Are you an enthusiastic video producer, a developer, a simple user or just a curious person about the Kdenlive project? Participate to our next on line meetings to stay updated, give your feedback, advises and opinions.
The first 20 minutes of our meetings are dedicated to an open discussion about any topic related to Kdenlive so everybody will have the chance to contribute to the discussion.
The Kdenlive community will be online on the Freenode IRC channel #kdenlive on Tuesday, January 16 and on Tuesday, February 20 at 9 pm UTC+1 (CET)
See you there!
Kdenlive cafés #25 and #26 – Everybody is invited
Heya Devs
Have you thought about using discord? Its been a big help for development in OpenToonz.
Link: https://discord.gg/yuuEk78
Do you have any plans for integration of other Open Sources ie. Natron, Blender, Audacity, OpenToonz, Krita (Animation) .SVG
Love your work devs
Adam Earle
Hi Adam,
There are no plans for using Discord at the moment. AFAIK the KDE project is looking for a new communication solution besides IRC, so we’ll probably tag along once that is a settled.
I would really like to see integration specially with Blender/Natron probably via some form of EDL. Unfortunately there are no devs available at the time being to implement this since all focus is on finishing the timeline refactoring.
Join us at the next café if you’d like to discuss this issues further. 😉
Cheers
dear kdenlive-team, someone made your app available as a flatpak https://github.com/flathub/org.kde.kdenlive on flathub. https://flathub.org/apps.html
It seems like you can now add the link to the .flatpakref file at https://flathub.org/repo/appstream/org.kde.kdenlive.flatpakref to the download section of your website.
This is definitely something we will discuss. Cheers
I recently figure how to make good chroma key using color selection, alpha operation and key spill mop up filters. I want to share this tutorial with the community.
https://youtu.be/hoVtJyV2Nf0
Could you share the Cafe log?
Hi Ryu,
I’m sorry for the late answer.
You can read the latest Café log here:
https://notes.kde.org/public/kdenlive-cafe25
Yes! I really like to know the details of this project. But since English is not my native language, and I do not know how to convert Cafe time to my country, I prefer to read it when it is finished.
Really, many thanks.
IRC seems a bit foo-fooey and old-fashioned… is there any chance of being present in this decade, and using Discord?
If not, I’m sure I can figure out IRC. Just wanted to put the request out there. 🙂
We have a room in Matrix for testing. We are waiting to see what the KDE project is going to choose among the more “modern” options available though before we decide on anything.
Is there anyway to have ffmpeg / mlt complied on the next release with nvenc enabled so most of us can make use of our nvidia cards gpu for rendering? Either in the release or the appimage verison?
Rick, in case you are using distribution-provided packages, you then need to ask your distro packagers. For the project’s *ubuntu ppa as well as the AppImages I’m not aware of any intention to enable nvenc support (probably because it increases maintenance complexity, but I may be wrong here; but also lack of nvideo hardware with the devs).
If I’m not mistaken, then there’s only nvenc support in ffmpeg necessary, but not in MLT — because MLT offloads the decoding and encoding to ffmpeg.
The problem with decode/encode offloading to GPUs is that you never know which profiles a specific card and driver code supports; seems that the GPU vendors only focus on what’s needed to decode the stuff that’s produced by the big media corp outlets. Encoding doesn’t seem to play any importance (IMHO), as the common YouTube and Facebook mobile video gets encoded on mobile devices with their integrated GPU ecosystem(s). Maybe with Amazon AWS and cloud deployment there may be more demand on GPU encoding in the cloud, which may improve the driver support. But that’s virtual screen streaming, not highly variant video production encoding.
Finally, you need to move a hell of frame data from main memory into the GPU, encode, then shuffle back for mass storage access. For modern multi-core CPUs encoding is a breeze, so you see instead the ugly GPU-main memory bottleneck when using nvenc and offloading. MLT will probably never support in-GPU processing unless we get a unified GPU programming model and people with money, time, and knowledge to tackle such an adventure. Because that’s still not required for the biggest market, that is gaming, I doubt I will see this in my own lifetime.
Unless you have a video workstation with a dedicated workstation GPU (meant for editing) you want to not use your graphics card for much more than playing games and doing things like antialiasing (which is needed for new OSes and all these billions of PDFs everywhere). Nvidia doesn’t care about ffmpeg’s thousands of utilities and libraries and effects, they don’t make video cards for that. If anything, normal consumer graphics cards end up being an annoyance. You want a really fast hard disk and TONS of RAM and lots of CPU power. Video editors on a PC are using your RAM/CPU, not your graphics card. There are workstation cards that can help that are designed specifically for video editing/rendering, for 3D modeling/ CAD type applications, or for image processing work. GeForce isn’t one o’ those.