there is much new features in the blender 2.8 beta> We've had text overlays in the VSE since 2.6. Does anyone have any new or current work arounds or official ways to get various fonts into the VSE?


2.8 now has font selection for the VSE and the change was backported to 2.7 branch as well.


We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.


There's the generic workaround where you make text in the 3d viewport, assign the font and settings, set the camera, and import it as a strip into the vse as a scene strip (then just assign as overdrop). It technically gives you the greatest possible control over your font, but it's cumbersome and we'd all rather just have the simple standard presets like other vse.

My other work arounds to date have been to create png overlays in inkscape and either save multiple formatted texts to a single png image (then transform and crop as necessary in the vse) or do 1 image per overlay and simply import into the vse and assign as an overdrop. Inkscape is quite nice when it comes to creating text since: you can scroll through the fonts dynamically and preview what they look like, you have an infinitely large workspace to see and micromanage things in, the text is made of vectors, you can apply linear and radial gradients as well as textures and specialized gradients, apply fill and stroke and gradients to those as well, and the best part is you can apply a myriad of confusing yet amazing preset filters to these and then change the font/size/style after the fact and have them update...cause these aren't raster images (Vectors are awesome!).

And the remaining work around was to use ffmpeg to render the text style/font from a subtitle file like a .ass file as png overlays. So technically if I created a .ass file or some other subtitle file with the fonts/styles and desired texts recorded, I could just render them all out as a series of transparent png text overlays with ffmpeg. This rendering process takes a long time though, but with blender we're used to planning for long renders anyways (planned renderhood).

ffmpeg -ss 00:17:11.88 -f lavfi -i "color=color=white@0.0:size=1920x1080,format=rgba,subtitles=subfile.ass:alpha=1" -frames:v 1 output.png

So you assign a time in the order of hours/minutes/seconds/milleseconds (17 hours 11 seconds 88 milleseconds) to match one of the times listed in the sub file. You change the size= to the desired resolution. And replace subfile.ass with the subfile you are using to do this. So you can cheat and use an existing subfile and just replace the texts with ur desired texts and go this route. And in this case it's shooting out a single transparent png image called output.png

If you want more information on any of these or pictures, I'll see what I can do.


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.