- Tanner Helland (dot) Com - http://www.tannerhelland.com -

Day 5 – 10 Days of Ubuntu 10.10 Feature Requests

Posted By Tanner On September 24, 2009 @ 11:50 am In Linux and Open Source,Ubuntu | 19 Comments

Today is day 5 of my “10 Days of Ubuntu 10.10 Feature Requests” series.  See the series introduction here [1].

Day 5 – Solid, Functional Video Editing

One of the recurring comments submitted to me while researching this article was the current lack of a great Linux video editor, and I have to agree.  The last 10 years have seen video editing move from the professional to the hobbyist to the casual user realm, and people with almost no computer experience are now using software like Windows Movie Maker and iMovie to compile home videos, travelogues, and wedding slideshows.

Unfortunately, Linux remains a mixed bag when it comes to video editing.  There are a lot of video tools available, but not really a “best” option.  Lifehacker’s recent Ubuntu wishlist [2] described the problem thusly:

“In Linux, there are a range of options, almost none of them with a finished feel, and all of them front-loaded with codec, dependency, and interface headaches galore… While Ubuntu isn’t in the video software business, the many folks who contribute time, thought, and sometimes money to the project could consider this a serious missing link in the Linux application space.”

Lifehacker certainly isn’t the first [3] to bemoan the lack of a go-to Linux video editor.  You can find essays on this same topic going back many years (here’s a good one from 2004 [4], for example) – and most of them sound exactly the same: “Linux offers a lot of options, but none feel complete.”

So now, in late 2009, what kind of state do we find Linux video editing in?  In some ways, much has improved – a number of new programs have arisen in recent years, for example – but in other ways we’re in the exact same spot.  The same old programs have the same old problems (I’m looking at you, Kino [5]) and there’s still not a good “go-to” app for video editing.

But this article series isn’t about where Ubuntu is today – it’s about features we want to see by the time 10.10 rolls around.

To that end, I’m going to give a quick assessment of 5 promising Linux video editors – Cinelerra, Kdenlive, OpenShot, PiTiVi, and Kino – and what I think the likelihood is that any one of these will become “the” Linux video editor.  (Note: there are many other options for Linux video editing, but I think these 5 are the most likely to fit the bill for this article – possibilities for default inclusion in Ubuntu 10.10.)

Cinelerra

Cinelerra is arguably the biggest, baddest video editor you’ll find in Linux.  Its feature list is impressive [6], it offers both a corporate and community supported version, and its screenshots [7] make it look competitive with most multi-hundred dollar commercial offerings for other OSes.

The downsides?  It’s user interface is ridiculously confusing, it has a very complicated licensing structure [8], and to use it in Ubuntu you’ll probably need to compile it from scratch.

Verdict: no way does Cinelerra fit into the same category as Windows Movie Maker and iMovie.  It’s only for the hard-core and extremely patient, neither of which seems to be a key demographic for Canonical. Also, the Lumiera rewrite [9] is a promising idea, but until it becomes stable it’s impossible to predict how the effort will payoff.

Kdenlive

Kdenlive is arguably the most popular video editing option for KDE users, as many distros include it by default.  Its interface integrates nicely within the KDE environment, and even beginners should be able to make sense of the default layout. Kdenlive also supports a wide variety of formats via the ffmpeg library, direct video import, active development, and – gasp! – it even has a user manual [10].

Downsides? For Ubuntu users, the KDE factor makes it an unlikely default choice. Stability issues [11] are an ongoing factor, no gstreamer, and at least in my experience I’ve found the program unfortunately sluggish.

Verdict: Kdenlive is a promising program with a ton of potential, but too many dealbreakers exist for it to be a likely contender for Ubuntu 10.10′s default video editor.

OpenShot

OpenShot has a classic FOSS story – a developer tries Ubuntu, likes it, looks for a video editor, can’t find one that fits his needs, so he resolves to write his own.  Most of those stories end with “it didn’t work out,” but not OpenShot.  This month the young project passed the 10,000 mark [12] for its .DEB installer and development continues at a frantic pace [13].  OpenShot’s feature list is shaping up nicely [14], the developers quickly comment on every bug and feature request in Launchpad, and the use of gstreamer and GTK is ideal for Ubuntu inclusion.

Downsides?  OpenShot isn’t available via PPA, which has prevented large-scale testing (though one is supposedly in the works).  It’s written in Python – a beautiful language, but bytecode-implemented and notoriously slow, not to mention that multi-threading in Python typically requires use of C extensions [15] (not sure how the OpenShot team feels about that).

Verdict: OpenShot is a strong contender for a default Ubuntu 10.10 video editor, but much hinges on the type of large-scale testing that’s just beginning to take place.  Also, if multiprocessor support cannot be easily supported, performance may become an issue.

PiTiVi

Our next strong contender comes in the form of PiTiVi, another GTK+ and gstreamer-based video project written primarily in Python. PiTiVi has the strong bonus of being sponsored by Collabora [16], a company who has already contributed enormously to gstreamer, telepathy, and other key components [17] of the modern Linux desktop (thank you, Collabora!). PiTiVi has the advantage of lead developers with proven gstreamer [18] experience, a very smart modular design, plus a PPA for individuals interested in testing.

Downsides? For a five-year-old project, it’s unfortunate that transitions, titling, and effects are still on PiTiVi’s wishlist [19].  PiTiVi’s gstreamer integration is second-to-none, but without these basic tools it is currently little more than an experiment in the most basic elements of video editing.

