Screencasts - Cheese

Jeff Spaleta jspaleta at gmail.com
Tue Jun 10 23:10:23 UTC 2008


2008/6/10 Clint Savage <herlo1 at gmail.com>:
> I agree, wholeheartedly.  My current problem is the lack of knowledge I have
> about gstreamer, and any other free video editing tools.  I've heard of
> fluendo, which I'm guessing is proprietary, and cheese with gstreamer.  What
> else is out there that we can use to edit ogg/theora video, and how can we
> make it possible to get ogg support as a whole available to embed within a
> browser?  I mean this shouldn't be something too difficult to put into
> epiphany, konqueror or firefox, should it?

For the time being stick with very simple video tasks that need very
little video editting.  if you can do it in one sitting with cheese or
istanbul do it that way for now.  We can also convert raw dv into
compressed theora video as well out of the box if you want to work a
dv camera. The pitivi application used just to convert dv to theora
with no editting basically works.  Though it's also possible to just
create a simple shell script using gst pipelines to convert raw dv
into theora, I should probably add that if someone else doesn't figure
out how to do that first by adapting the audio splice script.

I think there is room to build out a series of cookie cutter pipeline
scripts to do common tasks once we have a group of people want to
seriously poke at what gst lets you do out of the box.. but we aren't
there yet.  For example all the video affects that cheese does should
be applicable to gst pipelines to post-process video clips.  I think
gst even has a series of transitional affects like screen wipes and
fades that we can access via pipelines... i think. I haven't figured
out how to do it yet.

if you need to play with basic clip sequencing inside of an
application interface you can attempt to use pitivi. My understanding
there is a GSoC project related to pitivi this summer to spruce it up.
'We' should watch pitivi development very closely. Even if it doesn't
change direction and turn into a pro-sumer video tool...as a light
weight video sequencer and editor it may be enough for community
generate content to keep content flowing for our needs.  it just needs
a little love to get there.

I can't stop you from reaching for the more mature editting software
that is ffmpeg based and available in 3rd party repositories.  if your
employer needs to create higher production quality videos than what
you can do with the in distro tools, then you are free to do so.  I
had looked at kino at one point, but it does everything as raw dv. I
could have stripped out the ffmpeg exporters and some optional
features and gotten it into Fedora using a gst based theora exporter.
But since it doesn't handle mixed source content, its not necessarily
worth the effort to do. You can get a fully functional kino from 3rd
party repos that deal with ffmpeg.  I don't expect people to really be
able to collaborate by passing raw dv files as source material.  We
need to be able to work with lower quality content as source to get
our feet wet with working with video collaboratively. Pitivi lets you
mix theora and dv together as source material. Video purists will
cringe at the concept of doing that..but I'm okay with it to keep from
having the hosting burden of dealing with raw dv until we have a
clearer idea wtf we really want to do with collaborative video
production.  If kino could use theora as source material and could be
shipped in such a way that ffmpeg support could be optionally added
via an addon package... I would continue to work on getting kino
packaged in such a way that we could offer it out of the box.  But I
never identified a gstreamer module which let me convert back to raw
dv to use as a theora importer in kino to convert theora to raw
dv..which is what kino works with internally. Again video purist will
cringe at the concept...but it is what it is.  ffmpeg has the ability
to do this lossy conversion, but for the life of me I can't find a way
to get gst to do it.

I have no problems accepting appropriately licensed content in formets
we support made with tools we can't ship.  I understand that I have
multiple competing goals for community generated video content.
Dogfooding the open tools while producing high quality open content
won't be achievable simultaneously for a while until the shippable
tools get better.  But I will make it a point to have a policy which
highlight videos that are built solely using Fedora tools as the
become available to encourage people to continue to dogfood and
develop the tools.  The landscape is not going to get better until we
have an important reason to use video as a tool to help us achieve
larger project goals.

-jef




More information about the marketing mailing list