UXP - Full control over Premiere timeline - when?

Premiere Pro
,
1

I am a relatively new creative cloud developer. I need to be able to do is add video, text and images to the timeline and manipulate the media (resize, animate, change start/end points on timeline, etc) based on data in a JSON file.

So far, the UXP API is pretty limited. It’s surprising that you can’t add media to the timeline with start and end points in one shot. I am not sure what else folks are using the API for, but manipulating the timeline seems like the PRIMARY use case of the API.

The UXP API has been in beta since 2024, but basic features still haven’t been released. CEP is deprecated, but UXP still can’t perform the most basic of video editing functions. What is the timeline? What features are on the backlog? I don’t want to invest in CEP if a full UXP release is right around the corner.

Ideally, I would like for these tasks to be performed in Premiere - but if there’s no plan for implementation we will need to look at other options.

Other feedback:
Some of the things we tried to do in Premiere were not possible, so we wrote another script for AfterEffects - it just generates animated captions and annotations based on a text file. We’ve optimized it quite a bit, but AfterEffects is really slow at constructing the document. It still takes seconds to minutes for dozens of captions.

Still no SVG import in Premiere? I use Inkscape for all my SVG editing. I avoid Illustrator because it runs slowly and consumes a lot of recourses. I only use it when I need to prepare documents for print, which is rare because 99% of what I do is for the web. Having to convert SVG files to PNG or AI is an extra step and quite wasteful.

Where I am at now: Moving towards rendering video directly in the browser because of combinations of the issues listed above. I was surprised to learn that rendering a 4K captions video in the browser is FASTER than trying to set up a composition in AfterEffects to do the same thing. AfterEffects is a powerful program for time-based imaging, but it has some serious performance issues.

In Summary:

The documentation for UXP, CEP and ExtendScript is incomplete and at times, innacurate. The implementations also appear to be incomplete, buggy and have performance issues.

2

Two shots remain required (just like in ExtendScript) :

  1. set appropriate in/out points on source projectItem.
  2. add the projectItem to the timeline.

UXP still can’t perform the most basic of video editing functions.

Specifics, please?

That’s not what we’re hearing from the dozens of partners already shipping UXP plugins, but you’re correct that there is still functionality missing, from full 1:1 parity with ExtendScript.

3

Thanks for your reply! I have outlined a few use cases. Please let me know if I might be overlooking anything that would help solve my issues:

  1. I have a JSON file with image references. The images have been dragged into the project. My script put the images at the appropriate times into the timeline, but there doesn’t seem to be a way to adjust the endpoints in UXP from what I was reading. I’d be glad to know if I am mistaken!

  2. There is a text track in the timeline, look at the existing text track and line up the selected clips with each item in the text track (similar use case to where I was running into issues with #1).

  3. Create captions with word-level highlighting and animation, I have an AfterEffects script that does this. I had a proof-of-concept that produced the results I wanted in a few minutes of work, then I spent the better portion of a week optimizing the script so it could run on real-world examples. The performance still isn’t up to my standards - it takes a minute (with caps lock ON) to generate captions for a 5 minute video (and that’s not even rendering). There doesn’t seem to be a way to prevent AE from rendering stuff, telling users to press CAPS LOCK or it will take a long time isn’t a viable solution for us. I still feel even the “faster” 1-minute time is too slow. My reading tells me that there isn’t a ton of control of telling AE to render or not from ExtendScript, which seems to be where the main slowdowns are occurring. I am using undo groups and disabling previews/using fast preview where possible.

Dragging a SRT file into the track in Premiere works GREAT and It’s really fast. Adding capabilities to do per-word animation would be nice. I know there are caption generator tools that work very well for general use cases, but ours is education-specific and rather specialized.

4
  1. The premiere_api example shows how to change the start/end points, of a given trackItem. Note: a trackItem’s in and out points refer to the time range, within the source projectItem; a trackItem’s start and end, refer to the trackItem’s position, within the sequence.
  2. Text items = not yet supported via UXP. They were never supported via ExtendScript, so…parity achieved! :slight_smile:
  3. Captions and Transcription = same situation.