
- special exports (key images, time freeze) : same as export video.
Now there is some logic to the fact that the buttons on the individual video screens are only exporting this specific screen's video and images.
It keeps the function of the button similar to when there is only one screen.
Another glitch : when blending is enabled and you use the "dual export" (on common controls bar), it exports the side by side video, including the blending, which is redundant.
What about:
- If the blending is enabled, then the "dual export" video will actually export a single, merged video.
The logic would be that :
- Export buttons on the individual screens works on the individual videos. (same function as if there was only a single screen, no blend)
- Export buttons on the common controls bar :
--- If no blending, works by creating side by side images or videos.
--- If blending, works by creating blended outputs.
Then, a sequence button could be added to the common controls bar to implement sequence export. (side by side or blended depending on the current option)
?
Sounds very logical to me. That was the function I expected...;) I expected WYSIWYG on the blended single screen when I tried to export a sequence. Your solution seems good.
9 2011-02-04 07:33:56 (edited by Alexander 2011-02-04 07:36:15)
...Hello Joan. I got one more idea on this overlay function: would it be possible put 2-3 markers or marked areas (with some distinct objects) at the beginning of each video and the markers would "fixate" the frames relative to each other. The manual movement of frames would be automated. This combination of overlay, moving origin of coordinates would make visual analysis and comparison A LOT easier...
And overlaying the graphical objects without blending them into the video would be a great feature, too. Iam sorry for sounding like a kid before Christmas, wanting more and more...;)
- joan
- Admin
- Offline
Well, I think this would probably overlap too much with the superposition sub project.
I had done some "proof of concept" code last year to test the theory so I know it can work.
The idea of the current blend function was to provide a quick tool until this more general function is implemented, so I don't want to over complexify it or divert too much energy .
The quick blend will be useful when the two videos haven't been shot at the same place though.