MLT框架。 在视频上插入放大文字

Question

Need to achieve effect something like that: put zoomingIn text on video background (black with some animation) between other videos. Demo: https://www.dropbox.com/s/ehqwkjgu5u5e9rp/4%20Cultivar%20Road%20Brampton%2C%20Jessi%20Sandhu.mp4?dl=0

Trying to make is using dynamictext and affine filter:

"1.mp4" -mix 7 -mixer luma "2.mp4" -mix 7 -mixer luma "bg.mp4" -attach-clip dynamictext in=0 out=145 "text.txt" -attach-clip affine in=0 out=145 -attach-clip affine valign=middle halign=center scale=1 fill=1 geometry="0=10%,10%:80%x80%:0%;36=0,0:100%x100%:100%;109=-10%,-10%:120%x120%:100%;145=-10%,-10%:120%x120%:0%"

But i have unexpected results with infinite video and no animation at all...

Answer 1

Your syntax for the dynamictext filter is incorrect. The dynamic text filter requires an argument like this:

-attach-clip dynamictext:"This is some text"

Also, the dynamictext filter does not accept a text filename. The actual text must be specified as the filter argument.

Also, your syntax for the affine filter is incorrect. The geometry is applied to the encapsulated transition. So the geometry should be specified with: "transition.geometry=..."

I think your best path to success will be to use the qtext producer.

Try something like this as an example:

melt qtext:text.txt in=0 out=145 fgcolour=white size=200 -attach-clip affine valign=middle halign=center scale=1 fill=1 background=color:black transition.geometry="0=10%,10%:80%x80%:0%;36=0,0:100%x100%:100%;109=-10%,-10%:120%x120%:100%;145=-10%,-10%:120%x120%:0%"

Check the documentation for qtext for other parameters you might find useful: https://mltframework.org/plugins/ProducerQtext/