I’m going to record some videos soon (some interviews, and some scenery to use as background for the interviews). I will use blender to add a bar with the name of the person at the bottom of the screen, and maybe to do some color correction and greenscreening.
I’m also wondering about what kind of workflow or approach would be best.
What would be the best recording format?
What should I use as intermediate format for the video? (when slicing up the footage into parts)
Should I actually slice the footage up into parts? or use the whole strip several times but with different start- and end frames?
I record the audio with a separate microphone. Where in the process should I merge/sync it with the video stream? before or after the montage?
Where in the process should I do the color correction? at the beginning (the original footage)? or just before combining it into the final movie?
Re audio - normally recording dual system audio, you would sync the audio in your video editor before editing, then you’ve not got a nightmare to deal with later on, because they remain in sync throughout the edit. The audio can then be exported and final edits and mix done fairly painlessly.
Colour correction is normally done late in the process (usually the last step before sound mix I believe). And done shot by shot. In dedicated programs it’s easy to copy colour settings from one shot to another and tweak from there. Not sure of the best blender workflow though. If you think about it, you don’t want to be spending time correcting footage that may not make it into the final edit.
The other questions sort of depend on what format you’re shooting on, and how you’ll be editing… For the greenscreen, you’ll want to shoot in as minimally compressed a format as you can afford. More info??
If using Blender to edit (I would advise against this), try to avoid syncing your seperate audio and video via a Meta-strip. It doesn’t seem to work to well at the moment. That is when you cut it up, the audio outside the cut point (the non-existent sound) bleeds into the previous shot. Don’t know why, it was supposed to have been fixed earlier.
To sync it obviously you should do that at the start of the video, on a clapper snap in shot or a hand clap. After each pause of shooting you need to do another sync point and lay that up against the new clip. You then cut these at the same time and copy them to the new edit location in the master.
Check the blog in my sig there are some editing tips for Blender. But the best one is “edit it somewhere else”, really.
it depends. Blender can handle pretty much everything thanx to ffmpeg.
But do you have the luxury of recording in any format?
What should I use as intermediate format for the video? (when slicing up the footage into parts)
it depends. For example I convert my canon 550D quicktime movies into DNxHD.
Using proxies might also be a possibility. Finally, you could convert your material into stills and work with these.
Should I actually slice the footage up into parts? or use the whole strip several times but with different start- and end frames?
This is non linear editing so you could do both soft and hard cuts.
I record the audio with a separate microphone. Where in the process should I merge/sync it with the video stream? before or after the montage?
early on, ideally before you begin cutting because it might not be possible (or very laborious) later on.
Where in the process should I do the color correction? at the beginning (the original footage)? or just before combining it into the final movie?
That’s also doable in Blender. It’s called an adjustment layer. Add one in the VSE over an image/video strip, grade that, and clone it over the other strips. Depending on the shot type, it might require minimal tweaking.
Not sure of the best blender workflow though. If you think about it, you don’t want to be spending time correcting footage that may not make it into the final edit.
Blender’s VSE offers only basic color correction at the moment (3way color wheel). It is sufficient for primary and secondary color corrections though. It’s trickier in case you need masks but it still doable within the VSE. For more advanced tools (say RGB curves, Hue Correct etc) and greater control one should turn to the compositor.
Whenever I need more control than the 3way color wheels affords I render the specific frames from the shots after editing, grade in the compositor and replace the originals (be it image sequence or movie strips) in the VSE before the final render.
Btw, please note that what one can accomplish grading-wise with the simple 3 color wheel tool is surprisingly remarkable!
I am not suggesting that you buy another app there are more capable free ware editors. It’s just that blender is a bit buggy when using the vse. I have had timelines get corrupted and lost hours of work. Do try to iterate your saves so you can return to an earlier version if needed. If you look at my blog below there really are some useful tips.
Thanks for all the input.
I think the main order I understand from this, is:
record the video and audio
sync them (Mux the audio to the video stream)
Do the montage (either soft- or hard-cutting) and the greenscreening
Adjust the colors
render to final movie+audio
@3PointEdit: which one would you recommend? I’ve experimented with Pitivi and OpenShot, but I was not really happy with them. With Pitivi I couldn’t fade between shots (or I just couldn’t find the feature), and I don’t really remember why I quit OpenShot. I might give it a retry.
@Hammers: I’ll basically have three kinds of shots: Interview recordings, Interview recordings with greenscreen, and scenery shots (for the background).
@blendercomp: We don’t have a videocamera yet. We’ll either buy or rent one. So I’m also interested in what video format would be best, so I can take that into account when choosing a camera.
Didn’t realize that you needed to do greenscreening. In that case the greenscreening should be done before cutting as it’s done in the compositor and not within the VSE (AFAIK the sequencer plugins are not functional).
Adjust the colors
render to final movie+audio
@3PointEdit: which one would you recommend? I’ve experimented with Pitivi and OpenShot, but I was not really happy with them. With Pitivi I couldn’t fade between shots (or I just couldn’t find the feature), and I don’t really remember why I quit OpenShot. I might give it a retry.
Once you’re on linux and you want to experiment with other non-Blender softwares, IMHO Kdelive is the best alternative.
@Hammers: I’ll basically have three kinds of shots: Interview recordings, Interview recordings with greenscreen, and scenery shots (for the background).
VSE can handle all these without problems. Working with image sequences will most likely be the safest route.
@blendercomp: We don’t have a videocamera yet. We’ll either buy or rent one. So I’m also interested in what video format would be best, so I can take that into account when choosing a camera.
I see. Based on my experience, it’s hard to suggest this or that format really. Like I said Blender can handle almost everything. All consumer-level recordings are container format variations of h.264 which is highly compressed and not suitable for direct editing. Converting to an intermediate codec or using image sequences instead is a must.
Blender does not like AVCHD from Sony. I ended up cutting that format in Windows Movie maker!
You may have to much greenscreen. My sugesstion would be to cut/select the shots you really need then lift them from the timeline and render to a keyable strip then alpha over another strip in the VSE (means that you can slide the background as required).
