Is it possible to have for example a sequence that is 2160P, and then create another that is 720P?
If not, this is a much needed feature and I hope it’s on the roadmap.
It allows to edit different sections in their native resolution without needing to mix formats and the assembling them together at the end to the final resolution.
It’s also necessary when, for example, you create a piece that will end up being used on different platforms, like Youtube (horizontal) and TikTok (vertical) but the cut is the same. You simply edit once, then bring that into another sequence that is vertical and position/scale etc.
Sequences are part of the open project so their resolution/dimensions and fps etc. are the same across all sequences.
In order to use the same timeline for outputs to different platforms, you can use the Re-scale feature and/or the Aspect Ratio selection in the render dialog.
Once Kdenlive handles each asset with its original dimensions (i.e. stops squeezing them into the project dimension) this may be possible, or at least easier. But otherwise I don’t think this is
a much needed feature
Do other video editors have this feature? I know for sure it’s not on the roadmap.
The ‘aspect ratio’ option was added relatively recently, ostensibly to support this sort of One Cut Multiple Render Formats use - but the main issue with it is that it just blindly crops the source material to that aspect ratio …
So unless you’ve edited your source to fit the subject into just the central intersection of all your formats (which compromises them all to fit badly), things that you might have wanted visible are just going to fall out of the frame.
You might be able to create one ‘master sequence’, and add further manipulations to it on nested or duplicated sequences, but really you’re still going to need one sequence suitably positioned/cropped/zoomed for each output format that you want to render, in all but the most trivial cases.
thanks for your answers, I’ll do some experimenting. And as for the question of if other software does this, yes. Premiere, After Effects, FCP etc do this.
You would simply drag your clip onto a “new sequence” icon and it would create the sequence to match the framerate and dimensions of that clip. i would assume it would then require less calculations when working with it … since the footage is not upscaled or downscaled to match the sequence of a different size.
Ultimately I can used multiple sequence resolutions within a project. So the project doesn’t need to be fixed and you decide what format each sequence requires and output as necessary. For social clips if I drag a 1920 sequence into a vertical sequence it basicallly clips the wings. If I want to reposition, I do cut it on the timeline into pieces and move each as needed.
At risk of sounding like someone that uses proprietary applications and expects the same behaviour here, this is a screenshot from Shotcut, which doesn’t ask for a project frame rate, but creates one based on the first files you import, though I see that Kdenlive does give a dialogue to change it … regardless I don’t see the fixed project resolution as important, these decisions should be made by the ingested clips and the desired export
Nah, you’re not sounding that way, or at least not to me …
We live in a world where multiple frame sizes and rates is becoming the norm (even for footage shot on a single outing with a single camera). That might have been crazy talk, that nobody sane or professional would have imagined needing or wanting a few years ago, but well, here we are
I’m really looking forward to the day kdenlive hits its roadmap targets for improving how it handles source clips different to the project profile.
I think I actually prefer the way kdenlive handles this (selecting the project format when you create a new project, then warning you of a mismatch with the first clip added and offering to switch but not forcing you do or just silently doing it anyway).
If the first clip I add was shot at 4k@240fps, I probably don’t want the whole project edited/rendered at 240 frames per second, I want that one to be a slow motion clip rendered at a more usual viewing frame rate. Likewise if the intro title shot was in a different format to the main presentation.
I think there’s three profiles of interest for a project:
The project settings (which I agree with the feature request here that making that configurable per-sequence would be a Good Thing).
This is the format and reference frame for all manipulations that you are going to do on the timeline.
The source material format. Right now when this is different to the project format it is automatically massaged to fit the project frame, which is the source of several flavours of pain - but in the ideal world (and on the roadmap), the original size would be preserved and you would need to manually transform it to fit the project frame however you wanted to make it fit.
The output format for rendering. In the usual case this should default to the project settings, but there’s no reason that it always should. Most of my projects are all 4k, editing 4k source at various frame rates - but most often they’ll get rendered at 1080p. The existing Rescale option makes that easy. Being able to select any profile for output might be nice, but non-integer changes to a different frame rate or similar would make that a bit messy …
The aspect ratio rendering override definitely would be more useful if you could have a separate sequence manually transformed to fit that format how you want rather than just blindly edge cropping a different format sequence to fit.
It would also be nice to have a ‘virtual frame’ working space that can be larger than the visible project frame - right now if I add a transform to a clip and move it 500 pixels to the right, then add another transform and move it 700 pixels to the left, I lose the 500 pixels that get cropped after the first transform and get a 700 px black bar down the right hand side instead of just a 200px one.
That might seem like an insane self-inflicted injury - but the real use case for stacked transforms like that is things like using motion tracking keyframes to keep a selected subject steady in the frame (importing the inverted tracking position). Right now to make that work you generally need a lot more zoom that you otherwise might to hide what was ‘unnecessarily’ cropped by the initial imported transform.