Verdict: I have huge hopes for PiTiVi, but development pace continues to be a concern.  The 0.13 line has brought some welcome improvements [19], but the likelihood of PiTiVi reaching feature parity with other entry-level video editors by Ubuntu 10.10 is very, very slim.

Kino

Last but not least, we have classic Kino – one of the most commonly recommended entry-level video editors for Linux.  Kino is built with GTK+ but not gstreamer and is heavily based upon the premise of importing DV via firewire.  The latest version (1.3.4 as of this writing) has a solid feature list [20], and Kino boasts the most comprehensive user manual [21] I’ve seen for a Linux video editor (although it’s only accurate through v1.0).

Downsides? Kino development has lapsed greatly since its primary contributor switched his focus to MLT [22] (an excellent decision on his part, IMO). An aging codebase and architecture make it an unlikely candidate for forking, and my experience with the software has been plagued by frequent freezes and/or crashes.

Verdict: Kino continues to provide service to the Linux community as a simple entry-level editor, but its time has passed. Unless forked and/or rebuilt largely from scratch, I think Kino’s future contains only basic maintenance, not innovation.

Conclusion

Despite my complaints at the start of this article, there is actually a lot to be excited about when it comes to Linux video editing.  GStreamer [23] and MLT [24] in particular have made the likelihood of an excellent Linux video editor much more probable, and within the next 12 months I expect great things to happen in this arena.

In my opinion, OpenShot and PiTiVi are the most likely contenders for Ubuntu 10.10′s video editor of choice.  OpenShot offers a great vision, growing community excitement, and a rapidly-progressing feature set. PiTiVi offers a technically advanced backend, excellent architectural design, and a team of proven developers.

Which will win out?  I’m no psychic, but if I had to bet – my money’s on OpenShot. :)

But this is open source we’re talking about, so things can change rapidly.  Personally, I’d love to see PiTiVi, Kdenlive, and OpenShot all blossom into solid, proven projects by the time Ubuntu 10.10 rolls around.

For more information, each project’s homepage is available here:

And a more complete list of available Linux video editors can be found here:

<< Day 4 – Real Wine Integration [31]

Day 6 – Simple, Reliable, Integrated Backup Tool >> [32]


Article printed from Tanner Helland (dot) Com: http://www.tannerhelland.com

URL to article: http://www.tannerhelland.com/1283/day-5-ubuntu-10-10-features/

URLs in this post:

[1] here: http://www.tannerhelland.com/ubuntu-linux/10-ubuntu-10-10-features/

[2] Lifehacker’s recent Ubuntu wishlist: http://lifehacker.com/5355900/five-features-we-want-to-see-in-ubuntu

[3] isn’t the first: http://lunduke.com/?p=429

[4] here’s a good one from 2004: http://www.osnews.com/story/6443/Video_Production_with_Linux/page2/

[5] Kino: http://www.google.com/search?q=kino+linux+crashes

[6] Its feature list is impressive: http://heroinewarrior.com/cinelerra.php

[7] screenshots: http://heroinewarrior.com/cinelerra_shots.php

[8] very complicated licensing structure: https://wiki.ubuntu.com/UbuntuStudio/CinelerraReview

[9] Lumiera rewrite: http://en.wikipedia.org/wiki/Cinelerra#Lumiera_rewrite

[10] user manual: http://www.kdenlive.org/user-manual

[11] Stability issues: http://www.google.com/search?q=kdenlive+crashes

[12] the young project passed the 10,000 mark: http://www.openshotvideo.com/2009/09/over-10000-downloads-served.html

[13] development continues at a frantic pace: https://launchpad.net/openshot/

[14] feature list is shaping up nicely: http://en.wikipedia.org/wiki/OpenShot_Video_Editor#OpenShot_Features

[15] multi-threading in Python typically requires use of C extensions: http://www.python.org/doc/faq/library/#can-t-we-get-rid-of-the-global-interpreter-lock

[16] sponsored by Collabora: http://www.linux.com/archive/feature/152464

[17] other key components: http://www.collabora.co.uk/projects/

[18] lead developers with proven gstreamer: http://gnomedesktop.org/node/4082

[19] transitions, titling, and effects are still on PiTiVi’s wishlist: http://www.pitivi.org/wiki/Roadmap

[20] solid feature list: http://www.kinodv.org/article/static/2

[21] the most comprehensive user manual: http://www.kinodv.org/docbook/

[22] its primary contributor switched his focus to MLT: http://www.kinodv.org/article/view/157/1/7/

[23] GStreamer: http://www.gstreamer.net/

[24] MLT: http://www.mltframework.org/twiki/bin/view/MLT/

[25] Cinelerra: http://www.heroinewarrior.com/cinelerra.php

[26] Kdenlive: http://kdenlive.org/

[27] OpenShot: http://www.openshotvideo.com/

[28] PiTiVi: http://www.pitivi.org/

[29] Kino: http://kinodv.org/

[30] http://linuxappfinder.com/multimedia/videoeditors: http://linuxappfinder.com/multimedia/videoeditors

[31] << Day 4 – Real Wine Integration: http://www.tannerhelland.com/ubuntu-linux/day-4-ubuntu-10-10-features/

[32] Day 6 – Simple, Reliable, Integrated Backup Tool >>: http://www.tannerhelland.com/ubuntu-linux/day-6-ubuntu-10-10-features/

All text copyright © 2008 Tanner Helland. To reduce paper and ink usage, all comments, images, and videos have been suppressed on this printing.