Before we start we would like to thank the KDE e.V. (travel and hosting support) and the Carrefour Numérique (workspace) for helping make this events happen.
Paris Sprint 2018
From the 25th to the 29th of April, 5 members of our team had a sprint in Paris to focus on the future of Kdenlive. And it was fantastic! We met for the first time in person, made friends and worked a lot! But let’s start with the beginning. We were warmly welcomed by Mathieu at the Carrefour Numérique, part of the Cité des Sciences in Paris.
On the first day after a brief introduction, the team started working on the vision of the project, defining objectives, discussing technical issues and schedules and by the end of the day came up with a roadmap (see below) with a clear set of short, mid and long term goals post the refactoring release.
One of the biggest changes proposed for the refactoring release is the deprecation of hybrid clips, which means that now there is a strict clip structure in which you can’t move video clips to audio tracks and viceversa (see video below). Also a high priority list of tasks focused on stability (but with minor new features) was made for releasing the refactoring branch for the 18.08 release schedule which includes:
- Make a stable release.
- Don’t break compatibility with older versions.
- Finish implementing strict mode.
- Folder import. (New feature)
Besides all the planning and heavy bugfixing of the refactoring branch during the 4 days of the sprint. We also took steps towards improving our communication infrastructure by adding multilingual support to the website and started translation in French and Italian, defined fixed dates for the Kdenlive Cafés, created a bridged Telegram group with #kdenlive at IRC (also working on bridging it with Matrix) and a francophone Telegram group, and discussed a strategy for dealing with the bugtracker after the refactoring release.
We also had 2 public events:
- A meeting for people interested to help the project, which resulted in a nice contribution by Camille: a rewrite of the developer info in the wiki. We also received a patch adding the possibility to manage and download keyboard shortcut templates.
- A public presentation where we met several very enthusiasts users.
Some highlights of things we did during the sprint :
Audio and video are now always split for a cleaner workflow. You can see a preview in the following video:
The monitors now have a nicer overlay toolbar and support multiple layout guides:
At Libre Graphics Meeting
Roadmap
The following is our proposed roadmap for the coming releases after the refactoring is finished.
Short term
- Show audiowave in monitor
- Trimming operations
- Nested timeline
- Same track transition
- Corruption fixes
- Check consistency before save
- Refresh button
- Info dialog for clips, to be able to set in/out points precisely
- Highlight on snap
- Clip locking
Mid term
- Audio workflow
- Multicam
- Fast render (multicore)
- Secondary color correction
Long term
- Cross platform
- GPU processing
- Great bit-depth color support
- OpenRaster support (Krita/Gimp)
- Blender/Natron integration
- Ardour Integration
- Frame caching
- Tracking/Stabilization
- AI powered features
- Style transfer
- Segmentation
Hey, is there any plan to tackle the common crashes that come from MLT/FFMPEG/etc? Perhaps moving some stuff out of process?
Also, would be lovely to see kdenlive becoming the defacto tool for youtubers and other creators like that, the same way audacity is the defacto tool for audio.
Cheers and great work!
Hi Team,
Thanks a lot for the work done during all those events.
So many things on track :-).
Hope that I would be able to help ;-). @+
Hi,
There are plenty of stuff to do, from coding to documenting and everything else in between. What are you interesting in doing?
🙂
Thanks for the news! I love knowing any news about this project!
By the way, “Speed control via Ctrl + resize. (New feature)” refers to the playback speed on the timeline ???
very excited 🙂
and what do you mean by blender/natron integration?
“AI powered features”? Like what kinds of possible ideas are we talking about here?
PS I had to lie on the captcha to post this question. I actually am a robot.
PS If I might add a comment on the mid term goals… yes, different users are going to get excited by different things, however:
* A more granular audio workflow, would be used by some, but not most users.
* Multicam would be unused by all but a minority of users.
* Secondary color correction wouldn’t be used by anyone except a small handful of users.
Virtually *every* user though will want faster render times and will be using a machine with >1 core these days.
Don’t get me wrong, I would like (and likely use) all of these features. I hope they’re all developed. All of them pale in importance and universal helpfulness though compared to being able to use the full power of your machine to go from your edit to the finished render.
Something very similar can be said for GPU processing compared to the other long term goals.
Thanks for your consideration and hard work!
Unfortunately faster rendering and GPU processing are features needed in MLT, which is the framework that Kdenlive uses process video.
Hoo! Integration with Blender/Natron. It’s a Great news. Please implement OpenColor IO too.