Blender Video Editor And Compositor


I figure that most development should be focused on the actual 3d aspect of it… when there are so many better programs to do them in IE. Shake,Nuke,Inferno,Fusion,Avid MC, FCP

Any reason why its better to have them in blender?

What is the one major thing that makes blender different from Shake,Nuke,Inferno,Fusion,Avid MC, FCP?
Hint: I use blender but not Shake/Nuke/Inferno/Fusion/Avid etc because I can’t aford those programs :slight_smile: Make sense?

that and blender runs on all major os(windows, mac, linux and solaris) platforms, you also dont have to render everything out to go between ones 3d scene and the nodal compositor

obviously those programs are cost prohibitive. i can see doing small comps in blender but when you work with a few hundred elements in a shot that may sometimes be 2K blender just cant do that. as for the video editor i really cant say much in its favour.

yes i am a professional and i use blender for a significant percentage of my 3d work even though i have Maya and Houdini. most of the time my stuff ends up in maya for the final render because blender still needs work there but i find it much faster for modelling, scene layout and the like.

The Biggest Reason is full and utter intergration with the renderer, no dicking about with dozzens of images/ files, and in the sequencer you can use a scene strip and it will only rendeer the scene when you render the sequencer, its just saves a hell of alot of hassle and time, and with it we have Full 32bit throughout only nuke does that, shake and others convert to 8bit and back all the way through your composite.

PS, don’t think general 3d stuff is being ignored, thats far from the truth.

I don’t get why trolls start these threads. why? Why come onto a free site that supports a free software package and criticize it for not being as good as commercial packages, some of which cost thousands of dollars per seat. why?

(oh, and add a) the ability to do it for free, and b) no context swapping, all in one package advantage as well)

mfoxdogg is the sequencer full float? I thought there was some colour space issues going through there. I dont think that imported video is bumped up to float space for manipulation.

sequencer isn’t compositor is

This must be some kind of a ‘reverse phychology’ technique,
to try and generate development in the sequence editor :smiley:
I think that any one who’s ever made any edited animation with blender loves the sequence editor,
although I didn’t even realize that it was 8 bit… Sometimes I saw little gradations,
like if the ‘World’ background is a 2 color blend fade, but I just assumed it was from the compression.

Sequencer is fully float image aware. You can mix 32 bits and float freely, output then will become float too.

When an image has float colors, all imaging functions in Blender default to use that. This includes the Video Sequence Editor, texture mapping, background images, and the Compositor.

because they have a demand on software like nuke/shake/whatever and want it for free. kinda lazy.

i have to admit, i would love to see a foss compositor/sequencer beyond blenders capabilities too and as long as the critique is constructive and not redundant i don’t mind people posting like this.

Does Blender have option to show images as full, half, third and quarter like After Effects and other common Comp softwares?

Then continue to use those apps and maybe one day if a developer has an interest in developing blender’s VSE or Compositor further then blender might have more to offer than it already does. Rather than a lame ass attitude as to why blender should even have those things. :wink: Personally I’m very grateful. I’m not a professional and I don’t have cash to splash. But I’d still like to enjoy some of the high end features like high precision and the beginnings of a linear workflow that even well established NLE’s and some compositors still don’t have. Sure no plethora of filters and presets but so what.

as for the video editor i really cant say much in its favour.
32bit Float, Linear workflow and basic editing / grading functions.

yes i am a professional and i use blender for a significant percentage of my 3d work even though i have Maya and Houdini. most of the time my stuff ends up in maya for the final render because blender still needs work there but i find it much faster for modelling, scene layout and the like.
Good for you, enjoy, you get the job done so why fuss over why blender does or doesn’t this and that.


same reason why 3dsmax has Video Post

Haha :smiley: Video post is something from 90’s :wink:

Blender has powerful effects features, they are just locked up in the Node compositor, where interlace fears to tread! I just round tripped some colour correction work from FCP to Color and that sucks. Dragging a scene in is waaaay easier!

Nuke is but ugly, but fast… updates are kinda slow.

Name a Suite that updates as fast as blender, and has as cool/good developers ?
It’s like a party at the rich kids parents mansion… it’s gonna be awesome look nice but you cant go nutz and have a blast… well in blender party everything goes… I mean everything ^^ it’s like the art party some dudes throw together under a brigde in some weird place of the city and you go there and everything is ‘oh snap!’ they got a projector doing arty stuf at the walls. there’s free booze and wine. Kickass.

Yellow have you confirmed that the sequencer is all float now? I thought that you were seeing truncated colours, that was from sources so it could be the codec, what about blender internal images?

The sequencer allows us to convert a file to float by pressing the float button, bit of a pain with a few hundred inputs, a bit like the deinterlace but flexible I guess with differing inputs.

I don’t think blender reinterprets the values (filling in the gaps), unfortunately the histogram doesn’t work when a file is converted to float so can’t see the result without rendering out.

But this is all after FFMPEG has done the YCrCb to RGB colour conversion, whether the colour space conversion has been done to high precison I don’t know, it certainly scales the video inputs, causing what wasn’t clipping to clip, if one has videoed with care, just under the 100% zebra. Blacks below 16 crushed to 16 , super whites above 235 crushed to 235 and then both remapped 0-255 in RGB or just truncated and no chance of recovering them.

I don’t think the VSE is included in the 2.5 Colour Managment stuff, so the VSE doesn’t so far linearize the video inputs. Hopefully that will change or maybe it has already.

Going through nodes to scenes on the VSE, is not workable for video editing. Hopefully the VSE will get some love in 2.5.

I think it would be good to see video files converted from YCrCb to RGB at high precision, not scaled 0-255 (as an option) and option to linearize in the import to the VSE, rather than a secondary conversion to Float and linear after FFMPEG has finished with it.

Blender is not a media player so why does the 16-235 scaling happen arbitrarily, without choice, you loose information the camera has captured.

Although I’m more than happy to be told it’s pointless and I’m barking up the wrong tree but AVISynth conversion appears to be far better than blenders handling of it.

Yellow I was just reading on the Prolost blog that Apples Color also clips RGB files, like the QT files generated on the new Canon DSLRs! Imagine having beauifull images with lots of headroom crippled by the import process. Apparently the app. cant force reinterpret RGB as YUV.