I’m editing a project in .mp4 video shot at 29.97 FPS and want to speed that up only during the editing process, then slow it back to normal. It’s spoken monologue and I think it would be more efficient to do it this way so long as I can still make out what’s being said. The problem is that when I change the FPS the video changes independently from the audio so they no longer match.
Advice appreciated.
Sorry you can’t high speed scrub/playback in the VSE. The only way to speed up audio is to change the pitch (strip properties) but the cuts you allocate won’t line up with the original pitched strip.
Ah, I was afraid that might be the case. But at least I know for sure now.
Thanks for explaining.
You could pitch it up by a factor then place audio level keyframes on strip at each cut point (but don’t cut). Then restore pitch to 1. Goto dope sheet place playhead at start of strip and scale x by pitch factor. They should line up with cut points.
Thanks. I’m putting down this project for today, but will take a closer look at what you wrote tomorrow and give it a try.
I’m not really sure what’s the use of this but ideally you would need to export the audio, change its speed to 2x or 3x or whatever is still discernible for your purposes (using Audacity or any DAW app), import said audio file back into blender and apply a speed strip effect of the same factor (e.g. 2x). In theory audio should “match” the video. Now how usable is that for editing I don’t know but I wouldn’t really bet on it as the cuts will apply to the sped up version of the video so you would need to figure out how to transfer those cuts back to the proper fps. Sounds like a really messy deal.
How about activating the “draw waveform” chekbox and use that as a rough reference guide? Place markers along the timeline will enable you to segment the raw material into chunks. Btw, that’s how I usually work: during the first pass I make heavy use of markers to locate points of interest which are further trimmed and refined in subsequent passes.
Thanks blendercomp, the use of it is that I have all these multiple takes of someone speaking and I’m skimming through all this footage trying to get the best takes of everything. So for the most part I want it at normal speed, but sometimes all I want to know is whether the speaker is talking about something they mentioned already or not. In other words, has this topic been covered in one of the clips I’ve already set aside as usable. It occurred to me that I could speed up the process if I could adjust the playback speed as needed. So, I get what you’re saying exporting, changing speeds, reimporting, and all that. But it wouldn’t be practical in this case. I appreciate the idea though.
I do remeber at some point adding a driver to the pitch value that was driven from a speed effect so that as you ramped the speed the paitch altered too. Sadly that won’t help you as the strip needed to change length to function.
Ok now I see what you meant to do. Sadly Blender sucks both at editing multicam sources and at tagging clips. The good news is that it can be bent to do this kind of stuff though
Like I said the quick work around is to use markers.
Btw, you’ll be stunned to see how professional editors work: they keep extensive notes and plan the edits using pure analogue technology (paper and pencil). They review the shots using a viewer, take notes, and plan everything on paper before they move on to cutting.
I haven’t used pen and paper for 10+ years. Blender should have some basic metadata tracking for media.
lol, that’s not what other editors are saying though
Blender should have some basic metadata tracking for media.
+10000
Yeah, I am stunned. Are you sure about that? I was under the impression the process went more like this.
In what types of situations would they use the methods you mentioned?
As for taking notes I use the text editor in Blender so that I can reference the outline of my project quickly.
I do think most projects call for some accompanying notes, though not necessarily analogue.
And markers are quite helpful too.