Dev:Ref/Requests/Sequencer/Done

Features in SVN or Release

Title	Replace imbuf-Scaler
Summary	The imbuf-Scaler is very slow. It should be replaced with ppmqscale-code.
Link	Already done in SVN --Peter Schlaile 18:27, 6 August 2008 (UTC)

Title	Floating Point Precision FPS
Summary	Enable floating point precision on FPS as per Otomo. This has been enabled for some time. As such, I am removing this from the primary requests page. See the section titled "True NTSC and 24p-support" in the associated link.
Links	http://www.blender.org/development/release-logs/blender-246/sequencer-changes/ --Troy James Sobotka, 2008-09-27

Title	Add Theora Video and Vorbis Audio Support
Summary	Add FFMPEG-output to theora video and vorbis audio in ogg and ogm container formats.
Links	svn-16936 --mindrones, 2008-10-06

proxies

• works incredibly fast - how do they do it? Has fast preview, going through video, fast thumbnail generation. I know Vegas makes some temporary files, which maybe enable this...Today you cannot see the difference so much on new machines, but on a 500 mhz processor it's really visible.

Hmm, that was one of the features, that I definitely _liked_ about Blender. Nevertheless: I have been already told, that preview editing and working with EDL-files is mandatory in professional environments... To make it short: will include a feature, to import video strips at lower resolution maybe even directly from tape. (Peter)

It would be great to have some kind of proxy system, where a clip can have multiple sources. I already had this problem when I was editing my reel in Blender, my footage was at 1024x576 and that's what I wanted to produce the final video at, but my computer isn't fast enough to play that back in real time. So the only way around it for me, was the hellish procedure of saving out half-res versions of each strip, importing them into Blender and editing them, then switching over at the end when I wanted to render the final video. It was terrible!

It would have been much easier if each clip could have a high-res and low-res version defined, then a global 'display type' setting for the entire sequence that you could use to switch between low res and high res. That would still be really cumbersome, taking lots of unnecessary manual labour to make all the separate high and low res clips, so even better would for the computer to do it - i.e. select a bunch of clips and use a 'make low-res proxies' command, which would then generate (by scaling down copies) and link up low-res proxies for each selected clip. --Matt Ebb

snap

When dragging strips around in ULEAD, it will snap to clips when they come close to a corresponding strip or to the frame pointer. So if you have a bunch of cuts you can easily let one clip start directly after the previous clip without having to look at frame numbers to get an exact match. javawocky

cuts

Ability to set in points and out points. In Blender, you can just move the arrows on the ends of the video, but if you want to extract a 5 minute scene from a 2 hour movie, that's a lot of arrow moving. directedition

• You could just cut the strip (move frame pointer to desired point and press 'k') in half and delete the tail piece. I do it in ULEAD ;) But being able to have an exact point would be advantageous. javawocky
• I agree here, we need some better cutting features. Maybe this will have to come after asset management though. tbcpp

preview

• There is also an option to show "safe frame" margins in the video preview window. Raschko

logic of the sequencer

• It is at present very confusing to see, which strip is currently the output strip. In case you don't know: If there is an effect strip present, we take the upper most visible non-audio strip. If there is no effect strip present, we use the lowest strip. This is really confusing. intrr suggested just making every movie or metastrip, that is not input of an effect strip implicitly alpha-over. The strip-IPO is the global ALPHA-value. Additionally one must add the ability to mute a video strip, so that it can be used as a storyboard-strip explicitly. The whole idea of this is, that the sequence editor works more intuitively like the gimp (or even consistent with it's own audio strips!), which also layers all available strips on top of each other. Most people most likely won't notice, since they use at least one effect strip anyways, which will render the same result. If you have speed concerns: I have already a way, that makes this run as fast as the current sequencer. Peter Schlaile
  • What does this mean? How do we get it into a feature format? Sobotka 05:47, 1 July 2008 (UTC)
  • Well, nevermind. It is in SVN now :) --Peter Schlaile 18:28, 6 August 2008 (UTC)

Checking image renders with audio

• This isn't directly sequencer related, but when playing back animations and preview (hardware) renders, I'd like preview audio to also be played. Currently my workflow is render to PNG (not quicktime because this way I can replace frames more easily), load the image sequence and the mixdown in quicktime, copy the mixdown and paste it into the image sequence, play, then go back to Blender and tweak. Charles 10:29a 29 Nov 2006
• Audio sync / scrub should work. Or are we talking different parts of Blender? --Peter Schlaile 19:02, 29 November 2006 (CET)
  • I think he wants to sync a 3D animation with the audio in the sequencer. In that case the solution is using the Scene strip no? Or, if the audio is in the Audio view this isn't related to the sequencer. mindrones 2008-08-08